NUCLEAR ENERGY AND SUSTAINABLE ...

KATHOLIEKE UNIVERSITEIT LEUVEN FACULTEIT TOEGEPASTE WETENSCHAPPEN DEPARTEMENT WERKTUIGKUNDE AFDELING TOEGEPASTE MECHANICA EN ENERGIECONVERSIE Celestijnenlaan 300 A, 3001 Heverlee

Erik LAES ⎯ NUCLEAR ENERGY AND SUSTAINABLE DEVELOPMENT⎯ Oktober 2006

NUCLEAR ENERGY AND SUSTAINABLE DEVELOPMENT Theoretical reflections and critical-interpretative research towards a better support for decision making

Promotor : Prof. Dr. ir. W. D’haeseleer Prof. Dr. ir. R. Weiler

Proefschrift voorgedragen tot het behalen van het doctoraat in de toegepaste wetenschappen door Erik LAES

Oktober 2006

SAMENVATTING Erik Laes KERNENERGIE EN DUURZAME ONTWIKKELING. Theoretische reflecties en kritisch-interpretatief onderzoek naar een betere ondersteuning van de besluitvorming. Proefschrift aangeboden tot het verkrijgen van de graad van Doctor in de Toegepaste Wetenschappen. Promotor:

Prof. Dr. ir. W. D’haeseleer Prof. Dr. ir. R. Weiler

Oktober 2006

In dit proefschrift bestuderen we de rol van kernenergie in België in het licht van een brede maatschappelijke vraag naar meer duurzame ontwikkeling. Hoewel het concept ‘duurzame ontwikkeling’ reeds een lange ontwikkelingsgeschiedenis heeft doorgemaakt, verkreeg het pas in 1987 een ruime internationale weerklank met de publicatie van het rapport “Our Common Future” (ook wel het ‘Brundtland-rapport’ genoemd) door de Wereldcommissie voor Milieu en Ontwikkeling. Een concrete uiting van deze bezorgdheid om duurzame ontwikkeling vormen de internationale inspanningen om de wereldwijde uitstoot van broeikasgassen te verminderen, die het energiebeleid weer hoger op de (mondiale) politieke agenda plaatsten. De ‘V.N. Conferentie inzake Milieu en Ontwikkeling’ (Rio de Janeiro, 1992) en het daar ondertekende klimaatverdrag (‘Raamverdrag van de Verenigde Naties inzake de Klimaatverandering’) – gevolgd door het ‘Kyotoprotocol’ (1997), dat concrete emissiereductiedoelstellingen vooropstelt voor de ondertekenende landen in de periode 2008-2012 – vormen de fundering van de globale besluitvorming rond klimaatverandering. Maar het concept ‘duurzame ontwikkeling’ gaat verder dan enkel bekommernissen omtrent milieu of klimaat. Vele interpretaties blijven de ronde doen, al lijkt er toch een minimale consensus te bestaan dat duurzame ontwikkeling een geïntegreerde benadering van de economische, milieu-, sociale en institutionele componenten van ontwikkeling inhoudt, en dit met het oog op billijkheid binnen de huidige generatie en tussen generaties onderling. Duurzame ontwikkeling stelt de huidige vaak eenzijdige nadruk op economische groei in vraag en verlegt de aandacht naar de mogelijke gevolgen voor mens en milieu, zowel op korte als lange termijn. Bovendien moet bij het inschatten van deze gevolgen steeds een voorzorgsbenadering in acht genomen worden. Hoewel een consensus op het niveau van algemene principes haalbaar blijkt, is de toepassing van deze principes in concrete beleidsvraagstukken echter vaak omstreden.

ii

Kernenergie en duurzame ontwikkeling

1 Doel en relevantie van het onderzoek Doel van dit proefschrift is het debat rond de (vaak controversiële) invulling van ‘duurzame ontwikkeling’ op het gebied van energievoorziening in de Belgische context, en in het bijzonder de rol van kernenergie daarin, in eerste instantie te verhelderen door een systematische en geïntegreerde evaluatie van zowel filosofische, theoretische als praktische benaderingen in dit domein. In tweede instantie worden de resultaten van deze theoretische evaluatie aangewend om een concreet beleidsmodel voor te stellen dat meer geschikt is om de complexe vraagstelling rond duurzame energievoorziening het hoofd te bieden. Deze brede doelstelling omhelst volgende meer specifieke onderzoeksdoelen: 1. het ontwikkelen van een filosofisch ‘standpunt’ of ‘houding’ (namelijk het constructivisme) als uitgangspunt voor verdere theoretische en empirische beschouwingen; 2. een kritische test van dit filosofische standpunt door een confrontatie met andere gangbare perspectieven (met name de risicomaatschappij-theorie en vormen van sociaal constructivisme); 3. een empirische test van dit filosofische standpunt door gevalsanalyses omtrent het ‘gebruik’ van het duurzaamheidconcept in concrete beleidsrelevante contexten; 4. een onderzoek van het belang van de verworven theoretische en empirische inzichten bij de vormgeving van een nieuwe benadering voor het Belgische energiebeleid; 5. een eerste inschatting van de mogelijke implementatie en haalbaarheid van dit nieuwe voorstel. Met dit onderzoek sluiten wij ons aan bij een transdisciplinaire stroming in het wetenschappelijk onderzoek die vanuit een filosofisch geïnspireerde houding het belang van politieke planning, juridisch-institutionele benaderingen en/of de rol van de overheid in complexe technologische dossiers probeert te herdenken. We wensen ook te benadrukken dat we de feitelijke aanvaarding van duurzame ontwikkeling op het internationale, nationale en regionale beleidsniveau als startpunt voor ons onderzoek beschouwen. We ontwikkelen dus geen fundamentele kritiek op het concept ‘duurzaamheid’ als dusdanig, maar leveren eerder een bijdrage aan het ontwikkelen van een beter begrip van de concrete ‘werking’ van dit concept in een specifiek beleidsdomein (namelijk het energiebeleid), met het oog op het bevorderen van de operationalisering.

2 Probleemstelling en onderzoeksvragen De dynamiek van het duurzaamheiddebat maakt het voor een wetenschappelijk onderzoeker onmogelijk terug te vallen op zoiets als een universeel aanvaarde ‘duurzaamheidtheorie’ of ‘-taxonomie’ die ons in staat zou stellen om de posities die in dit debat worden ingenomen netjes te ‘overschouwen’. In dit proefschrift argumenteren we daarom dat het wetenschappelijk onderzoek zich zou moeten toeleggen op het bestuderen van duurzame ontwikkeling als veranderingsproces, eerder dan zelf specifieke

Samenvatting iii

doelstellingen (op het gebied van cultuur, economie, ethiek, milieu, enz.) voorop te stellen. Dit houdt ook in dat ons onderzoek zich meer richt op het ‘procedurele’ aspect van het duurzaamheiddebat (‘hoe kan dit debat kwalitatief beter ingericht worden als een collectief leerproces?’), eerder dan dat we zelf een ‘inhoudelijk’ standpunt innemen (hoewel beide aspecten zeker niet scherp te scheiden zijn). We verleggen onze aandacht dan ook naar de formele en informele bestuursstructuren en –praktijken (governance) die in de praktijk vorm geven aan de interpretatie, visievorming en operationalisering van duurzaamheid. We willen evenwel niet de indruk wekken dat alle mogelijke problemen i.v.m. duurzame energie simpelweg op te lossen zijn door een verbeterde vorm van overheidsbestuur. Integendeel: relevante beslissingen worden op velerlei niveaus genomen, gaande van alledaagse gedragspatronen van consumenten en/of burgers over de bestuursraden van multinationals tot internationale politieke verdragen. Het is voor het beleid dan ook noodzakelijk om modulerend en reflexief in te spelen op veranderingsprocessen op deze verschillende niveaus. Deze benadering van de duurzaamheidproblematiek houdt tevens in dat we onze probleemformulering voldoende ruim moesten kiezen om veranderingsprocessen adequaat in kaart te kunnen brengen. We formuleerden onze centrale onderzoeksvraag als volgt: “Kan kernenergie bijdragen aan een veranderingsproces gericht op meer duurzaamheid – indien ja, onder welke voorwaarden; en hoe kan deze vraag best beantwoord worden?”. Deze centrale onderzoeksvraag omhelst drie deelvragen: 1. Hoe wordt het principe van duurzame ontwikkeling geïnterpreteerd, zowel in algemene zin als in de meer specifieke context van het energiebeleid?; 2. Welke wetenschappelijke methodes en theorieën worden naar voren geschoven om een beleid gericht op duurzame energie te ondersteunen en in welke mate komen deze methodes tegemoet aan alle dimensies van de duurzaamheidproblematiek – en in het bijzonder de institutionele aspecten daarvan?; Wat zijn de sterke en zwakke punten van deze methodes, en welke lessen kunnen eruit getrokken worden voor de ondersteuning van het beleid?; 3. Hoe kunnen antwoorden op de vorige twee vragen bijdragen aan de constructie van een ‘betere’ beleidsondersteuning op het gebied van duurzame energievoorziening?. Bij het beantwoorden van deze vragen gebruikten we het debat rond kernenergie telkens als concreet aanknopingspunt. Waar in de jaren vijftig en zestig nog een grote politieke, industriële, wetenschappelijke en brede maatschappelijke consensus bestond rond het nut van deze technologie, brokkelde deze steun gaandeweg af. Ook de beleidscontext veranderde grondig. Waar het elektrische energiebeleid vroeger vooral een interne aangelegenheid was tussen de overheid, vertegenwoordigers van de elektriciteitssector en georganiseerde belangengroepen (werkgeversorganisaties, grote industriële elektriciteitsverbruikers en vakbonden), werd dit overlegmodel in het voorbije decennium opengebroken door de liberalisering van de elektriciteits- en gasmarkten, zoals vastgelegd door Europese directieven die omgezet werden in nationale wetgeving. Deze liberalisering heeft verstrekkende gevolgen voor wat betreft de actiemogelijkheden van nationale

iv


overheden. Zo verliest de uitrustingsplanning – het centrale sturingsmechanisme voor de elektriciteitsproductie voor de periode van de liberalisering – haar bindend karakter en wordt eerst nog ‘indicatief’ opgesteld, later volledig afgeschaft voor wat betreft de productie van elektrische energie. Tevens wordt het Europese niveau in een vrijgemaakte markt het meest relevant voor de planning van productie-eenheden door de spelers in een vrije markt, terwijl het nationale niveau belangrijke bevoegdheden blijft behouden om de energievraag bij te sturen (alhoewel ook op dit laatste vlak Europese initiatieven te zien zijn). Nationale overheden blijven evenwel de mogelijkheid behouden om de inzet van bepaalde technologieën op hun grondgebied te verbieden. Zo vertaalde het gebrek aan politiek en maatschappelijk draagvlak voor kernenergie zich in 2003 in een parlementaire beslissing om op termijn (in de periode 2015-2025) de bestaande kerncentrales (4 in Doel, 3 in Tihange) te sluiten. Maar ondanks het feit dat de kernuitstap in 2003 in een wet verankerd werd, lijkt dit debat ons verre van definitief beslist. De eventuele post-Kyoto verplichtingen, het belang van lage en stabiele elektriciteitstarieven vooral voor de grote industriële bedrijven in België, de te verwachten stijging van de internationale marktprijzen van fossiele brandstoffen, en de gevaren van een te grote afhankelijkheid van de elektriciteitslevering op basis van één enkele brandstof (aardgas) bij sluiting van de kerncentrales creëren een klimaat dat het mogelijk maakt voor sommige politici, wetenschappers en belangengroepen om nu en dan met gerichte communicaties die wet op de kernuitstap in vraag te stellen. Tegenstanders blijven de onherroepelijke sluiting van de kerncentrales bij een veertigjarige levensduur eisen. Het lijkt ons dus onvermijdelijk dat de overheid vroeg of laat opnieuw een stelling zal moeten innemen. Mogelijk creëert het overheidsstreven naar meer duurzame ontwikkeling een kader waarin het (kern)energiedebat op een meer gestructureerde manier aan de orde kan komen. Dit proefschrift probeert een bijdrage aan dit nieuwe debat te leveren.

3 Methodologische uitgangspunten In het theoretische deel van dit proefschrift argumenteren we dat de benadering van duurzame ontwikkeling als een veranderingen/of leerproces best onderbouwd kan worden door een constructivistische wetenschapsopvatting (zie verder). Binnen zulke opvatting verwijst duurzaamheid niet naar een vaststaande – door wetenschappers, politici of ethici te bepalen – ‘natuurlijke gegevenheid’, maar naar een voortdurende collectieve confrontatie waarbij diverse duurzaamheidconstructies met elkaar in botsing komen. Dit proces brengen we gedeeltelijk in kaart, maar tevens is onze analyse ook meer dan louter beschrijvend. We formuleren voorstellen om signalen afkomstig van deze collectieve confrontatie beter te articuleren en te interpreteren ter ondersteuning van een beleidsproject dat het bedenken van verschillende visies op duurzame energievoorziening toelaat en uitprobeert. Door deze constructivistische wetenschapsopvatting situeren we ons in het kritisch-interpretatieve bereik van de humane wetenschappen. De kritische component van ons onderzoek wordt gevormd door onze aandacht voor de manier waarop bestaande interpretaties en praktijken op het gebied van duurzame energievoorziening vorm krijgen door dominante waarden en normen, en de uitsluitingsmechanismen die hiermee intrinsiek

Samenvatting v

verbonden zijn. Deze component wordt vooral benadrukt bij het ontwikkelen van het theoretische luik van dit proefschrift. De interpretatieve component berust op onze aandacht voor de manier waarop betekenissen worden toegekend aan bepaalde gebeurtenissen of concepten, en wordt vooral benadrukt in de gevalsstudies van dit proefschrift.

4 Uitwerking van de onderzoeksvragen in een drieluik De drie onderzoeksvragen worden beantwoord in drie verschillende onderzoekslijnen die we iteratief met elkaar hebben verweven. In een eerste onderzoekslijn leveren we een bijdrage aan het filosofische debat rond het conceptualiseren van de relatie tussen (technologische) ontwikkeling (die meestal verbonden wordt met dieper gelegen veronderstellingen en waardeoordelen aangaande de toekomst van het moderniseringsproject) en de milieuproblematiek. Tegen de achtergrond van dit ‘gesprek’ tussen verschillende filosofische perspectieven werken we een eigen onderzoekskader uit waarin we duurzame ontwikkeling als leerproces situeren ten opzichte van vier mogelijke bestuursmodellen (theories of governance). De tweede onderzoekslijn bestaat uit drie gevalsstudies, die elk een beeld schetsen over hoe kernenergie gepositioneerd wordt ten opzichte van andere energieopties in het duurzaamheiddebat. Binnen deze onderzoekslijn verzamelen we gegevens aangaande drie relevante analyse-eenheden of niveaus, namelijk het ontwikkelen van beleidsgerichte wetenschap, het ontwikkelen van beleidsmaatregelen en het algemene maatschappelijke debat aangaande kernenergie. Op het niveau van beleidsgerichte wetenschapsbeoefening analyseren we een grootschalig Europees onderzoeksproject (ExternE) dat een inschatting probeert te geven van de totale maatschappelijke kost (de som van de private en externe kost) van verschillende energiebronnen. Op het niveau van het energiebeleid reconstrueren we vervolgens de beleidsontwikkelingcyclus in de context van de Belgische parlementaire beslissing (2003) om op termijn uit de kernenergie te stappen. Tenslotte analyseren we op het niveau van het maatschappelijk debat rond kernenergie in België de posities die door de voornaamste belanghebbenden worden ingenomen. Door deze gevalsstudies proberen we telkens de beleidspraktijk vanuit een andere hoek te belichten. Tenslotte verbinden we in een derde onderzoekslijn onze theoretische en praktische bevindingen in een praktijkgericht voorstel voor een nieuw bestuursmodel op het gebied van duurzame energievoorziening.

5 Samenvatting, conclusies en aanbevelingen Antwoorden op de eerste twee onderzoeksvragen groeperen we onder de sectie ‘samenvatting en conclusies’, terwijl de antwoorden op de derde onderzoeksvraag geformuleerd worden als concrete aanbevelingen voor het beleid.

vi


5.1

5.1.1

Samenvatting en conclusies

Constructivisme als (meta-)theoretisch uitgangspunt

Zoals boven vermeld wordt de vraag naar de rol van kernenergie in een toekomstig energiebeleid door ons gesitueerd binnen de brede dynamiek van het duurzaamheiddebat. Dit duurzaamheiddebat vatten wij in dit proefschrift op als het geheel aan inspanningen en pogingen die ondernomen worden om duurzaamheidproblemen te omschrijven en oplossingen voor deze problemen uit te denken en een kader te scheppen om deze uit te voeren. Kenmerkend voor dit debat is de circulatie van een grote hoeveelheid aan theoretische inzichten, opinies en discoursen met betrekking tot het eigenlijke probleem (‘waarom is onze energievoorziening onduurzaam?’) en de wijze waarop dit probleem moet worden opgelost. Ook zijn de belangentegenstellingen tussen de groepen die bij het debat betrokken vaak groot en niet altijd volledig transparant. Dit gegeven stelt een onderzoeker voor een aanzienlijke uitdaging: kunnen we zomaar aannemen dat uit die veelheid aan perspectieven het ene ‘juiste’ kan gekozen worden waarbij het (energetische) duurzaamheidprobleem in zijn ‘ware dimensies’ wordt onthuld? Met andere woorden: vragen van (meta-)theoretische aard dringen zich in het geval van de studie van de duurzaamheidproblematiek op een bijzonder pertinente manier op. In eerste instantie verhelderen we het duurzaamheiddebat door een onderscheid te maken tussen drie niveaus waarop dit debat gevoerd wordt: het niveau van een (diffuus, algemeen) ‘manifest beeld’ van socio-materiële levensvormen, het niveau van (inherent politieke) visievorming, en het niveau van een (technische) uitwerking van duurzaamheid als een beleidsdoel. Om goed te begrijpen hoe een bepaald perspectief of theorie duurzaamheidproblemen opvat is het bijgevolg nodig dit perspectief op deze drie niveaus te bevragen. In dit proefschrift onderwerpen we een aantal courante (meta-)theoretische perspectieven aan dergelijke kritische bevraging. De risicomaatschappij-theorie (vooral uitgewerkt door de Duitse filosoof en socioloog Ulrich Beck) thematiseert de huidige overheersende ervaring van onzekerheid (zowel op het gebied van industriële en/of technologische risico’s als op het gebied van het in verval raken van traditionele rolpatronen) als een effect dat inherent verbonden is met de moderne mogelijkheid om de natuurlijke en maatschappelijke realiteit steeds beter te manipuleren. Met andere woorden, de hedendaagse ervaring van een soort ‘existentiële onzekerheid’ is geen toevallige, eenvoudig te controleren neveneffect van een in wezen onproblematisch moderniseringsproces, maar daagt de basisveronderstellingen en instellingen van de moderne maatschappij zelf uit. De kern in het risicomaatschappij-discours is dat deze patstelling enkel overwonnen kan worden door een omslag in het moderniseringsproces: waar de ‘eerste’ of ‘klassieke’ moderniseringsgolf de omslag van een feodale naar een industriële samenleving mogelijk heeft gemaakt, moet deze nu opgevolgd worden door een ‘tweede’ of ‘reflexieve’ modernisering van de industriële samenleving. In dit proefschrift verdedigen we de stelling dat deze theorie al te gemakkelijk uitgaat van een essentialistisch-structurele verklaring op macroniveau om een in wezen politieke visie te

Samenvatting vii

ondersteunen, waarbij onvoldoende aandacht besteed wordt aan empirische onderbouwing aan de hand van concrete politieke contexten of controversen rond technologische opties. Een tweede mogelijk (meta-)theoretisch perspectief, het sociaal constructivisme, besteedt wel ruimschoots aandacht aan concrete praktijkvoorbeelden, maar valt ten prooi aan een andere vorm van essentialisme. Het sociaal constructivisme gaat namelijk uit van een verklaringsmodel louter in termen van de sociale belangen van de actoren die bij de ‘constructie’ van een bepaald wetenschappelijk feit, technologie of besluitvormingsinitiatief betrokken zijn, maar is daardoor niet in staat een verklaring te geven hoe keuzes ooit gelegitimeerd kunnen worden. In tegenstelling tot beide bovenvernoemde perspectieven pleiten we in dit proefschrift voor een constructivistische (zonder het ‘sociale’) wetenschapsopvatting, ontdaan van essentialistische en grondingaspiraties (dit perspectief leunt dicht aan bij de werken van de Franse antropoloog en filosoof Bruno Latour). Dergelijke opvatting laat ons toe het duurzaamheiddebat op te vatten als een collectieve confrontatie van (al dan niet wetenschappelijke) perspectieven die elk als ‘bemiddelaar’ optreden voor de hybride entiteiten die door Latour ‘actanten’ worden genoemd (een notie waarmee hij het traditionele onderscheid tussen sociale ‘actoren’ en ‘objecten’ opheft). We vullen dit perspectief vervolgens aan met inzichten ontleend aan de justificatietheorie van de Franse socioloog Luc Boltanski en de econoom Laurent Thévenot (le modèle des cités de justification). Boltanski en Thévenot gaan ervan uit dat mensen in hun zoektocht naar een rechtvaardige oplossing in betwiste situaties gebruik maken van bepaalde ‘grammaticale’ regels, die echter verschillend worden uitgewerkt al naargelang het specifieke justificatieregime dat geacht wordt van toepassing te zijn. Deze theorie laat ons toe om de ‘collectieve constructie’ van een duurzame energietoekomst aan een dergelijke grammaticale lezing te onderwerpen. Uiteindelijk kunnen we onze keuze voor deze theorie niet in absolute termen verdedigen (dit zou uiteraard in tegenspraak zijn met de kernelementen van een constructivistische houding), maar moet het ‘nut’ van deze theorievorming blijken uit criteria zoals de bruikbaarheid ervan bij het ordenen van empirische gegevens (de gevalsstudies), de morele aanvaardbaarheid bij het bevorderen van collectieve leerprocessen en de praktische haalbaarheid ter ondersteuning van het nieuwe bestuursmodel dat wij in dit proefschrift voorstellen. 5.1.2

Vier bestuursmodellen gewikt en gewogen

Deze constructivistische blik laat ons toe om een aantal (traditionele) benaderingen van beleid in complexe technologische aangelegenheden aan een nieuw onderzoek te onderwerpen. Hierbij worden deze benaderingen of modellen niet zozeer beschouwd als methodes die ons zouden toelaten de geschikte ‘middelen’ te kiezen voor een gegeven problematiek, maar eerder als een samenwerkingsverband (cooperative scheme) – een complex geheel van praktijken, attitudes, interactieregels enz. – dat bepaalt hoe een bepaalde situatie tot een beleidsprobleem wordt omgevormd (geconstrueerd). Hierbij worden steeds bepaalde veronderstellingen gemaakt, bijvoorbeeld aangaande relevante kennisperspectieven, en/of ‘toelaatbare’ argumenten en de rolverdeling bij het bedenken en uitvoeren van oplossingsstrategieën. We onderwerpen vier (in ideaaltypische zin opgevatte) bestuursmodellen aan een dergelijk kritisch onderzoek.

viii


Het expertmodel (expert-based governance) gaat uit van de mogelijkheid om bij het oplossen van beleidsvraagstukken de politieke autoriteit en verantwoordelijkheid scherp te scheiden van de wetenschappelijke, die berusten op competentie in bepaalde (voor het probleem relevant geachte) wetenschappelijke disciplines. Dit expertmodel benadrukt de overdracht van expertkennis naar beleidsmakers en gaat veelal uit van het belang van centraal gestuurde maatregelen ter bevordering van duurzaamheid. Dergelijke oplossingsstrategieën bewijzen hun nut doorgaans in beleidsvraagstukken die gekenmerkt worden door een grote eensgezindheid aangaande de ‘feiten’ en ‘waarden’ die op het beleidsprobleem in kwestie van toepassing zijn. In het vaak controversiële debat rond de duurzaamheid van energieopties botst dit model – hoewel het zeker nog de dominante benadering vormt – op inherente grenzen. Het risico bestaat dan namelijk dat waardengeladen keuzes stilzwijgend gedelegeerd worden naar expertgroepen, of dat de op de wetenschappelijke rationaliteit geënte oplossingsstrategieën in de praktijk onwerkbaar blijken. Een andere veelgebruikte beleidsstrategie is het aggregatiemodel (governance by aggregation). In dit model wordt een directe confrontatie van sterk ethisch geladen perspectieven vermeden door beleidsproblemen op te vatten naar het model van een ‘onderhandeling’ tussen verschillende ‘belangen’, waarbij verondersteld wordt dat deze belangen op één of andere manier vergelijkbaar kunnen gemaakt worden. Het model toont echter zijn begrenzingen bij toepassing op problemen waarbij een beroep gedaan wordt op diepgewortelde ethische overtuigingen (zoals het geval is in het debat rond kernenergie) en door wetenschappelijke onzekerheden die de eenduidige identificatie van ‘belangen’ bemoeilijken. In tegenstelling tot het aggregatiemodel worden waardengeladen debatten in het pacificatiemodel (governance by pacification) niet vermeden. Dit model gaat uit van de verdediging van bepaalde maatschappelijke waarden door een aantal georganiseerde en door de staat erkende groepen, die via een georganiseerde overlegprocedure tot een compromis proberen te komen aangaande de gewenste beleidsvoering. Het georganiseerde overleg tussen overheid en vertegenwoordigers van de elektriciteitssector, werkgeversorganisaties, grote industriële energiegebruikers en vakbonden in overlegorganen zoals het ‘Controle Comité voor Elektriciteit en Gas’ (CCEG) en het ‘Nationaal Comité voor Energie’ (NCE) was – tot voor de opening van de energiemarkten – een typevoorbeeld van het pacificatiemodel. Maar ook in de context van het huidige duurzaamheiddebat neemt de overheid opnieuw een toevlucht tot dit pacificatiemodel, bijvoorbeeld door de oprichting van een ‘Federale Raad voor Duurzame Ontwikkeling’ (FRDO) met vertegenwoordiging van de belangrijkste maatschappelijke actoren (de ‘traditionele’ actoren uitgebreid met de milieubeweging, derdewereldorganisaties en enkele andere groepen). Het pacificatiemodel – dat in het vroegere energiebeleid zeker een cruciale rol heeft gespeeld – kan echter niet zomaar ‘overgeplant’ worden naar het nieuwe debat dat duurzaamheid vooropstelt. Problemen stellen zich bijvoorbeeld bij de georganiseerde vertegenwoordiging van entiteiten zoals ‘het milieu’ of ‘toekomstige generaties’. Bovendien leert het historische voorbeeld van het energiebeleid ons dat de

Samenvatting ix

‘pacificatie’ van waardenconflicten vaak ten koste ging van de transparantie van de gebruikte gegevens en wetenschappelijke modellen. Het deliberatiemodel (deliberative governance) (dat wij in dit proefschrift voornamelijk hebben uitgewerkt aan de hand van Habermas’ theorie van het communicatieve handelen) legt vooral de nadruk op het belang van een ‘machtsvrije’ en ‘inclusieve’ dialoog aangaande zowel de in het geding zijnde ‘feiten’ als ‘waarden’, en lijkt dus uitermate geschikt om de duurzaamheidproblematiek het hoofd te bieden. Het model heeft echter onvoldoende aandacht voor de politieke dimensie van het duurzaamheiddebat, die gekenmerkt wordt door een onuitwisbare spanning en conflictueuze interactie. Tevens is het ook hier onduidelijk hoe op basis van het deliberatiemodel een ‘stem’ gegeven kan worden aan ‘het milieu’ of ‘de toekomstige generaties’. Als besluit kunnen we stellen dat elk model zekere verdiensten heeft, maar toch ook telkens belangrijke tekortkomingen vertoont in de context van duurzaamheid. Het bestuursmodel dat wij in dit proefschrift uitwerken ontleent bijgevolg enkele sterke elementen aan elk van deze modellen, maar wijkt er toch ook weer in belangrijke mate van af. 5.1.3

Het marktmodel als ultieme scheidsrechter?

Voor onze eerste gevalsstudie kozen we een grootschalig Europees onderzoek naar de externe kosten van verschillende energiebronnen (het ExternE project). Dit onderzoek berust in wezen op het doordenken van het aggregatiemodel (zie boven) tot in haar uiterste consequenties. De basisfilosofie achter het ExternE project is namelijk een monetaire evaluatie toe te kennen aan alle mogelijke impacten van de verschillende opties voor elektriciteitsproductie. De idee is op zich aanlokkelijk: het berekenen van de externe kosten van verschillende energieopties zou het beleid toelaten om consistente en rationele keuzes te maken, die bovendien gemakkelijk te communiceren zouden zijn naar een breder publiek. Bovendien worden de sturingsmogelijkheden van overheden door de vrijmaking van de energiemarkten ook systematisch verlegd naar het gebruik van marktinstrumenten (taksen, subsidies, handel in emissierechten, enz.), zodat ook hier het ExternE project tegemoet komt aan een duidelijke beleidsbehoefte. Critici formuleren echter ethische bezwaren tegen het volledig kaderen van energiekeuzes in een marktmodel, en verwerpen dan ook de theorie van de externe kosten als een betrouwbare basis voor het beleid. In deze gevalsstudie leggen we de verschillende ‘bemiddelingen’ (veronderstellingen, modelberekeningen, meetinstrumenten, enz.) bloot die aangewend worden om de berekening van externe kosten mogelijk te maken. We argumenteren dat het verwerpen van het hele ExternE project op ethische gronden voorbijgaat aan het belang van een contextuele evaluatie van het mogelijk nut van een dergelijke benadering. Zo maken wij een onderscheid tussen een ontdekkingscontext, een rechtvaardigingscontext en een toepassingscontext. In de ontdekkingscontext (context of discovery) bewijst het onderzoek naar de externe kosten van energieopties haar nut door op een systematische manier op zoek te gaan naar ‘entiteiten’ die totnogtoe aan de economische berekening ontsnapt zijn. Zodoende draagt

x


dit onderzoek bij aan de vereiste aan inclusiviteit die we in ons voorstel voor een nieuw bestuursmodel vooropstellen. In de rechtvaardigingscontext (context of justification) is de bovenvernoemde ethische kritiek terecht, tenminste als de externe kosten als een soort ‘pensée unique’ worden voorgesteld. Als de externe kosten ‘verhaallijn’ (een notie die we introduceren om de nadruk te leggen op alle noodzakelijke veronderstellingen en simplificaties die nodig zijn om alle ‘actanten’ in een marktlogica in te brengen) echter geplaatst wordt naast andere mogelijke ‘verhaallijnen’ (die bijvoorbeeld ontwikkeld kunnen worden op basis van het model van Boltanski en Thévenot) is het – mits de nodige procedurele garanties – mogelijk om tot een meer constructieve interactie te komen aangaande de gewenste beleidsmaatregelen. Dit kan dan gebeuren door in de toepassingscontext (context of application) compromissen te zoeken tussen de verschillende ‘zuivere’ verhaallijnen, door bijvoorbeeld het gebruik van externe kost berekeningen te limiteren tot bepaalde impacten of beleidsdomeinen. 5.1.4

Een blik op de beleidvormingspraktijk

Een tweede gevalsstudie betreft de concrete beleidvormingspraktijk in het geval van de Belgische parlementaire beslissing om op termijn (in de periode 2015-2025) uit de kernenergie te stappen. Hier tonen we aan dat de officiële rechtvaardiging van deze beslissing zich baseerde op het justificatieschema dat we typisch aantreffen in het expertmodel (zie boven). Dit was bijvoorbeeld duidelijk in de tijdens de parlementaire discussies zelfverklaarde onderbouwing van de beslissing door de expertopinie (met verwijzing naar het AMPERE rapport), een zeer geringe bereidwilligheid om een debat over ethische waarden aan te gaan, en de bureaucratische behandeling van de verdere opvolging van de wet op de kernuitstap. Een gedetailleerde analyse onthult echter dat deze ‘technische’ behandeling van het beleidsprobleem in officiële beleidsdocumenten ten koste ging van enkele ambiguïteiten en weglatingen. Deze bevinding werd bevestigd door onze derde gevalsstudie, die het maatschappelijk debat rond (kern)energie betreft. Op basis van interviews met belanghebbenden in dit debat (geselecteerd op basis van de lijst van de in de FRDO vertegenwoordigde organisaties) reconstrueerden we drie ideaaltypische perspectieven op de rol van kernenergie in een duurzame energietoekomst. Het managementperspectief (management perspective) ziet verdere economische groei en technologische ontwikkeling als de motor achter elke vorm van duurzame ontwikkeling. Elke (overheids)actie die deze groei in gevaar brengt moet bijgevolg vermeden worden. Het beleid moet volgens dit perspectief vooral een stabiel kader bieden voor ondernemingen, die hierdoor hun verantwoordelijkheid naar de maatschappij (via vormen van ‘duurzaam ondernemen’) kunnen opnemen. Volgens dit perspectief is er geen reden waarom elektriciteitsbedrijven die nucleaire reactoren uitbaten geen deel van de oplossing zouden kunnen vormen. Het controleperspectief (controllist perspective) besteedt meer aandacht aan de institutionele inbedding van technologie in de maatschappij. Binnen dit perspectief wordt

Samenvatting xi

de vrees geuit dat kernenergie ‘onvermijdbaar’ zal worden in de toekomst, tenminste als de overheid vasthoudt aan de dominante ‘technische’ probleemomkadering – namelijk het tegelijk respecteren van eventuele post-Kyotonormen en economische groeidoelstellingen. Dit perspectief toont zich voorstander van een meer ‘democratische’ vorm van probleemomkadering: aanvaarding of verwerping van kernenergie zou dan eerder gebaseerd moeten worden op overleg met de voornaamste betrokken partijen, onder garanties van voldoende transparantie. Vooral aan die transparantie schort er volgens dit perspectief veel in de huidige besluitvorming: de verdeling van kosten en baten van kernenergie in het verleden, de kosten voor de ontmanteling van kerncentrales en de berging van het radioactief afval, de echte kosten van het ‘business-as-usual’ scenario, enz. worden gezien als ‘grote onbekenden’ in het energiedebat. Het hervormingsperspectief (reformist perspective) ziet de evolutie van de Belgische elektriciteitssector als een vastgeroest proces waarbij de ontwikkeling van wetenschappelijke kennis, technologische innovatie (of het gebrek daaraan bij hernieuwbare energiebronnen) en winst voor de dominante marktspeler elkaar versterken – een proces dat in dit perspectief enkel verklaard kan worden door de uitoefening van politieke en economische macht. Dit alles zorgt er volgens dit perspectief voor dat perfect valide technologische opties (zoals hernieuwbare energiebronnen en warmtekrachtkoppeling) onvoldoende ontwikkeld worden, rationeel energiegebruik ontmoedigd wordt, de ‘echte’ kosten van de verschillende energieopties verborgen blijven, en tenslotte dat burgers geen stem krijgen in het debat. Kernenergie vormt binnen dit perspectief de ware belichaming van wat er mis gaat in de Belgische energiewereld. Uit de analyse van het maatschappelijk debat rond de toekomstige rol van kernenergie blijkt bovendien duidelijk dat er nauwelijks transversale verbanden tussen deze verschillende perspectieven aanwezig waren – met andere woorden elk van deze perspectieven baseert zich op een andere probleemomkaderingen, gebruikt andere gegevens, methodes, wetenschappelijke rapporten, enzovoort. Dergelijke omstandigheden stimuleren enkel een vorm van antagonistisch leren – d.i. een vorm van strategisch leren gericht op het ‘verschalken’ van de tegenstanders.

5.2

Aanbevelingen

Wie de analyse tot dusver heeft gevolgd kan niet anders dan tot een eerder pessimistische conclusie komen. Het debat rond de plaats van kernenergie in een toekomstgericht energiebeleid en meer algemeen in een globaal concept van duurzame ontwikkeling is tot vandaag opnieuw een duidelijk voorbeeld van polarisatie tussen voor- en tegenstanders, waarbij vele aspecten van de internationaal aanvaarde visie op duurzame ontwikkeling onderbelicht blijven. Nochtans is de kwaliteit van dit debat bepalend voor de mogelijkheid van een samenleving om signalen te interpreteren in functie van toekomstgerichte projecten zoals duurzame ontwikkeling. In navolging van auteurs zoals Bruno Latour proberen wij vanuit een constructivistisch gedachtegoed de kwaliteit van de interacties te verhogen. Kwaliteit staat hierbij niet meer in het teken van een loutere overdracht van expertkennis of het implementeren van een centraal bepaald sturend beleid, maar wel van het bevorderen

xii


van allerlei collectieve leerprocessen die de zelfreflexiviteit van het collectieve debat proberen te verhogen. In de laatste hoofdstukken van dit proefschrift formuleren we aanbevelingen in verband met de vormgeving van dergelijk collectief leerproces. 5.2.1

Een ‘agentschap voor een “duurzame” energievoorziening’

Aanbeveling nr. 1: Inclusieve leerprocessen gericht op duurzame ontwikkeling moeten institutioneel verankerd worden. De oprichting van een ‘agentschap voor een “duurzame” energievoorziening’ is in dit opzicht absoluut noodzakelijk Dit agentschap moet een adviserende rol opnemen ten dienste van de formele democratische instellingen, terwijl het tegelijk toch een zekere afstand ten opzichte van deze instellingen moet in acht nemen.

Een kader voor het debat over de toekomst van onze energievoorziening zou op zeer uiteenlopende wijzen vorm kunnen krijgen. De ruimte van denkbare ‘constructies’ wordt o.i. omspannen door twee uitersten. Aan de ene kant zou men een zuiver conceptueel kader kunnen vooropstellen; een door alle betrokken partijen gedeelde opvatting over hoe dit soort kwesties best aangepakt worden. Aan de andere kant zou men een soort ‘commissie’ of ‘raad’ kunnen oprichten die als een soort ‘nationaal geweten’ zou fungeren en beslissingen neemt over hoe knopen moeten worden doorgehakt. Op basis van onze analyse zijn we eerder geneigd een tussenpositie in te nemen: het creëren van een organisatie die, op basis van gedegen informatie en onderzoek, probeert articulatie- en discussieprocessen te bevorderen, met adviserende functie naar parlement en overheid (dus zonder beslissingen te nemen in de plaats van de politiek verantwoordelijken). Aldus moet dit ‘agentschap voor een “duurzame” energievoorziening’ (DEV agentschap) functioneren als een ‘knoop’ in een netwerk dat verschillende ‘werelden’ (wetenschappelijke instituten, georganiseerde maatschappelijke groepen, de politieke wereld) en ‘niveaus’ (gaande van internationale besluitvorming, over nationale en regionale bevoegdheden tot alledaagse socio-materiële praktijken) probeert te kanaliseren. Dit kanaliseren betreft in essentie twee vragen (Latour): ‘met hoeveel (actanten) zijn we?’ (m.a.w. welke diversiteit aan belangen, (wetenschappelijke) perspectieven, probleem-kadreringen enz. laten we toe in het debat?) en ‘kunnen we samenleven?’ (m.a.w. is er een vorm van voorlopige overeenstemming te vinden?). Het spreekt vanzelf dat aan de voorbereiding en organisatie van een deliberatie aangaande deze vragen de grootst mogelijke zorg besteed moet worden. De volgende paragrafen gaan in op enkele essentiële elementen van kwaliteitsborging.

Samenvatting xiii

5.2.2

Open en flexibele relaties met de ‘buitenwereld’

Aanbeveling nr. 2: In overeenstemming met een constructivistische grondhouding (die immers elke essentialistische- of grondingaspiratie afwijst) is het belangrijk open en flexibele relaties te onderhouden met concrete praktijken die plaatsvinden buiten de muren van het DEV agentschap. In het bijzonder bevelen we aan om het voorzorgsprincipe formeel in te schrijven in het mandaat van dit agentschap als een noodzakelijke (maar niet voldoende) randvoorwaarde bij de uitvoering van haar opdrachten.

De ruimtelijke metafoor die we in deze aanbeveling vooropstellen moet niet al te letterlijk opgevat worden: bedoeling is het doorbreken van de manier waarop een vanzelfsprekend taalgebruik, dominante actoren en hun symbolische referenties, dominante wetenschappelijke perspectieven en vaste spelregels van overleg doorwerken in de argumenten en in de uitkomst van het debat. In dit proefschrift argumenteren we dat een verplichte toepassing van het voorzorgsprincipe (bvb. door dit voorzorgsprincipe expliciet in te schrijven in de oprichtingsstatuten van het DEV agentschap) een noodzakelijke maar niet voldoende voorwaarde vormt voor open en flexibele relaties met de ‘buitenwereld’. Toepassing van het voorzorgsprincipe heeft immers het voordeel dat aangeleund kan worden bij de zich op internationaal vlak ontwikkelende interpretatie en beleids- en juridische toepassingen van dit principe, wat een consistente en niet-arbitraire toepassing van beslissingsregels principieel mogelijk maakt. Niettemin moet de toepassing van het voorzorgsprincipe aangevuld worden met andere dimensies van de duurzaamheidproblematiek zoals de (internationale) verdelende rechtvaardigheid van risico’s en voordelen van technologische ontwikkeling, en de (internationale) institutionele dimensie. In deze zin geven we ook enkele concrete institutionele aanbevelingen voor het concreet tot stand brengen van ‘openheid naar de buitenwereld’, bijvoorbeeld door publieke diensten uit te rusten met een ‘early warning’ functie, de onafhankelijkheid van regulerende organismen te garanderen, en/of onafhankelijk academisch onderzoek te ondersteunen. 5.2.3

Ondersteuning van deliberaties door criteria voor een “duurzame” energievoorziening en multi-criteria technieken

Aanbeveling nr. 3: De taakuitvoering van het DEV agentschap moet ondersteund worden door specifieke beleidswetenschappelijke ‘objecten’ (bvb. waardenbomen, criteria en indicatoren voor een duurzame energievoorziening) en ‘technieken’ (bvb. multi-criteria mapping).

In dit proefschrift argumenteren we dat het absoluut noodzakelijk is de deliberaties aangaande een duurzame energietoekomst te ondersteunen met een duidelijke lijst van

xiv


duurzaamheidcriteria en –indicatoren, die een kader vormen voor het debat. Hierbij is het van belang dat deze lijst legitiem is voor de verschillende soorten actoren die een rol spelen (belanghebbenden, wetenschappers, beleidsmakers, enzovoort). Zo is het bijvoorbeeld duidelijk dat experts beter de grenzen van hun competentie kunnen respecteren naarmate beleidsvragen beter afgebakend worden. De kwaliteit van een expertadvies zal er dan ook op vooruitgaan naargelang pogingen ondernomen worden om ongestructureerde beleidsproblemen met duidelijk gedefinieerde en gemeenschappelijk onderhandelde criteria om te zetten in meer gestructureerde deelproblemen (die met behulp van wetenschappelijk gevalideerde methodes aangevat kunnen worden). Bij het opstellen van deze lijst van criteria is het van belang om voldoende oog te hebben voor de concrete dynamiek van het duurzaamheiddebat. Al te vaak worden duurzaamheidcriteria zonder verdere aanpassing overgenomen van internationale (beleids)initiatieven ter zake. Zo loopt men het risico om belangrijke aspecten van dit debat, zoals bijvoorbeeld de institutionele dimensie van de organisatie van het energiesysteem (centraal versus decentraal), weg te laten. In tegenstelling tot deze gangbare praktijk construeren we in dit proefschrift een lijst van criteria voor duurzame energie in nauwe interactie met betrokkenen in het debat. We passen hierbij ook een multi-criteria methode toe (multi-criteria mapping) die ons toelaat om op participatieve manier ‘heen-en-terug’ te redeneren tussen toekomstgerichte energiescenario’s en de mogelijke criteria om de impacts van deze scenario’s te evalueren. Dergelijke technieken kunnen uiteraard nooit als vervanging van publieke besluitvorming gebruikt worden, maar enkel als ondersteuning. 5.2.4

Hervorming van toekomstverkenningspraktijken

Aanbeveling nr. 4: Strategische prioriteiten voor een “duurzame” energievoorziening moeten gebaseerd zijn op een ‘brede’ kosten-batenanalyse. Deze analyse moet ondersteund worden door een wetenschappelijk verantwoorde vorm van toekomstverkenning, op voorwaarde dat de nu gangbare toekomstverkenningspraktijken aangepast worden om normatief geïnspireerde analyses mogelijk te maken.

Toekomstverkenning is een onmisbaar element bij het beantwoorden van de tweede vraag (‘kunnen we samenleven?’). In dit proefschrift argumenteren we dat een antwoord op deze vraag best gezocht kan worden door te vertrekken van een aantal normatieve toekomstgerichte energiescenario’s, die in een breed gedefinieerde kosten-batenanalyse (contrasterend met de ‘nauwere’ vorm van kosten-batenanalyse die we bespraken aan de hand van het ExternE project) deliberatief tegen elkaar uitgespeeld kunnen worden. We tonen aan hoe opnieuw het model van Boltanski en Thévenot nuttig aangewend kan worden om een invulling te geven aan een wensbare energietoekomst vanuit de verschillende justificatieregimes geïdentificeerd door deze auteurs. Bovendien biedt dit model een gemeenschappelijk referentiekader voor de discussiepartners en laat het toe mogelijke compromissen of onoverbrugbare verschillen beter in kaart te brengen. Dergelijke praktijk van scenariobouw verschilt van de gangbare omdat hier veel sterker de

Samenvatting xv

nadruk gelegd wordt op de justificatie van energiekeuzes eerder dan op extrapoleren van trends of econometrisch modelleren van de toekomstige energievoorziening (zonder nochtans het belang van deze vormen van toekomstverkenning bij de keuze van concrete beleidsmaatregelen te willen minimaliseren). 5.2.5

Het belang van vertrouwen

Aanbeveling nr. 5: De overlegprocedures van het DEV agentschap moeten zodanig opgesteld worden dat vertrouwensvolle interacties gestimuleerd worden. Het is daarom van belang de interacties die leiden tot de identificatie, formulering, diagnose, en rangschikking van beleidsproblemen onderworpen worden aan een analyse vanuit dit opzicht.

Voor het welslagen van de discussies die binnen het ‘agentschap voor duurzame energie’ plaatsvinden is een zekere mate van vertrouwensvolle interactie tussen de deelnemers van groot belang. Er bestaat immers geen fundamentele grens aan het ‘rationeel’ in twijfel trekken van werkelijkheidsdefinities. Dit betekent dat de deelnemers aan deze discussies tot op zekere hoogte een gemeenschappelijk gedeelde identiteit moeten kunnen opbouwen. Het benadrukken van vertrouwen betekent niet dat wij een visie van politiek vooropstellen die volledig op vertrouwen zou gebaseerd zijn. Politiek en bestuur zijn immers onlosmakelijk verbonden met processen van machtsverwerving en/of bestendigen van machtsongelijkheid, en (verwachtingen omtrent) deze interacties in de ‘echte wereld’ zijn onmogelijk buiten de muren van het energieagentschap te houden. Om vertrouwen op te bouwen is het daarom van belang deze machtsrelaties enigszins op te schorten. In de aanbevelingen die we uit het onderzoek afleiden formuleren we enkele noodzakelijke voorwaarden om dit vertrouwen tot stand te brengen. Zo moeten er sterke procedurele richtlijnen in acht genomen worden in verband met de productie van betrouwbare gegevens en het gebruik van theorieën waarop analyses gestoeld zijn. Het is van zeer groot belang dat de werking van het DEV agentschap ondersteund wordt door een onafhankelijk en voldoende toegerust wetenschappelijk secretariaat. Aan de deelnemers aan het debat (maatschappelijke groepen) moet de mogelijkheid geboden worden om hun standpunten zo goed mogelijk te onderbouwen. Dit betekent bijvoorbeeld dat, waar nodig, financiële middelen moeten kunnen vrijgemaakt worden om externe expertise in te roepen. Daarnaast is het ook van belang om zo transparant mogelijk te zijn omtrent de rolverdeling in de collectieve deliberaties. Naast de georganiseerde belangengroepen (die ‘wegen’ op de besluitvorming en uitvoering van beleidsmaatregelen) identificeren we in dit proefschrift nog een aantal rollen die ingevuld moeten worden om kwalitatief hoogstaande interacties tot stand te brengen: probleemgerichte expertise, ethici, procesbewakers (administrators), procesbegeleiders (facilitators) en beleidsmakers. Elk van deze rollen zou idealiter een stem moeten krijgen in elke fase van de onderhandelingen (probleemafbakening en –definitie, definitie van strategieën en identificeren van kansrijke beleidsmaatregelen).

xvi


Tenslotte is het ook van belang om lessen te trekken uit de geschiedenis. Maatschappelijke actoren baseren hun inschatting van de evolutie in controversiële politieke dossiers op basis van de ervaring die ze hebben met vroegere claims en beloftes in dat dossier. Het is dan ook belangrijk deze ervaringen (die uiteraard vaak gepaard gaan met ‘emotionele’ interacties zoals wederzijdse verwijten, beschuldigingen, enz.) bespreekbaar te maken en er lessen uit te trekken. In verband met het kernenergiedebat denken we bijvoorbeeld aan de financiering van de nucleaire sector in het verleden (waarbij de sterke perceptie bestaat van een mechanisme van ‘privatisering van de winsten en collectivisering van de kosten’), het gebrek aan kadering van kernenergie in een globaal energiebeleid dat ook energiebesparing en hernieuwbare energiebronnen in overweging neemt, het ontstaan van ‘nucleaire passiva’, enzovoort. 5.2.6

Tijdelijke afsluiting van debatten

Aanbeveling nr. 6: Politiek bekrachtigde concrete doelstellingen zijn nodig om vooruitgang inzake duurzame ontwikkeling te meten. Deze doelstellingen moeten echter zodanig geformuleerd worden dat er nog ruimte is voor verdere articulatie en leerprocessen. Beleidsmaatregelen gericht op de verwezenlijking van deze doelstellingen moeten onderzocht worden op flexibiliteit, diversiteit en omkeerbaarheid .

De onderhandelingen binnen het DEV agentschap moeten erop gericht zijn om overeenstemming te bereiken over de na te streven doeleinden (criteria) en over de middelen die daartoe moeten ingezet worden. Naarmate het agentschap erin slaagt om overeenstemming te bereiken zal ook de relevantie en legitimiteit in het politieke en maatschappelijke debat toenemen. Hierbij is het echter niet van belang dat ten allen prijze een consensus bereikt wordt over elke afzonderlijke aanbeveling naar de overheid. Omdat conflicten of dissensus vaak een teken zijn dat bepaalde materies of keuzes in het brandpunt van een samenwerkingsverband staan is het net van belang om deze conflicten scherp en transparant te krijgen. Tevens is het van belang dat het ‘sluiten’ van het debat toch ruimte laat aan leerprocessen. Dit kan men bijvoorbeeld bereiken door criteria en doelstellingen zodanig te formuleren dat ze ruimte laten voor verdere interpretatie en articulatie in het licht van nieuwe bevindingen. Om collectief leren te bevorderen is een periodieke evaluatie van zowel de beleidsmaatregelen (aan de hand van de duurzaamheidcriteria en –indicatoren), de interactieregels als de kennisbasis waarop de interacties gebaseerd zijn nodig. Niettegenstaande de beste pogingen om gedegen beslissingen te nemen zal altijd een onvermijdelijke onzekerheid blijven bestaan. Dit gegeven zet aan tot een reflectie omtrent diversificatie van opties en omkeerbaarheid van beslissingen. Eenmaal ingezet vertonen technologische trajecten immers een grote vorm van weerbarstigheid t.o.v. veranderingen (d.i. het fenomeen van de technologische ‘lock-in’).

Samenvatting xvii

5.2.7

Een collectief leerproces

Aanbeveling nr. 7: Collectieve leerprocessen kunnen bevorderd worden door de volgende punten in acht te nemen: een wederzijdse uitwisseling van informatie tussen het DEV agentschap en de ‘buitenwereld’; het bevorderen van concrete socio-materiële experimenten; en het verzekeren van een koppeling tussen alle stadia van de beleidsvormingscyclus.

Het oprichten van een nieuw orgaan (een DEV agentschap), met adviserende functie naar de politieke overheid (zonder de politieke verantwoordelijkheid over te nemen) is dus volgens ons een absoluut noodzakelijke (en tot nu toe ontbrekende) maatregel om de nodige continuïteit en kwaliteit in het tot nu toe weinig gestructureerde (kern)energiedebat te injecteren. De oprichting en werking van dergelijk energie-agentschap vormde ook de kern van onze aanbevelingen. Toch wensen we hier enkele kanttekeningen bij te plaatsen, die we binnen het bestek van dit proefschrift echter niet verder konden uitwerken. Ten eerste kan het niet de bedoeling zijn dat dit orgaan over het gehele spectrum van het (kern)energiedebat een coördinerende rol opneemt. Ook hier zal een ‘subsidiariteitsbeginsel’ in acht moeten genomen worden. De vele organisaties actief op het terrein van energie zouden de doelstelling van duurzame ontwikkeling moeten integreren in hun alledaagse taakuitvoering. Het oprichten van specifieke organen mag niet als alibi fungeren voor het ontlopen van die verantwoordelijkheid. Tevens kan het oprichten van het ‘agentschap voor duurzame energie’ een aanleiding vormen om bestaande organen te herorganiseren, bijvoorbeeld in het geval van overlappende competenties of verantwoordelijkheden. Bovendien is het van belang om de operationalisering van beleidsmaatregelen of beloftevolle technologieën (die bijvoorbeeld geïdentificeerd werden in de strategische visievorming) op te volgen in concrete socio-materiële contexten. Hierbij kan eventueel aansluiting gezocht worden bij bestaande (maar vooral academische) modellen zoals het ‘strategisch niche management’, waarbij een beloftevolle technologie eerst uitgetest wordt in ‘beschermde’ omstandigheden alvorens ze in een eventueel aangepaste versie aan de werking van competitieve markten wordt blootgesteld. Tenslotte hebben we de vraag opengelaten in hoeverre allerlei procedures voor burgerparticipatie in technologische keuzes (zoals burgerjury’s, consensusconferenties, lokale Agenda 21-initiatieven, enzovoort) kunnen bijdragen aan collectieve leerprocessen. Dergelijke procedures hebben een sterke opgang gemaakt in het afgelopen decennium, maar de theorievorming omtrent hun concrete impact op beleidvorming of het veranderen van gedragpatronen staat nog in de kinderschoenen.

xviii


6 Algemeen besluit De vraag naar een meer duurzame ontwikkeling is bijzonder complex. Meer bepaald zet zij aan tot een geïntegreerde bevraging van gangbare praktijken op zowel een metatheoretisch, theoretisch als empirisch niveau. Het duurzaamheiddebat verloopt voorlopig te fragmentarisch en weinig consistent, waarbij belangrijke aandachtspunten (met name de institutionele dimensie van duurzame ontwikkeling die we in dit proefschrift behandeld hebben) onderbelicht blijven. Zonder een gestructureerd beleidsdebat geleid vanuit een streven naar duurzame ontwikkeling blijft het zeer moeilijk de optimale rol van kernenergie in België te bepalen. Nochtans is deze vraag te belangrijk om over te laten aan de dynamiek van een gepolariseerd debat, waar enkel diametraal tegengestelde voorstellen een kans maken. De voorziening van een samenleving met energie en met elektriciteit is immers nauw verweven met de dimensies van duurzame ontwikkeling.

CURRICULUM VITAE Erik LAES Adres: Schomstraat 39 2600 BERCHEM Geboren : Wilrijk, 4 oktober, 1975

E-mail: [email protected] Tel. : 00 32 486 90 00 75

Nationaliteit : Belg

1.

Professioneel Sept. 1999 - Sept. 2003:

Aspirant Wetenschappelijk Medewerker (AWM) bij het SCK•CEN (Studiecentrum voor Kernenergie/Centre d’Etude de l’Energie met maatschappelijke zetel in de Herrmann Debrouxlaan 40-42, 1160 Brussels en laboratoria in de Boeretang 200, 2400 Mol). Onderzoeksopdracht in het kader van een doctoraatsproject onder begeleiding van de KULeuven (Prof. Dr. ir. W. D'haeseleer en Prof. Dr. ir. R. Weiler). Okt. 2004 –April 2005 : Deeltijds onderzoeker KULeuven (Afdeling Toegepaste Mechanica en Energie-conversie) / SCK•CEN in het kader van een onderzoeksproject in opdracht van viWTA. Vanaf april 2005: Wetenschappelijk medewerker bij het SCK•CEN. Onderzoeksopdrachten in het kader van Europese Kaderprogramma’s en in opdracht van viWTA. 2.

Opleiding 1998-1999 : KULeuven Gediplomeerde in Aanvullende Studies (GAS) milieukunde-milieubeheer

1993-1998 : KULeuven Burgerlijk scheikundig ingenieur

1997-1998 : KULeuven Certificaat Filosofische Academie

xx


3.

Publicaties PUBLICATIE IN INTERNATIONAAL TIJDSCHRIFT OF PROCEEDINGS MET PEER REVIEW 1.

2.

3.

4. 5. 6. 7.

Laes, E. and Meskens, G. (2001), “Nuclear energy and sustainable development : contradiction or challenge ?”, Proceedings of the Global 2001 International Conference on "Back-End of the Fuel Cycle: From Research to Solutions" (Paris, Sept. 9-13) (CDROM). Laes, E. and Meskens, G. (2001), “Sustainable development and radioactive waste management : challenges for the future”, Proceedings of the 8th International Conference on Radioactive Waste Management and Environmental Remediation (ICEM '01) (Bruges, Sept. 30 – Oct. 4) (CD-ROM). Laes, E., Meskens, G., D’haeseleer, W. and Weiler, R. (2002), “The Belgian nuclear phase out as a strategy for sustainable development: Unstructured problems, unstructured answers?”, in Biermann, F., Campe, S. and Jacob, K. (Eds.), Proceedings of the 2002 Berlin Conference on the Human Dimensions of Global Environmental Change “Knowledge for the Sustainability Transition. The Challenge for Social Science”, Global Governance Project, Amsterdam/Berlin/Potsdam/Oldenburg, pp. 271-284. Laes, E. and Meskens, G. (2003), “An architecture for sustainability assessment : the case of nuclear energy in Belgium”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 501-511. Bovy, M., Laes, E. and Eggermont, G. (2003), “Science, objectivity and neutrality in nuclear research: Shared references or ideal types?”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 433-443. Laes, E., Meskens, G., D’haeseleer, W. and Weiler, R. (2004), “Trust as a central paradigm for advisory science: The case of the Belgian nuclear phase out”, International Journal of Sustainable Development, Vol. 7, No. 1, pp. 1-26. Laes, E., D’haeseleer, W. and Weiler, R. (2005), “Addressing uncertainty and inequality in nuclear policy”, Journal of Enterprise Information Management, Vol. 18, No. 3, pp. 357376.

ONDERZOEKSRAPPORTEN 1.

Laes, E., Chayapathi, L., Meskens, G. en Eggermont, G. (2004), Kernenergie en maatschappelijk debat, rapport in opdracht van het Vlaams Instituut voor Technologisch Aspectenonderzoek (viWTA), viWTA, Brussel.

HOOFDSTUK IN BOEK 1.

Laes, E., Meskens, G., and Ruan, D. (2004) “Planning for sustainability in the Belgian electricity sector: a Multi-criteria mapping exercise”, in D. Ruan and X. Zeng (eds.) "Intelligent Sensory Evalution – Methodologies and Application", Springer Verlag, Berlin/ Heidelberg/New York, pp. 311-326.

It is not a question of annihilating science, but of controlling it. Science is totally dependent upon philosophical opinions for all of its goals and methods, though it easily forgets this. (Friedrich Nietzsche)

People disagreeing everywhere you look, Makes you wanna stop and read a book. (Bob Dylan)

VOORWOORD

Een proefschrift mag dan al het resultaat zijn van vele uren “eenzame arbeid”; toch zou het nooit tot stand komen zonder de steun, de inspiratie en de inzet van vele mensen. Dit geldt natuurlijk in nog sterkere mate voor de poging tot interdisciplinair werk waarvan u hier de neerslag vindt. Iedereen die mij in dit werk heeft bijgestaan wil ik van harte danken. In het bijzonder wil ik mijn beide promotoren Prof. Dr. ir. William D’haeseleer en Prof. Dr. ir. Raoul Weiler danken om de mogelijkheid die zij geboden hebben om dit interdisciplinaire onderzoek uit te voeren. Hun interesse en kritische lezing van de verschillende ‘tussenproducten’ was van onschatbare waarde. Verder gaat mijn dank ook uit naar de collega’s van het PISA programma op het StudieCentrum voor Kernenergie (SCK). Gilbert Eggermont, initiatiefnemer van PISA, dank ik om meer dan wie ook mijn horizon verbreed te hebben met zijn veelheid aan ideeën en kritische inzichten. Gaston Meskens, mijn mentor op het SCK, voor zijn niet aflatend zoeken en denken buiten de lijnen. Mijn collega(post-)doctoraatsonderzoekers en dagelijkse gesprekspartners: Michel, Gunter, Steven, Isabelle, Chloée en Koen. En verder de PISA-collega’s die bijgedragen hebben tot een aangename, stimulerende en dynamische werkomgeving: Ludo, Benny, Frank, Geert, Cathérine, en Véronique. Ook de collega’s van STEM, Matthieu Craye en Lieve Goorden, die het empirische luik van dit onderzoek methodologisch hebben begeleid mogen niet in het lijstje ontbreken. Dit proefschrift draagt ook de herinnering aan de boeiende contacten met de vele respondenten in het kader van mijn empirisch onderzoek. Ik heb hun vrijwillige inzet enorm gewaardeerd. Ook de leden van de stuurgroep die dit onderzoek vooral in een vroeg stadium hebben begeleid wil ik danken. Mijn ouders en broer dank ik voor de steun bij al mijn studies. Sabine en Ruben: dank voor de fijne momenten samen weg van de zwaarwichtigheid van een proefschrift en bij deze sorry voor mijn momenten van afwezige aanwezigheid.

Erik Laes Oktober 2006

CONTENTS

SAMENVATTING 1

Doel en relevantie van het onderzoek........................................................................ii

2

Probleemstelling en onderzoeksvragen.....................................................................ii

3

Methodologische uitgangspunten .............................................................................iv

4

Uitwerking van de onderzoeksvragen in een drieluik .............................................v

5

Samenvatting, conclusies en aanbevelingen .............................................................v 5.1

Samenvatting en conclusies....................................................................................vi

5.1.1

Constructivisme als (meta-)theoretisch uitgangspunt .....................................vi

5.1.2

Vier bestuursmodellen gewikt en gewogen....................................................vii

5.1.3

Het marktmodel als ultieme scheidsrechter?...................................................ix

5.1.4

Een blik op de beleidvormingspraktijk ............................................................x

5.2

Aanbevelingen ........................................................................................................xi

5.2.1

Een ‘agentschap voor duurzame energie’.......................................................xii

5.2.2

Open en flexibele relaties met de ‘buitenwereld’..........................................xiii

5.2.3

Ondersteuning van deliberaties door criteria voor duurzame energie en multi-criteria technieken................................................................................xiii

6

5.2.4

Hervorming van toekomstverkenningspraktijken .........................................xiv

5.2.5

Het belang van vertrouwen.............................................................................xv

5.2.6

Tijdelijke afsluiting van debatten ..................................................................xvi

5.2.7

Een collectief leerproces ..............................................................................xvii

Algemeen besluit…………………………………………………………………..xviii

CHAPTER 0 - PROBLEM FRAMING, RESEARCH QUESTIONS AND METHODOLOGY 1

Purpose and relevance................................................................................................2

xxvi

Nuclear energy and sustainable development

2

Sustainability from a bird’s-eye view ....................................................................... 5 2.1

Historical and conceptual background .................................................................... 5

2.2 The international political agenda (from Rio 1992 to Johannesburg 2002 and beyond) ..................................................................................................................... 6 3

Problem delineation and research questions ........................................................... 8

4

Methodological considerations ................................................................................ 13

5

Structure.................................................................................................................... 14

CHAPTER 1 - THE PRINCIPLE(S) OF SUSTAINABLE (META-)THEORETICAL FRAMEWORK

DEVELOPMENT:

TOWARDS

A

1

Introduction .............................................................................................................. 18

2

Characteristics of sustainable development ........................................................... 22 2.1

3

Appeals to sustainability ....................................................................................... 23

2.1.1

Sustainability as a ‘manifest image’ of a form of life.................................... 26

2.1.2

Sustainability as a ‘vision’ ............................................................................. 29

2.1.3

Sustainability as a ‘policy target’ or ‘goal’.................................................... 30

2.2

Defining sustainable development? ...................................................................... 32

2.3

An operational list of characteristics..................................................................... 34

Sustainable development as a principle of justification ........................................ 36 3.1

The ‘commonwealth model’ of Boltanski and Thévenot...................................... 38

3.1.1

Framing the problem of justification ............................................................. 38

3.1.2

The ‘common grammar’ of the commonwealths........................................... 40

3.1.3

From the commonwealth(s) to the common world(s).................................... 42

3.1.4

Criticism, conflict, clarification, bargain and compromise............................ 45

3.1.5

Theoretical implications and evaluation ........................................................ 47

3.2

Beck’s theory of risk society................................................................................. 49

3.2.1

A brief guide to risk society........................................................................... 50

3.2.2


3.3

(Social) constructivist studies of science and technology ..................................... 58

3.3.1

A brief guide to (social) constructivist studies of technology........................ 59

3.3.2


3.4

Constructivism as a (meta-)theoretical framework ............................................... 71

Contents xxvii

3.5 4

In search of the ‘sustainable commonwealth’?......................................................79

Summary and conclusions........................................................................................85

CHAPTER 2 - IN

SEARCH OF SUSTAINABLE ENERGY POLICY:

FOUR

THEORIES OF

GOVERNANCE

1

Introduction...............................................................................................................88

2

Technology policy problems and policy-making strategies...................................90

3

2.1

A typology of uncertainty......................................................................................90

2.2

Structuring policy problems ..................................................................................93

2.3

The policy development cycle ...............................................................................97

Four governance schemes for sustainable development......................................101 3.1

3.1.1

Theoretical background................................................................................102

3.1.2

A historical example: nuclear safety regulation ...........................................107

3.1.3

The contribution of expert-based governance to the sustainability debate ..115

3.1.4

Conclusion....................................................................................................118

3.2

Governance by aggregation .................................................................................120

3.2.1


3.2.2

Contribution of governance by aggregation to the sustainability debate .....125

3.2.3

Conclusion....................................................................................................140

3.3

Governance by pacification .................................................................................141

3.3.1


3.3.2

An example: investment planning in the Belgian electricity sector .............143

3.3.3

Contribution of governance by pacification to the sustainability debate .....147

3.3.4

Conclusion....................................................................................................151

3.4

4

Expert-based governance.....................................................................................102

Deliberative governance ......................................................................................152

3.4.1


3.4.2

Contribution of deliberative governance to the sustainability debate ..........156

3.4.3

Conclusion....................................................................................................168

Summary and conclusions......................................................................................169

xxviii


CHAPTER 3 - THE EXTERNE

METHODOLOGY AS A DECISION SUPPORT FOR

SUSTAINABLE ENERGY POLICY

1

Introduction ............................................................................................................ 174

2

Theoretical background......................................................................................... 175

3

4

5

6

2.1

Definitions........................................................................................................... 175

2.2

Major difficulties................................................................................................. 178

2.2.1

Representing the interests of affected populations ...................................... 179

2.2.2

Quantifying the interests of affected populations ........................................ 186

The ExternE project............................................................................................... 191 3.1

Life cycle analysis............................................................................................... 192

3.2

The impact pathway approach ............................................................................ 193

3.3

Results................................................................................................................. 195

3.4

Major uncertainties and general robustness of ExternE results .......................... 199

3.4.1

Health impacts of air pollution .................................................................... 199

3.4.2

Global warming ........................................................................................... 200

3.4.3

Discounting.................................................................................................. 201

3.4.4

Conclusion ................................................................................................... 202

Externalities of the nuclear fuel cycle................................................................... 202 4.1

Major nuclear accidents ...................................................................................... 207

4.2

Decommissioning and dismantling of nuclear power plants............................... 208

4.3

Nuclear waste management................................................................................. 210

4.4

Proliferation of nuclear materials........................................................................ 214

4.5

Conclusions......................................................................................................... 215

Externalities as a decision support........................................................................ 216 5.1

Improvement of the scientific policy base .......................................................... 217

5.2

Energy and pollution taxes.................................................................................. 219

5.3

Subsidies ............................................................................................................. 221

5.4

Standard setting................................................................................................... 222

5.5

Research and development planning................................................................... 224

5.6

Voluntary action.................................................................................................. 224

Summary and conclusions ..................................................................................... 225

Contents xxix

CHAPTER 4 - A RECONSTRUCTION OF THE CASE OF THE BELGIAN PHASE-OUT LAW

POLICY DEVELOPMENT CYCLE IN THE

1

Introduction.............................................................................................................229

2

Socio-political background to the phase-out law .................................................231

3

4

2.1

General structure of electricity sector..................................................................232

2.2

Liberalisation of the energy markets ...................................................................234

2.3

European perspectives on nuclear power ............................................................236

2.4

Public opinion......................................................................................................237

The policy development cycle in the case of the Belgian nuclear phase out ......238 3.1

The AMPERE report ...........................................................................................240

3.2

The federal plan for sustainable development .....................................................243

3.3

Justification of the phase-out law as a well-structured problem..........................244

3.3.1

The phase-out law in a nutshell....................................................................244

3.3.2

Recasting the problem in a technical mould.................................................246

Interviews with members of the FRDO ................................................................250 4.1

Observations ........................................................................................................251

4.2

Interview results ..................................................................................................253

4.2.1

Economic issues ...........................................................................................253

4.2.2

Technology...................................................................................................255

4.2.3

Governance...................................................................................................257

4.2.4

Socio-cultural reflections .............................................................................273

4.3 5

Discussion............................................................................................................275


CHAPTER 5 - A PROPOSAL FOR A NEW GOVERNANCE STRUCTURE AS A SUPPORT FOR SUSTAINABLE ENERGY POLICY

1

Introduction.............................................................................................................286

2

Requirements ..........................................................................................................291

3

A model process ......................................................................................................295 3.1

Selection and structuring of policy problems ......................................................296

3.1.1

Invoking the precautionary principle ...........................................................297

xxx


3.1.2 3.2

Constructing scenarios ................................................................................. 308

3.2.2

Definition, comparison and selection of strategic options ........................... 310

3.2.3

Definition, comparison and selection of tactical options............................. 315

Political decisions................................................................................................ 318

3.3.1

The choice of standards ............................................................................... 319

3.3.2

The choice of potential measures................................................................. 320

3.3.3

Tactical considerations in political decision making ................................... 322

3.4

Initiation of policy instruments and organisation................................................ 324

3.4.1

Monitoring ................................................................................................... 325

3.4.2

Evaluation .................................................................................................... 328

3.5

5

Definition, comparison and selection of options................................................. 307

3.2.1

3.3

4

Structuring problems for sustainability........................................................ 302

Detection of matters of concern .......................................................................... 329

Quality assurance ................................................................................................... 331 4.1

Fairness ............................................................................................................... 332

4.2

Competence......................................................................................................... 332

4.3

Uncertainty management .................................................................................... 333

4.4

Transparency ....................................................................................................... 334

Roles of different actors ......................................................................................... 335 5.1

Scientific experts ................................................................................................ 336

5.2

Ethicists............................................................................................................... 337

5.3

Stakeholder representatives................................................................................. 338

5.4

Administrators..................................................................................................... 340

5.5

Facilitators........................................................................................................... 341

5.6

Politicians............................................................................................................ 342

6

A practical proposal ............................................................................................... 342

7

Possible pitfalls ....................................................................................................... 347

8

7.1

‘Your proposal will suffer from a lack of political and/or social support’.......... 347

7.2

‘Your proposal is unethical’................................................................................ 348

7.3

‘Your proposal expects too much of intentional political planning’................... 349

7.4

‘Your proposal quite simply comes too late’ ...................................................... 350


Contents xxxi

CHAPTER 6 - CRITERIA

AND SCENARIOS AS A SUPPORT FOR SUSTAINABLE

ENERGY GOVERNANCE

1

Introduction.............................................................................................................355

2

A structured value tree for sustainable energy criteria.......................................356

3

2.1

Methodological approach ....................................................................................356

2.2

Construction of the combined value tree .............................................................358

Sustainable energy scenarios .................................................................................362 3.1

3.1.1

Methodological approach .............................................................................363

3.1.2

Choice of energy model ...............................................................................367

3.1.3

Limitations of the modelling approach ........................................................368

3.2

Background information on framing assumptions ..............................................369

3.2.1

Demographics...............................................................................................370

3.2.2

Assumptions for energy service demands ....................................................371

3.2.3

Resource availability and energy prices .......................................................373

3.2.4

Discount rates...............................................................................................374

3.3

Using the ‘building blocks’ and ‘framing assumptions’ to construct scenarios ..377

3.3.1

Strategic goals ..............................................................................................377

3.3.2

Demand side analysis ...................................................................................378

3.3.3

Supply side analysis .....................................................................................380

3.3.4

Policy measures............................................................................................385

3.3.5

Visions beyond 2050 ....................................................................................385

3.4

4

Introduction .........................................................................................................362

Results .................................................................................................................386

3.4.1

Selected results per option............................................................................387

3.4.2

Comparison between scenarios ....................................................................401

3.4.3

Detailed results (including sensitivity analysis) ...........................................404


CHAPTER 7 - LONG-TERM

OPTIONS FOR CRITERIA MAPPING EXERCISE

1

BELGIAN

ENERGY POLICY:

A

MULTI-

Introduction.............................................................................................................419

xxxii

2


Methodology............................................................................................................ 423 2.1

2.1.1

Identification of stakeholders....................................................................... 426

2.1.2

Identification of options for action .............................................................. 427

2.1.3

Identification of decision attributes ............................................................. 429

2.2

3

4

Operation of the MCM model............................................................................. 430

2.2.1

Identification of empirical indicators for the attributes (‘scoring’) ............. 431

2.2.2

Assigning weights to the attributes (‘weighting’)........................................ 432

2.2.3

Analysis of results........................................................................................ 433

Results ..................................................................................................................... 434 3.1

The scenarios....................................................................................................... 434

3.2

Selection of criteria ............................................................................................. 436

3.3

Scoring ................................................................................................................ 439

3.4

Weightings .......................................................................................................... 441

3.5

Rankings.............................................................................................................. 444

3.6

Sensitivity analysis.............................................................................................. 449

3.6.1

Sensitivity analysis: uncertainties in scoring ............................................... 450

3.6.2

Sensitivity analysis: importance weightings ................................................ 452

Discussion ................................................................................................................ 453 4.1

The ‘product’....................................................................................................... 453

4.1.1

The combined value tree.............................................................................. 453

4.1.2

A new direction for scenario development .................................................. 454

4.1.3

Eliciting new and inclusive knowledge perspectives................................... 456

4.1.4

The ranking profiles..................................................................................... 458

4.2

5

Development of the MCM model ....................................................................... 426

The ‘process’....................................................................................................... 459

4.2.1

Reflections related to the ‘experimental’ nature of the exercise.................. 459

4.2.2

The importance of trust................................................................................ 460

4.2.3

Interaction and deliberation in the participative process.............................. 462

4.2.4

Using the MCM method for decision support ............................................. 463


Contents xxxiii

CHAPTER 8 - SUMMARY,

CONCLUSIONS, RECOMMENDATIONS AND FURTHER

RESEARCH

1


2

Recommendations...................................................................................................476

3

Suggestions for further research ...........................................................................485

4

Epilogue ...................................................................................................................487

REFERENCES & BIBLIOGRAPHY…………………………………………………489 ANNEX 1………………………………………………………………….………507 ANNEX 2………………………………………………………………………….519 ANNEX 3………………………………………………………………………….521 ANNEX 4………………………………………………………………………….533 ANNEX 5………………………………………………………………………….537 ANNEX 6………………………………………………………………………….547

LIST OF FIGURES

Figure 1. The policy development cycle............................................................................. 100 Figure 2. Ambiguity in the ranking of electricity supply options in the literature on externalities ............................................................................................................... 187 Figure 3. The impact pathway approach ........................................................................... 194 Figure 4. Reconstruction of policy development in the case of the Belgian nuclear phase out ............................................................................................................................. 231 Figure 5. The evolution of electricity demand in Belgium................................................. 233 Figure 6. Organisation of the electricity sector in Belgium (in 2003)............................... 234 Figure 7. Determination of the demand for energy service ............................................... 372 Figure 8. Primary energy demand – scenarios M.K.P.CS and R.K.P.CS......................... 388 Figure 9. Electricity production – scenarios M.K.P.CS and R.K.P.CS ............................ 389 Figure 10. Primary energy demand – scenario M.K.LCS and R.K.LCS .......................... 391 Figure 11. Electricity production – scenario M.K.LCS and R.K.LCS ............................... 393 Figure 12. Primary energy demand – scenario M.K.P.LCS and R.K.P.LCS..................... 395 Figure 13. Electricity production - scenario M.K.P.LCS and R.K.P.LCS ........................ 397 Figure 14. Primary energy demand – scenario M.K.P.LCS.I and R.K.P.LCS.I ............... 399 Figure 15. Electricity production - scenario M.K.P.LCS.I and R.K.P.LCS.I .................. 400 Figure 16. Final energy use (all scenarios)....................................................................... 402 Figure 17. Electricity production (all scenarios)............................................................... 403 Figure 18: Possible dynamics resulting from the confrontation of different networks...... 422 Figure 19. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants ................. 445

LIST OF TABLES

Table 1. Pezzoli’s ten categories of literature on sustainable development........................12 Table 2. The six worlds in Boltanski and Thévenot’s theory of justification.......................44 Table 3. The different dimensions of uncertainty. ...............................................................93 Table 4. A typology of policy problems and approaches ....................................................96 Table 5. The multi-dimensional character of sustainability..............................................180 Table 6. External costs of non-renewable electricity production technologies in Belgium ....................................................................................................................................197 Table 7. External costs of renewable electricity production technologies in Belgium......198 Table 8. External costs of the nuclear fuel cycle in Belgium (without reprocessing) .......204 Table 9. Distribution of impacts from routine operation of the nuclear fuel cycle ...........204 Table 10. Summary of the reasoning behind the Belgian nuclear phase out ....................244 Table 11. Methodological differences in the approaches used by the AMPERE Commission and the Wuppertal Institute ..................................................................265 Table 12. Strengths and weaknesses of different modelling approaches to the representation of carbon emissions and energy use of a national economy..............368 Table 13. Number of households and population (millions) .............................................371 Table 14. Evolution of the demand for energy services – all scenarios (% average annual growth) .......................................................................................................................373 Table 15. Resource prices for all scenarios ......................................................................374 Table 16. Sector specific discount rates in ‚Market Drive’ scenario group .....................376 Table 17. Key differences between ‚Rational Perspective’ and ‚Market Drive’ scenario 376 Table 18. Technical energy saving potential (industry) ....................................................378 Table 19. Technical energy saving potential (residential and service sector) ..................379 Table 20. Constraints on technological options in both worldviews ................................385 Table 21. CO2 emissions – all sectors (scenario M.K.P.CS and R.K.P.CS ).....................390 Table 22. CO2 emissions – all sectors (scenario M.K.LCS and R.K.LCS ) ......................394 Table 23. CO2 emissions – all sectors (scenario M.K.P.LCS and R.K.P.LCS ) ...............398 Table 24. CO2 emissions – all sectors (scenario M.K.P.LCS.I and R.K.P.LCS.I)) ..........401 Table 25. Resource prices (sensitivity analysis)................................................................404 Table 26. External costs of air pollutants .........................................................................405 Table 27. Impacts on public health caused by air pollution .............................................406 Table 28. Impacts on occupational health (coal and gas fuel cycle) ................................406

xxxvi


Table 29. Radiological health impacts of nuclear fuel cycle ............................................ 407 Table 30. Need for very long-term management of high-level waste (HLW) ................... 408 Table 31. Impacts on ecosystems caused by air pollution ................................................ 408 Table 32. Environmental impacts caused by solid waste from coal fuel cycle ................. 409 Table 33. Catastrophic risk (nuclear)............................................................................... 410 Table 34. Intensity of energy use ...................................................................................... 410 Table 35. Security of energy supply .................................................................................. 411 Table 36. Total system costs ............................................................................................. 411 Table 37. Marginal costs of electricity ............................................................................. 412 Table 38. The need for intermediary storage of spent fuel ............................................... 413 Table 39. Use of non-renewable resources....................................................................... 413 Table 40. Degree of decentralisation of electricity production ........................................ 414 Table 41. Diversity of electricity production park ............................................................ 414 Table 42. Weightings used by all participants (intermediate level).................................. 442 Table 43. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on average scores).......................................................................................................... 444 Table 44. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on ‘optimistic’ scores) .................................................................................................... 450 Table 45. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on ‘pessimistic’ scores)................................................................................................... 451 Table 46. Uncertainties expressed for each scenario according to each participant) ..... 451

LIST OF ABBREVIATIONS ABVV ACLVB ACV AVN BCEO BS BSE CCEG CCS CDM CHP CREG CV DAD DPWB DSM EC ECCS EEA EPR ES ETS EU ExternE FANC FBA FGD FPB FRDO GDP GHG GMO GT-MHR GW(e)

Algemeen Belgisch VakVerbond (Belgian socialist labour union) Algemene Centrale der Liberale Vakverbonden van België (Belgian liberal labour union) Algemeen Christelijk Vakverbond (Belgian christian labour union) Association Vinçotte Nucléaire BeheersComité van ElektriciteitsOndernemingen (Belgian management committee of electricity producers) Belgisch Staatsblad (Belgian official journal) Bovine Spongiform Encephalopathy (or ‘mad cow disease’) ControleComité voor Elektriciteit en Gas (Belgian control committee for electricity and gas) Carbon Capture and Storage Clean Development Mechanism Combined Heat and Power production Commissie voor de Regulering van Elektriciteits-en Gasmarkten (Belgian regulatory commission for the electricity and gas markets) Contingent Valuation Decide Announce Defend Dienst voor de Planning van het WetenschapsBeleid (Belgian federal services for the planning of science policy) Demand Side Management European Commission (of the European Union) Emergency Core Cooling System European Environment Agency European Pressurised water Reactor Environmental Space Emission Trading System European Union Externalities of Energy Federaal Agentschap voor Nucleaire Controle (Belgian federal agency for nuclear control) Furnace Bottom Ash Flue Gas Desulphurisation Federaal Planbureau (Belgian federal planning bureau) Federale Raad Duurzame Ontwikkeling (Belgian federal council for sustainable development) Gross Domestic Product GreenHouse Gas Genetically Modified Organism Gas Turbine-Modular Helium Reactor GigaWatt (electric)

xxxviii


HTR IAEA ICDO ICRP IEA IGCC ILO IMF IPCC kWh LCA LLFP LOCA LRTAP MA MHTGR MIT MW(e) NCE NEA NGO NIMBY NIMTOO NIRAS NMVOC NPT OECD PBMR PFA PISA PM PSA pTA PWR RARG SCK•CEN SCOT SME STAG SuND TMI

High-Temperature Reactor International Atomic Energy Agency Interdepartementale Commissie Duurzame Ontwikkeling (Belgian interdepartmental commission for sustainable development) International Commission on Radiological Protection International Energy Agency Integrated Gasification Combined Cycle International Labour Organisation International Monetary Fund International Panel on Climate Change kiloWatt-hour Life Cycle Analysis Long-Lived Fission Products Loss-Of-Coolant Accident Long-Range Transboundary Air Pollution Minor Actinides Modular High-Temperature Gas-cooled Reactor Massachussets Institute of Technology MegaWatt (electric) Nationaal Comité Energie (Belgian national energy committee) Nuclear Energy Agency (of the OECD) Non-Governmental Organisation Not In My Backyard Not In My Term Of Office Nationale Instelling voor Radioactief Afval en verrijkte Splijtstoffen (Belgian agency for radioactive waste and enriched fissile materials) Non-Methane Volatile Organic Carbon Non-Proliferation Treaty Organisation for Economic Cooperation and Development Pebble-Bed Modular Reactor Pulverised Fuel Ash Program of Integration of Social Aspects into nuclear research Particulate Matter Probabilistic Safety Assessment Participatory Technology Assessment Pressurised Water Reactor Risk Assessment Review Group StudieCentrum voor Kernenergie / Centre d’Etude de l’énergie Nucléaire (Belgian nuclear research centre) Social Construction of Technology Small and Medium-sized Enterprises STeam And Gas power plant Sustainability and Nuclear Development Three Mile Island

List of abbreviations xxxix

UK UN UNCED UNCSD UNEP UNFCCC US USACRS USAEC USC USDOE USEPA USNRC USSR VBO viWTA VREG WCED WHO WTO WSSD

United Kingdom United Nations United Nations Conference on Environment and Development United Nations Commission on Sustainable Development United Nations Environmental Programme United Nations Framework Convention on Climate Change United States (of America) United States Advisory Committee on Reactor Safeguards United States Atomic Energy Commission Ultra Super-Critical coal-fired power plant United States Department of Energy United States Environmental Protection Agency United States Nuclear Regulatory Commission (former) Union of Socialist Soviet Republics Verbond van Belgische Ondernemingen (Belgian employers’ organisation) Vlaams Instituut voor Wetenschappelijk en Technologisch Aspectenonderzoek (Flemish parliamentary technology assessment institute) Vlaamse commissie voor de Regulering van Elektriciteits- en Gasmarkten (Flemish commission for the regulation of electricity and gas markets) World Commission on Environment and Development World Health Organisation World Trade Organisation World Summit on Sustainable Development

CHAPTER 0 PROBLEM FRAMING, RESEARCH QUESTIONS AND METHODOLOGY In the academic year 1998-1999, while I was working on my master’s thesis, I spent a considerable amount of time trying to somehow get an overview of sustainability-related literature in order to develop a coherent account of the concept. With hindsight, I must admit that I was largely unaware of the complexity of the task I had set for myself. What I remember mostly from this period is a mixed feeling of frustration and pleasure that came out of trying to follow the innumerable strands of thinking to which the concept of sustainability seemed to be linked 1 . And then, in the fall of 1999, the co-promoter of this dissertation pointed out a possibility for doing further research in this field. At that time the Belgian nuclear research centre (SCK•CEN) had just started a project on the integration of social sciences in nuclear research (later named the ‘Program of Integration of Social Aspects into nuclear research’ (PISA)), and in one of the project’s research tracks (‘Sustainability and Nuclear Development’ (SuND)) the opportunity was offered for starting PhD research regarding the question whether nuclear power could contribute to a sustainable energy future. To me this seemed to be the ideal opportunity to continue the reflections I had taken up earlier, so I seized it with both hands. At that time, the question to me still seemed to be relatively straightforward. As I took it, the title of the research track in which I was engaged (‘Sustainability and Nuclear Development’) suggested the image of two sets of concepts – one related to ‘sustainability’, the other to ‘nuclear development’. Following this line of reasoning, my task then consisted in finding the concepts (if any) in the section of both sets (finding the ‘and’ in SuND) – i.e. a subset of ‘nuclear developments’ which could be qualified as being ‘sustainable’. However, in the course of my research I slowly started to realise (again, to my frustration) that this approach to the problem basically starts from the wrong premise, in the sense that it erroneously leaves unexamined the notion of a ‘set’ – i.e. an integrating principle from which all concepts related to ‘sustainability’ or ‘nuclear development’ could be seen in the ‘right’ perspective. Not only value conflicts (related to the choice of what is considered to be a ‘sustainable’ option), but also opposing ‘factual’ assessments point out that the definition of such an undisputed ‘set’ (or ‘set of sets’, or etc. – cf. Section 3) remains elusive 2 . Of course, the mere existence of such diverging sets does not necessarily imply that one should resign to this situation. One could acknowledge the existence of different views and methodologies, but still argue in favour of only one perspective or method (or a non-contradictory combination of methods), according to its ‘obvious’ capacities for

1

Perhaps we should set out right at the beginning of this dissertation that we use the two terms ‘sustainable development’ and ‘sustainability’ as synonyms. Admittedly, some authors make rigid distinctions, but there is no evident consistency of difference. There is no consensus on which term is broader, which term carries most ‘undesirable baggage’, etc. The debates about whether and how the usages have differed (or should differ) are unresolved (and, quite frankly, seem to be of peripheral interest to us). 2 A ‘set’ could be taken to mean ‘perspectives’, ‘pillars’, ‘methods’, ‘approaches’, ‘discourses’, ‘criteria’, ‘stakeholder interests’, ‘spatio-temporal scales’ – all of which figure prominently in the sustainability debate.

2

Chapter 0

solving the problem at hand. In other words, we 3 first had to tackle the question whether we could assume (and on which grounds) that there is some criterion of objective validity (i.e. one that any ‘reasonable’, ‘well-informed’, etc. person should adhere to) which dictates how to solve the question of the role of nuclear power in a sustainable energy future. Answering this question in itself already poses a significant challenge; but, unfortunately for us, the problems did not end there. Intimately related to the first question is a second one, which relates to the role of the researcher trying to solve problems of sustainability: can one assume that a sufficiently intelligent person is able to generate answers to the research question by him- or herself, or does a solution (or part of the solution) necessarily emerge from a process of deliberation among different perspectives? Taken together, these two questions (further specified in section 3) have been the fundamental ‘driving force’ behind our research efforts. If anything, the pages yet to come try to illuminate these questions from different angles (i.e. by applying philosophical methods, institutional research, case-study approach, etc.). This introduction is structured as follows: firstly, we begin by stating the purpose of the dissertation and we argue for its relevance (Section 1). Secondly, we will briefly introduce the concept of sustainable development (Section 2), from which we will derive a more precise specification of the problem we wish to address and the resulting research questions we wish to explore (Section 3). These sections are followed by some methodological considerations (Section 4) and a brief discussion of the layout of this dissertation (Section 5).

1. Purpose and relevance As mentioned above, the overarching purpose of the present thesis is to investigate to what extent the ongoing debate on sustainable development – and in our case, the more specific debate related to the question of sustainable energy provision and the role of nuclear power therein – can be illuminated by a systematic and integrated evaluation of philosophical, theoretical and practical approaches in the field of technology assessment, governance and communication (Chapters 1-4). Insights gleaned from these evaluations will subsequently be translated into a concrete governance proposal (Chapters 5-7). The overall purpose can be further specified as: 1. to introduce a particular philosophical framework (i.e. constructivism) which constitutes a vantage point for subsequent theoretical and empirical elaborations; 2. to defend this framework against possible ‘competitors’ (i.e. Beck’s theory of risk society and social constructivism);

3

I (we) switch here from the ‘personal/confessional’ tone of the ‘I’-form to the less subject-centred ‘we’-form – a form which according to us invokes better the constructivist perspective of the researcher/analyst as a ‘mediator’ for an unspecified collective of ‘actants’ (cf. Chapter 1).

Problem framing, research questions and methodology 3

3. to analyse the use of the sustainability principle in combined scientific-political practices (i.e. a case-study approach); 4. to investigate the relevance of the theoretical and case-study research in developing guidelines for governance in the field of sustainable energy; and 5. to make a first attempt at evaluation of the practical implementation of this new governance scheme. At the root of this investigation lies, as the reader will surely notice when reading through the chapters still to come, a dissatisfaction with how the discussion on sustainable energy has been conducted so far. Compared to the more ‘substantial’ side of the debate (concerned with the ‘content’ of sustainability in the field of energy policy), i.e. the perspective offered by either the ‘hard sustainability sciences’ (e.g. environmental sciences, environmental and resource economics, industrial ecology, etc.) or by the more ‘soft’ philosophical approaches (e.g. eco-philosophy, environmental ethics, epistemological or cultural critiques of modernity, etc.), analyses on how to link theoretical philosophical perspectives to the conditions under which political planning, legal-institutional approaches and/or state initiatives operate in the field of environmental and/or technological decision making constitute a more recent (albeit quickly evolving) field of scientific inquiry. In order to bring the ‘added value’ of our investigation more sharply into focus, let us briefly point out what we believe to be the three most important research projects belonging to the latter category of inquiry. The first project focuses on changing the decision-making processes of government agencies that deal with environmental issues. Scientists working in this area have developed and tested numerous models for decisionmaking processes trying to integrate scientific analysis and community deliberation into a comprehensive strategy for decision making (see e.g. Renn et al. 1994; NRC 1996; Bergmans 2005). The general aim is to define a democratic method for the development of policy that recognises the link between social rationality (i.e. concerning practical, normative or aesthetic concerns normally left out of scientific rationality) and public involvement. While useful, we nevertheless believe this project to be limited. Because its focus is on administrative decision making and small group processes, it fails to deal with the vast power differentials that exist outside of the process. Also, it fails to consider the larger decision-making structures in society. A second project focuses on the relationship between political planning initiatives and their potential to realise a deliberative and democratic practice. Scientists working in this area analyse existing power and institutional relationships from the vantage point of theories of deliberative democracy, and try to develop a planning practice that is both sensitive to issues of power and ethics, and able to assess issues in situated political processes (see e.g. Forester 1993, 1999). Again, while we certainly acknowledge some common ground between this project and our own, our approach is much more critical of the need to base political planning in the field of environmental and/or technological questions on deliberative-democratic ideals 4 . The final

4

Chapter 2 includes a critical analysis of theories of deliberative democracy – most notably Habermas (1996).

4

Chapter 0

project attempts to deal with society-level decision-making processes by taking into consideration institutional structures and the role they play in either promoting or hindering the furtherance of the environmentalist cause. Some of the major works in this area focus on the role of economic structures in environmental degradation (e.g. Jaeger 1994), risk decision making (e.g. Beck 1992), and the public sphere and the role of social movements (e.g. Torgerson 1999). Most of the work characteristic of this project tends to be critical of the possible role of the state in mitigating environmental degradation. One of the main exceptions is Eckersley (2004), whose analysis explicitly focuses on harnessing statecentred policy mechanisms for environmental purposes. Eckersley’s general approach perhaps comes closest to our own, but is much more outward-looking in its subject coverage: it discusses the role of the state both in national and international environmental policy, and does not focus specifically on energy policy. Arguably it is not to be called sharp-edged science anymore when we call for a processoriented implementation of sustainable development. But more specific and systematic reflection and guidance (as opposed to ad hoc approaches) on how to translate a general commitment to sustainability into a workable set of practical assessment criteria, process design characteristics and implementation methods seems to be lacking 5 . Thus, we consider it to be our task to set out the practical means – i.e. the essential characteristics of deliberative decision-making processes, the necessary and useful tools, the openings for adjustments, etc. – by which the considerable challenges of energy technology assessment in the context of sustainability can be met. To put it a bit colloquially: once society has decided to change its course (which, at least in words, seems to be the case as witnessed by the widespread adoption of sustainable development as an official policy principle), it has to be decided upon not only where we want to go from here, but also how fast, through which muddy detours, and with whom. Perhaps one further clarification is warranted in order to prevent misunderstandings from arising at an early stage: our investigation of the concept of sustainable development starts from the observation that it is a well-established fact in many areas of national, regional and international law and policy. Therefore, we will not enter into a critique of the concept as such. We start from the factual acceptance of sustainability and inquire into how this concept might be interpreted and possibly contribute to an improved understanding in view of its implementation. Thus, if the reader so pleases, my arguments can be understood as conditional: if we are to apply the principle of sustainable development to technology assessment and governance, then the research findings presented in this dissertation could offer valuable suggestions for improvement. But let us first provide some necessary (albeit very limited) background on this simple fourteen-letter word which yet has caused us so much trouble…

5

For a recent similar (but more general) attempt, see Gibson et al. (2005). These authors do not however explicitly discuss the philosophical premises from which their recommendations are derived.


2. Sustainability from a bird’s-eye view

2.1. Historical and conceptual background Although the concept of ‘sustainability’ certainly has a longer history, it was not earlier than in the 1987 report of the ‘World Commission on Environment and Development’ (WCED 1987), entitled “Our Common Future”, that ‘sustainable development’ enjoyed a breakthrough in the international political development discourse 6 . This report (also colloquially called the ‘Brundtland report’, after the commission’s chairwoman Gro Harlem Brundtland) systematically investigated the link between the prevailing (industrial) model of economic development and major environmental problems. In order to tackle these problems, the Brundtland report called for institutional and economic changes both at the local and the global level. It defines sustainable development as a development that “…meets the needs of the present without compromising the ability of future generations to meet their own needs…” 7 . It contains within it two key concepts: − the concept of needs, in particular the essential needs of the poor (e.g. food, jobs, energy, water and sanitation), to which overriding priority should be given; − the idea of limitations, imposed by the state of technology and (social, political and economic) institutions on the environment’s ability to meet present and future needs. In other words, the key message of the Brundtland report was that, even though economic growth historically has led to widespread environmental degradation, this should not necessarily be the case in the future, on condition that (economic) development would be brought in line with the requirements of the new sustainability paradigm. Moreover, a reorientation of economic growth towards a more environmentally friendly path was advocated based on the premise that future growth itself might be jeopardised by further environmental degradation. Concerning the role of technology – the issue of most importance to us – it can be argued that the view presented by the Brundtland report was somewhat ambiguous. In a sense, the role of technological progress is described there as a double agent. On the one hand, technologies can induce a more intensive exploitation of resources but on the other hand technological progress can contribute to a more efficient use of resources and/or provide alternative solutions (Glasbergen and Blowers 1995). There has been a great deal of speculation as to why the Brundtland report has become such a popular reference in the international development discourse. The story of how this happened is certainly complicated and there is no universal agreement on how it should be

6Which

we will not discuss in the present dissertation, see e.g. Lélé (1991) and Pezzey (1992) for classical historical analyses. Commonly, the historical roots of the concept of sustainable development are traced back to the late 60s, when the limitations of the assimilative capacity of the environment as a ‘sink’ for wastes resulting from economic activities became apparent. This conceptualisation of environmental conservation and economic growth as competing activities culminated in the first Club of Rome report (Meadows et al. 1972), called “The Limits to Growth”. 7 See chapter 1 for a discussion on the sense and non-sense of trying to define sustainable development.

6

Chapter 0

interpreted. However, at a very ‘basic’ level, we believe part of the explanation lies in a socio-political and cultural shift in the industrialised countries: the rise of ‘postmaterialistic’ values (Inglehart 1977, 1990) in affluent societies (often expressed as a feeling of discomfort with the ethical, social, spiritual, etc. dimensions of the contemporary interpretation of progress), combined with increasing scientific evidence that the consequences of consumption patterns could be observed to affect the whole world (e.g. the hole in the ozone layer, the enhanced greenhouse gas effect), has increasingly put environmental problems high on the political agenda of ‘rich’ countries. Another part of the explanation according to us lies in the fact that the Brundtland report based itself on the premise of economic growth, which made it acceptable to both developed and developing countries. In other words, the Brundtland report did not so much want to embrace a specific ecological principle (e.g. respecting the ‘carrying capacity’ of the earth), but rather a set of socio-economic objectives such as providing access to resources and ensuring a more equitable distribution of the costs and benefits of (economic) development (while limiting the overall negative consequences of this development).

2.2. The international political agenda (from Rio 1992 to Johannesburg 2002 and beyond) The concern about the environment, equity and development was echoed at the ‘United Nations Conference on Environment and Development’ (UNCED) in Rio de Janeiro (1992). As an outcome of this conference, the principle of sustainable development was formally adopted (and pinned down in 27 principles) and the so-called ‘Agenda 21’, a comprehensive plan of action for the 21st century, was signed by delegations of 178 countries (including Belgium). This formal adoption consequently trickled down to the regional and national policy levels (e.g. at the European Union (EU) level, inclusion of the concept of sustainable development in the 1997 Amsterdam Treaty). Also, the ‘United Nations Framework Convention on Climate Change’ (UNFCCC) was signed, setting the scene for the international community’s efforts to combat anthropogenically induced climate change (mainly as a result of the combustion of fossil fuels). But although the interdependence of economy and the state of the environment was now formally acknowledged by the global community, the North and South continued to emphasise their own preoccupations. While the North (and in particular the EU) stressed the need for negotiating multi-lateral environmental treaties, the South emphasised the North’s responsibility for the current state of the environment and demanded resources (e.g. financing development, technology transfer, etc.) for the eradication of underdevelopment. In particular, developing countries remain dissatisfied by the lack of commitment from ‘donor countries’ to increase the level of official development assistance up to the promised level of 0.7% of their gross domestic product (GDP). The major question is of course whether this formal recognition of sustainability as a guiding principle for the world’s future development has had a concrete impact on the above-mentioned negative dynamics. Most commentators (see e.g. Nierynck et al. 2003; Gibson et al. 2005) would agree that, while (slow) progress has been made in certain areas,


the overall results have been rather disappointing 8 . While thousands of specific initiatives have been undertaken at all levels from the local neighbourhood to the planetary level, they have so far remained mostly counterpoints to dominant practice. In comparison with the background situation at the time of publication of the Brundtland report, it is certainly not justified to conclude that any substantial progress has been made in curbing the threefold negative tendencies of population growth not accompanied by parallel economic developments (in the South), unsustainable consumption and production patterns (in the North) and a continued increase of inequity (between North and South). Some would even claim that the situation has actually deteriorated: in the 1999 Human Development Report (UNDP 1999) for instance, the ongoing process of globalisation is criticised for engendering yet a new form of inequity – i.e. a ‘technology gap’. According to this report, the benefits of especially new technologies (e.g. information technology, biotechnology) are not equally shared. At the same time, it also states that today’s globalisation, while opening opportunities to combat poverty, is still primarily driven by market expansion “…outpacing governance of these markets and their repercussions for people…” (UNDP 1999, p. 2). In this regard, it can be seen as a writing on the wall that in the most recent sustainability high mass, the ‘World Summit on Sustainable Development’ (WSSD) (Johannesburg, 26 Aug. – 4 Sept. 2002), a simple reaffirmation of the Rio acquis was already seen as a success. In contrast to the Rio conference ten years earlier, which benefited from a burst of optimism related to the fall of the Berlin wall, the WSSD took place in a glum international climate, dominated by the tensions created by the terrorist attacks of 11 September 2001 against the United States (US), the world economic crisis, and the then lagging ratification of the Kyoto protocol. Also, the ‘disappointing’ results of the WSSD can be ascribed to an obvious divergence in the US and EU international political agenda. Sustainability is no objective in any official US policy whereas the EU have made it a key to their institutional activities, both at the level of international treaties and in supplying national and EU strategies for sustainable development. Also the US and the EU have different opinions on the value and applicability of the precautionary principle (one of sustainability’s subsidiary principles), with the US putting much more emphasis on the use of ‘sound science’ for decision making (Dratwa 2002) 9 . Furthermore, at the WSSD the EU was practically left as the only political block which emphasised the need for an action-oriented outcome with clear and measurable objectives. In the final political Johannesburg declaration, some targets were included for some areas, but not for energy 10 . On energy, an agreement was reached to take joint actions to improve access of the poor to energy, and an agreement to

8

E.g. the Kyoto protocol, setting national emission reduction targets for greenhouse gases for industrialised countries, entered into force in 2005. The US, one of the major emitters of greenhouse gases, however has not ratified the Kyoto protocol. 9 Thus implicitly acknowledging the danger that a precautionary approach would go against the fundamentals of sound science. For a rebuttal of this argument, we refer the reader to chapter 5. 10 Concerning access to basic sanitation (halving the proportion of people lacking such access by 2015), minimisation of health and environmental impacts from the production and use of chemicals (by 2020), the restoration of fish stocks to sustainable levels (by 2015), a renewed commitment to national strategies on sustainable development (by 2005), and halting the loss of biodiversity (by 2010).

8

Chapter 0

increase ‘urgently’ and ‘substantially’ the global share of renewable energy sources was concluded. Both agreements remained without concrete targets however. The EU (with a group of like-minded non-European countries) continues to lead a political alliance for renewable energy. EU countries have shown a willingness to establish and fulfil quantified indicative targets within a given timeframe at both the national and regional levels (Directive 2001/77/EC). For instance, the indicative target for Belgium is to increase the share of renewable energy sources for electricity production to 6% by 2010. The EU remains divided on the issue of nuclear power: on the international level, it did not promote nuclear power as a ‘sustainable’ technology under the UNFCCC; however, on the regional level, a ‘green paper’ issued by the European Commission (EC 2001c) has pointed out the dangers of abandoning the nuclear option in light of energy security goals (cf. Chapter 4). Furthermore, the decision to invest in or close down nuclear power plants falls under national sovereignty (provided of course that applicable international treaties and jurisdiction are respected). Finally, it is perhaps worth mentioning that the ‘United Nations Commission on Sustainable Development’ (UNCSD) is devoting its 2006-2007 working programme to the issue of sustainable energy.

3. Problem delineation and research questions In view of the broad range of subjects and problematic situations covered by the sustainability discourse, it will hardly come as a surprise that it is nearly impossible to produce anything even remotely resembling a ‘state-of-the-art’ in the ‘theory of sustainability’ (as is common practice in an introduction to a dissertation). In the years following publication of the Brundtland report, sustainability debates proceeded vigorously both in words and deeds. Acres of literature were produced ranging from theoreticalspeculative outlooks to practically-oriented guidelines, covering almost every aspect of interactions between human activities and the environment. To add to the problem, not all sustainability-relevant literature necessarily bears the sustainability label explicitly; and conversely, for opportunistic reasons research or policy initiatives and projects were often relabelled and discussed as exercises in ‘applied sustainability’ (Zaccaï 2002). Despite this impressive amount of energy spent in trying to make sense of sustainability, we seem to be little further down the road of reaching a consensus. Fundamentally of course, this is because the debates about the meanings and implications associated with the concept imply (to some extent) a competition among interests and value positions seeking to defend or promote established or new priorities and understandings. For the purposes of our research project, it does not matter whether we endorse a particular perspective on the future orientation of society. As we will argue in chapter 1, the key lesson is that new perspectives constantly claim an ‘entry right’ into sustainability debates ahead of the legislative process, and this means that sustainable development will remain a contested question that will continue to engage firmly with established cultural, habitual or political practices in reaching judgments about the environment. In this process, an academic researcher simply cannot have the final say.


Neither can we ‘transcend’ the debates (in an abstract, universal sense) by proposing some kind of definitive ‘taxonomy’ to help the reader sort through the many competing conceptions of sustainability. Sustainability-related literature counts many examples of such attempts. For instance, one early typology that has enjoyed continued popularity is the ‘sustainability spectrum’ proposed by Pearce and Turner (1990). It identifies four ‘basic’ positions in the sustainability debate, following a ‘worldview continuum’ ranging from ‘technocentric’ to ‘ecocentric’. This depiction suggests that the key difference is one between underlying ‘ethical’ stances. However, a quite similar taxonomy is often accepted, labelling the positions along a ‘weak’ to ‘strong’ sustainability spectrum, suggesting that the core distinction is more ‘managerial’ than ‘ethical’ 11 . The debate then (suddenly) turns to whether or not we should have much confidence that economic capital and technological innovation will be able to provide substitutes for the services provided by nature. Perhaps the most familiar taxonomy (or rather, family of taxonomies) is the one that adopts an architectural metaphor, suggesting that sustainability rests on a number of interconnected ‘pillars’. There have been lively debates on the number of pillars, their mode of interconnection, and their relative importance; with differences of opinion reflecting contrasting preoccupations. For instance, approaches have been proposed that envisage an integral view on sustainable development in the manner of a ‘Russian doll’ model, in which ‘the environment’ successively encapsulates ‘human activities’ (Guijt et al. 2001), ‘society’ and the ‘economy’ (Munda 1997) or ‘society’ and the ‘economy’↔’polity’ couple (sic) (Deblonde 2001). Other approaches depict the ‘pillars’ as intersecting circles, thus suggesting a room for ‘win-win’-type solutions and a need for ‘balancing’, since none of the pillars taken apart enjoys an obvious predominance over the others. Such approaches of course are always contestable because they inherently propose different emphases and values; furthermore, accepting one or another ‘pillar architecture’ from the outset only serves to conceal the processes contributing to the production and recognition of sustainability problems, and to their solutions. Furthermore, any approach that puts concepts in separate categories – as if the essence of (political) decision making comes down to making a choice from a menu as in a restaurant – tends to obscure what is overlapping and shared and forecloses opportunities for innovative solutions. The degree to which such ‘transcendent’ models can be proposed is further eroded by the degree to which the relative strengths of social, economic, institutional, environmental (etc.) claims are influenced by contextual factors (e.g. geographical locations, levels of policy making, etc.). Sustainability assessment and decision making involves countless choices – e.g. whether some anticipated effect is deemed to be ‘significant’, whether an option should be included in the assessment, whether further research should be encouraged in a certain field, etc. – which simply disappear from sight in most taxonomies. As we will argue more rigorously in chapter 1, this observation suggests that, at best, consensus can be reached about broad directions of change, and that more attention should go towards conceptualising a view of sustainability as a process of change instead of specifying ‘end products’ (in social, economic, cultural, or environmental terms). This 11

The so-called ‘three-capital-model’ discussed in chapter 1 (Section 3.5).

10

Chapter 0

shift of attention impinges directly on the values we wish to uphold and the institutional structures and practices required to secure them. At issue here are questions of collective organisation and administration, or, more properly, issues of governance. Perhaps these questions are not obvious to raise in a democratic society, but in our view (and as our analysis in the pages to come will hopefully make clear), if we are to move in the direction of more sustainability, we cannot take for granted that existing decision-making structures place sustainability concerns at their centre. It is for this reason that we propose to position governance – i.e. the formal and informal structures and practices through which the collective is ‘produced’ – at the centre of attention. Accepting the principle of sustainable development as a starting point for reflection (which can only be substantiated further by a constant dialogue and deliberation in practical cases) also implies accepting a correspondingly ‘open’ problem framing. The overarching research question guiding our research therefore can be formulated as follows: ‘Can nuclear energy contribute to a development process leading to sustainability – if so, under which conditions; and how can this question be answered?’. From this overarching question the following subquestions can be derived: • ‘Which meanings are attributed to sustainability in general and in particular in the context of energy policy?’; • ‘Which scientific approaches and theories have been advocated and to what extent do they address the full scope of questions raised by demands for more sustainability, and in particular those concerning the institutional dimension of public policy making? What are the strong and weak points of these approaches; which lessons can be learnt?’; • ‘What then is an ‘appropriate’ scientific/political methodology or procedure to address sustainability questions?’. Such broad subject coverage certainly necessitates some a sound methodological approach. Before going deeper into these methodological considerations in section 4, we should first outline the focus and limitations of our PhD research. With regard to subject coverage, we developed our reflections based on the practical case of the ongoing (nuclear) energy debate in Belgium. But, as will become clear when reading through this dissertation, nuclear power issues cannot be seen in isolation from general energy issues, which, in turn, touch upon most of the questions raised by the more general sustainability debate 12 . Therefore, the question of (nuclear) energy can to some extent be seen as a convenient point of entry towards a larger set of questions raised by the demands for sustainability. With regard to scientific disciplines we applied in our research, focussing on the governance dimension necessarily implies adopting an interdisciplinary perspective. Nevertheless, since one cannot be a specialist in every single discipline, choices had to be made. These generally emerged in a rather ‘organic’ way from the questions raised in the course of our research project. For illustrative purposes, we will use yet another oft-cited 12

The following issues immediately come to mind: the depletion of non-renewable resources (fossil fuels and uranium), problems of local pollution (e.g. particles of fine dust), steadily growing problems of cumulative (both regional and global) pollution (e.g. acid rain, ozone depletion, the enhanced greenhouse gas effect), wide and growing inequalities between rich en poor nations, the key importance of adequate energy provision – including security of supply concerns – in national economies, etc.. For an overview, see Elliott (2003).


taxonomy, proposed by Pezzoli (1997). Pezzoli identifies ten clusters of literature that emphasise either managerial, technical or philosophical/political solutions to sustainability problems and challenges. These clusters are reproduced in Table 1; indicated in bold are the disciplinary categories we believe we have contributed to most in the present dissertation. Of course, we are not claiming here that the overall and case-specific unsustainability (if any) of prevailing current practices in the energy field will not be corrected by merely imposing an improved governance model aimed primarily at the level of national (federal) policy making. Decisions are made at many different levels and in many different contexts; therefore, a turnaround to sustainability will also require a host of other initiatives, e.g. to develop appropriate incentives; to foster greater adherence to the concept in everyday behaviour; to link national policy making with international, regional and local initiatives; etc 13 . And the energy debate could undoubtedly be enlightened further by investigating some of the remaining non-bold-faced categories in Table 1. Nevertheless, we believe an important role could be played by the accumulation of decisions on programmes, plans and policies at the state level. Therefore, we believe it to be a sensible option to begin with some clarification of the ‘contribution to sustainability’ test for these decisions.

13

We are referring here to Beck’s (1992) notion of ‘sub-politics’ (further developed in chapter 1).

12

Chapter 0

Cluster Applied perspective with managerial focus

Category 1. Managerialism, policy and planning Legal-institutional terrain and state initiatives Civil society and NGO’s Urban and regional planning and development Natural resources and rural development Indicators of sustainable development 2. Social conditions Population Human behaviour and social learning Environmental health 3. Environmental law Property and development laws Legal issues concerning environmental racism, equity and justice

Technical perspective: the hard sciences of sustainability

4. Environmental sciences 5. Eco-design and the built environment 6. Ecological economics Environmental and resource economics Eco-tourism Industrial ecology

Philosophical or ‘structural-transformative’ perspective

7. Eco-philosophy, environmental values and ethics Epistemology, science, culture and language Philosophy, policy and development Environmental justice and racism Eco-feminism 8. Environmental history and human geography/ecology 9. Utopianism, anarchism and bioregionalism 10. Political ecology Globalisation and eco-politics Urban and regional development Rural studies Critical social movements and empowerments Theory building and agendas for research and action

Table 1. Pezzoli’s ten categories of literature on sustainable development (Source: Pezzoli 1997) (indicated in bold are the categories further developed in this dissertation)


4. Methodological considerations The questions raised in the previous sections are answered in three different (but nonetheless intermingling) research strands. These strands have been developed and further refined in the course of the ongoing research activities. Therefore, the ‘linear’ account of the research given here – first the (meta-)theoretical considerations, then case studies and finally practically-oriented recommendations – should be seen as a reconstructive account of a less straightforward ‘weaving’ movement of constant adjustment and refinement (with regard to research questions, methodology, conceptual framework and data analysis). In a first strand, we make a contribution to the ongoing debate between different (meta-) theoretical philosophical perspectives on the relationship between (technological) development (a project underpinned by deeply-rooted assumptions about the nature of modernity) and the environmental challenge. We adopt a constructivist stance, enabling us to conceive of sustainable development in terms of a never-ending collective learning experience, based on ‘agonistic’ interactions between different perspectives on sustainability 14 . In short, constructivism builds on a critique of positivism, or more generally, any kind of supposedly ‘value-neutral’ theorising (e.g. rational choice theories of decision making). Claims that there is an ‘objective-reality-out-there’ are interpreted as always and unavoidably political (in the sense of expressing a choice on the ‘right’ ordering of a collectivity), contingent (to some extent) and filtered through different frames. Constructivism (as we have developed it here) broadly falls into what we might call the ‘critical-interpretative’ perspective in the social sciences (hence the title of this dissertation). It shares with critical theories a concern for questioning the values and norms that are internal to existing understandings and practices; exposing unfulfilled emancipatory promises and opportunities (by extending the project of emancipation towards ‘non-human others’); and exploring what possible changes in thought and practice might be permitted, facilitated or enhanced in terms of further emancipation (Eckersley 2004) 15 . With the interpretative (or hermeneutic) social sciences constructivism shares a concern for understanding how meanings are created and attributed to certain events or concepts, though its empirical focus on combined socio-material practices is somewhat different from hermeneutics usual concern with texts (e.g. narratives, discourses, etc.). Ultimately the vantage point offered by constructivism, when applied to governance mechanisms at the state-level (our principal remit in this dissertation), enables us to locate the (diffuse) demand for more sustainability in the context of ‘communicative justice’ 16 as a regulative ideal – i.e. the creation of a deliberative space in which decisions regarding

14 The key terms in this sentence – i.e. ‘constructivism’, ‘positivism’ and ‘agonistic learning’ – will be explained in detail in chapters 1 (constructivism/positivism), 2 and 5 (agonistic learning). 15 Though constructivism fundamentally differs from critical theory in its critique of critique as the ultimate ground for judgment (cf. Chapter 1). 16 We put the term here between quotation marks to indicate that we are using it in a different sense than it was originally intended by Habermas (cf. Chapter 2).

14

Chapter 0

sustainability take place in ways that are acceptable to all ‘differently situated others’ which stand to be affected by the decision. The second research strand consists of a number of case studies, exploring how nuclear energy is positioned in practice in the sustainable energy debate 17 . We felt that a casestudy approach would be greatly helpful in raising awareness about how decisions are arrived at in practice (instead of staying in some abstract (meta-)theoretical realm). Empirical data were collected on three levels. Firstly, on the level of ‘science for policy making’, we analysed a large-scale European research programme attempting to reveal the ‘true costs’ of different energy options. Secondly, on the level of national energy policy, we made a reconstruction of the policy development process leading to the Belgian decision to phase out nuclear power (framed as a move towards sustainable development). Thirdly, on the level of the societal debate on (sustainable) energy policy in Belgium, we analysed the positions taken by some of the most important stakeholders in the debate. In this case, it was instructive to analyse if and how the advent of sustainable development as a guiding idea for future developments could produce any significant changes in a historically polarised debate (Laes et al. 2004c). Taken together, each of these cases is meant to enlighten questions of decision making from a different angle. However, we are certainly not claiming any form of completeness here; in particular, we do not discuss more ‘elusive’ (but no less influential) forms of decision-making strategies such as lobbying, the relationships between administration and ministerial delegates, the use of rhetorics, etc. Finally, in the third research strand we try to combine the insights gleaned from both theoretical and empirical investigations into a practical proposal for a new governance scheme for sustainable energy set in the institutional context of Belgian energy policy. Here, we will look for emancipatory (in the enlarged sense of the word) potential available in contemporary processes or developments (e.g. the growing importance of the precautionary principle in national and international law) and suggest how these might be goaded and sharpened in ways that might bring about structural political changes aimed towards making energy governance more responsive to demands for more sustainability. This final research strand is of course the most speculative of the three, and is therefore also less easily described in terms of a methodological approach. Perhaps Eckersley’s (2004, p. 4) description of ‘disciplined imagination’ – i.e. drawing out a normative vision that has some points of engagements with emerging practices and understandings – comes closest to the perspective adopted in this final research strand.

5. Structure In the eight chapters that follow, we will reconstruct our search for answers to the questions we set out for ourselves. In the first chapter we present the results of a literature search on

17

Questions pertaining case-study methodology are discussed in Annex 1.


sustainable development. This result will take the form of a (meta-)theoretical perspective on sustainable development, i.e. a perspective which explains why different authors can defend such widely diverging interpretations of the concept. In particular, we will use Boltanski & Thévenot’s ideas on justification, Beck’s theory of ‘risk society’, and (social) constructivist theories of science and technology to arrive at a (meta-)theoretical position which incorporates at the same time pragmatic, empirical and moral dimensions. This first chapter deals with sustainable development on an admittedly rather abstract level. Nevertheless, we deem such analysis to be indispensable to our overall undertaking, since the conceptual deep structures in a piece of theorising determine, among other things, its specific validity claims and, correspondingly, the distribution of covert and overt burdens of proof. For example, Beck’s theory of ‘risk society’ (discussed in Chapter 1) certainly is an original and persuasive device for throwing a new light on some recent societal, scientific, and political evolutions in a cogent framework. But if it is more than simply an expository device – and surely, as we show in chapter 1, this is the case – one cannot avoid asking what are the specific more normative and practically-oriented claims made in this theory. Abstract and formal as such questions may appear at first sight, they need to be answered if we are ever to arrive at an assessment of the strength and practical innovative character of a concept such as sustainable development. In chapter 2, we start our search for a sustainable energy governance scheme. This chapter builds on the (meta-)theoretical perspective developed in chapter 1, and shows how we can derive from literature four ideal-typical governance schemes. We will show the internal connection between a particular way of conceiving sustainable development, of framing policy problems and of devising solutions; thus revealing the necessary ‘construction’ or ‘repair’ work that has to be performed for keeping a policy issue within the boundaries of one governance scheme, or moving from one scheme to another. The next two chapters (chapters 3 and 4) leave the path of theoretical musing in order to consider more practical cases of energy technology assessment, decision support and decision making. They are meant to tell ‘stories’ of how decisions live and die in ‘real’ circumstances. Following Law and Mol (1995), we conceive of constructivism as a strategy for drawing together these distinct stories to form a ‘patchwork’ of sometimes overlapping or sometimes juxtaposed practices, discourses and organising principles. In chapter 3, a first combined political-scientific practice (aimed at providing support for sustainable energy policy) is put to the test. The largest part of this chapter will be devoted to a detailed and critical examination of the ExternE methodology, a large-scale research effort funded by the EC designed to evaluate the external (or social) costs of different energy carriers. This allows us to show how and which political and/or moral conceptions inevitably enter into the work performed by scientists engaged in the project, enabling us to reflect on the use and limitations of such methods as a support for (sustainable) energy policy. In contrast to chapter 3, chapter 4 examines the use of science in a political setting. In this chapter, we will analyse how the Belgian government has justified its decision to phase

16

Chapter 0

out nuclear power (as a strategy for achieving sustainability in the energy sector). We will focus in particular on the role of scientific expertise (i.e. the AMPERE report) and the way expertise is brought into decision making. Lastly, we will present a summary of the interviews we conducted with members of the Belgian ‘federal council for sustainable development’ (FRDO), in order to map the present-day controversy on nuclear power and its role in a sustainable energy system (in the Belgian context), and analyse how different perspectives relate to ‘official’ decision making. Based on the critical investigations and evaluations in previous chapters, we will propose in chapter 5 an ‘improved’ governance scheme for sustainable energy development, aimed at providing guidance on dealing with ‘surprises’, recognising a multitude of perspectives, upholding the value of ‘sound science’, and setting out the practical means and tools (in terms of institutional arrangements, scientific ‘tools’, etc.) for arriving at ‘productive’ decisions. The final chapters of this dissertation are devoted to a preliminary assessment of our ‘improved’ governance scheme. In the context of this dissertation, this assessment of course could not be carried out in full (only so much can be set out in general rules and procedures; ‘reality’ is always the ultimate test) and was conceived as a limited ‘pilot exercise’, focussing on only some of the aspects of the policy development cycle. The pilot exercise aims to give input towards the process of agenda-building through a better selection and structuring of the policy problem posed by the demand for a more sustainable energy system. This was done with the aid of scenario exercises (Chapter 6) and a multicriteria mapping exercise, in cooperation with members of the FRDO (Chapter 7). We end with a summary of the main findings, recommendations for policy making and suggestions for further research (Chapter 8). Of course, we are not claiming that we have said the final word on these issues. Claims to ‘final solutions’ are always somewhat pretentious (if not dangerous), but especially in the case of sustainability they are outrightly preposterous. Paradoxically, the concept of sustainability is at the same time very simple to understand (hence, its inherent appeal to an ‘everyday’ or ‘manifest’ understanding) and hyper-complex (cf. Chapter 1). It must deal at once with the obvious and vast areas of uncertainty or even mystery. Nevertheless, assessments must be made and decisions must be taken. No doubt (or hopefully?) we will learn a great deal from further experience and more careful analysis and deliberation. Perhaps the initial steps taken here on the road towards sustainability assessment and governance in the field of energy policy will soon seem hopelessly primitive. I can live with that: this strange mix of pleasure and frustration seems to be my predicament…

CHAPTER 1 THE PRINCIPLE(S) OF SUSTAINABLE DEVELOPMENT: TOWARDS A (META-)THEORETICAL FRAMEWORK In the first chapter of this dissertation, we will submit the concept of sustainable development to a (meta-)theoretical questioning18. After a brief introductory section (Section 1), we argue that appeals to the principle(s) of sustainable development generally assume one of the following three forms: an appeal to a manifest image, a (political) vision or a policy target (Section 2.1). As a consequence, sustainable development needs to be critically questioned on each of these levels. We argue that it is impossible and furthermore unnecessary to try to capture the meaning of sustainable development in one integrated definition (of the kind of “sustainable development is…”, followed by one or more sentences laying down its definitive meaning) (Section 2.2). Next, in order to delimit our scope somewhat (in view of our particular subject matter – i.e. decision making in complex technological questions), we derive a list of characteristics of sustainable development. This list is derived from the different appeals to sustainability and from existing attempts at definition. At the same time we take care not to exclude possible interpretations on a prioristic grounds (Section 2.3). At first, we simply lay out this list of characteristics without making any attempt at systematisation, i.e. without arranging them in dimensions, outlining identities and differences, constructing hierarchies, etc. This systematisation will be introduced afterwards, when we review a number of different (meta-)theoretical outlooks. Each of them conceptualises the interplay between science, technology and society in distinct ways, thus proposing different answers to the fundamental questions concerning the vantage point from which such conceptualisation can take place, and concerning the role of the analyst in this process (cf. Chapter 0). In particular we look for ‘blind spots’, gaps or dichotomous oppositions within each theory which might impose a certain interpretation of sustainability. In section 3.1, we first propose Boltanski and Thévenot’s theoretical work on justification (the ‘commonwealth model’) as a candidate (meta-)theoretical model for our purposes. Boltanski and Thévenot submit that argumentation in view of reaching a justified agreement follows certain ‘grammatical’ rules. These rules apply to different types of argumentation – be it technical, market-based, political, etc. – and thus at first sight seem to offer an interesting prospect for analysing appeals to sustainability. At the same time, this model throws a new light on the inherent weaknesses of sustainable development when serving as a legitimacy principle. 18

The difference between a ‘meta-theory’ and a ‘theory’ is not absolute but rather gradual and ultimately vague: every ‘theory’ depends on the simultaneous use of different conceptual schemes, including an (implicit) reasoning as to why reference to the ‘theory’ is justified in the particular circumstances under consideration (hence, a ‘meta-theory’). Thus, it depends on the particular context in our text whether we are ‘explicitly’ evaluating other theories (in this case, our perspective becomes ‘meta-theoretical’) or whether we are explaining or reconstructing a certain ‘phenomenon’ (in which case our perspective becomes ‘theoretical’). However, such explicit separations will be used with exception in the main body of the present chapter, since they are not necessary for a better understanding. Therefore, we have opted to name our own perspective ‘(meta-) theoretical’. All of this will become clearer in section 3.4.

18

Chapter 1

Boltanski and Thévenot’s theoretical framework is then confronted with critical analyses of other theoretical outlooks, which either take issue with the ‘commonwealth model’ and its focus on justification at the factual/descriptive or at the normative level, regarding the desirable attitude of the policy analyst. In particular, we analyse Beck’s theory of ‘risk society’ (Section 3.2) and (social-)constructivist perspectives on science, society and technology (Section 3.3). We conclude this discussion with a (meta-)theoretical outline for our work, thereby clarifying our own (inseparably moral, empirical as well as pragmatic) position as a policy analyst (Section 3.4), and we explore the possibilities and limitations of sustainability when serving as a principle for guiding action (Section 3.5). We conclude with summarising our view on sustainability, which we describe as a ‘legitimacy compromise in the making’ (Section 4)19. The key significance of this chapter therefore lies in the fact that it provides the intellectual background of concepts, orientations and generalisations that shape policy for sustainable development. The aim is to develop a mental ‘roadmap’, which will aid us to find our way in the ‘landscape’ of sustainability debates which will be highlighted in the chapters to come.

1 Introduction In an age supposedly marked by an absence of ‘big stories’ (variably dubbed the ‘postindustrial’, ‘post-modern’ (and so on) age), sustainable development nonetheless seems to be one of the last pretenders to that title. Since the end of the ‘80s, following the publication of the ‘Brundtland report’ (WCED 1987), sustainability has rapidly risen in prominence as a central theme in the debate on the shaping of global governance structures, introducing global environmental health and justice alongside (and connected with) other guiding principles such as democracy, equality, social justice and (economic) welfare20. Following the ‘Rio Conference’ (1992), and the subsequent publication of ‘Agenda 21’ (UNCED 1992) – a rather ambitious implementation programme that explicitly calls upon states to elaborate national strategies, plans, policies and processes to make development more sustainable – reference to sustainable development really became unavoidable on an international level21. Sustainability has become generally accepted as a planning goal in European environmental policy (see e.g. EC 2001a), and was enshrined in the treaty

19

Concurring with the position taken by Godard (2003). Pallemaerts (1995) rightly notes instances where the concept of sustainable development is used before the publication of the Brundtland report itself. For instance, in as early as 1975 a decision of the governing council of the ‘United Nations Environmental Programme’ (UNEP) stated that “…Environmental management implies sustainable development of all countries, aimed at meeting basic human needs without transgressing the outer limits set to man’s endeavour by the biosphere…” (quoted in Pallemaerts 1995, p. 383). It is however clearly as a result of the Brundtland report that the concept has enjoyed a breakthrough in the political (and later juridical) agenda and the wider ‘popularised’ discourse on global environmental problems. 21 The concept figured in most major UN conferences (concerning food, health, population, social development, etc.), and a special body was created (the ‘United Nations Commission on Sustainable Development’ (UNCSD)) in order to follow up progress on implementation. Other international organisations, such as the World Bank, the ‘International Monetary Fund’ (IMF), the ‘World Trade Organisation’ (WTO), the ‘Organisation for Economic Cooperation and Development’ (OECD) and the ‘International Labour Organisation’ (ILO) regularly organise forums where sustainable development is discussed, or publish action programmes, declarations, etc. 20

The principle(s) of sustainable development: Towards a (meta-) theoretical framework

19

establishing the EU22. Belgium, as a signatory to these international agreements, institutionalised the concept in the 1997 law on the coordination of the federal policy on sustainable development (published in the Belgian Official Journal (BS) on 18 June 1997). This law inter alia foresees a two-yearly publication of a federal report on sustainable development by the federal planning bureau’s (FPB) task force on sustainable development and a four-yearly publication of a ‘federal plan on sustainable development’23. Nevertheless, despite these institutional landmarks, many questions remain regarding the correct interpretation and practical implications for policy making24. This situation opens up the way for harsh criticism and often sarcastic disdain from numerous observers and policy makers alike. Broadly speaking, two opposed camps have seized the opportunity to launch a frontal attack on the concept of sustainability. One camp fears that the sustainability approach is too ambiguous and impractical to serve as a base for sound (‘scientific’) decision making; and, if nonetheless applied on ‘ideological’ grounds, threatens to paralyse economic growth and technological (and even human) progress25. The other camp sees in the essentially vague formulation of the concept a strategic means to cover up inherent contradictions between the prevailing model of world economic development and ‘true’ sustainability, thus relegating sustainable development to the status of an ‘empty box’ at best, or an ‘ethical wrapping’ for otherwise destructive global development patterns at worst26. According to this view, such political newspeak is made possible only by a fundamental

22

Treaty on European Union (signed in Maastricht on 7 Feb. 1992, published in the Official Journal of the European Communities (C-191) on 29 June 1997). Art. 2 of the treaty reads: “…The Community shall have as its task, by establishing a common market and an economic and monetary union and by implementing the common policies or activities referred to in Articles 3 and 3a, to promote throughout the Community a harmonious and balanced development of economic activities, sustainable and non-inflationary growth respecting the environment,…”. Notice however the specific placement of the word ‘sustainable’ next to ‘noninflationary growth’, with ‘respect for the environment’ as an additional requirement; most commentators would perceive this as a pleonasm, since sustainability automatically implies a ‘respect for the environment’. 23 Based on a draft plan submitted by the ‘interdepartmental commission on sustainable development’ (ICDO), subject to approval by the council of ministers after public consultation and consultation of the FRDO. 24 For the Belgian situation, see in particular Berloznik et al. (1996), Mormont (2002), and Niestroy’s (2005) analysis of the Belgian strategy for sustainable development in comparison with nine other European countries. 25 Consider for instance the position taken by the US representatives during the climate change negotiations that the emission reduction targets agreed upon in the Kyoto Protocol (1997) would be ‘bad for the American economy’. The Republican-dominated US Congress (2000-2004) remained opposed to ratifying the protocol before major developing countries, like China and India, would ‘meaningfully participate’ in efforts to cut greenhouse gas. Besides business representatives or politicians, scientists and philosophers have also joined the fray, some of them receiving ample attention in the media. To give but two examples: the Danish statistician Björn Lomberg has acquired some fame for his denunciation of the ‘green litany of catastrophes’, which, according to him, the environmental movement is constantly proclaiming for mostly political motives. Lomberg argues on economic grounds that the resources spent on combating climate change could and should be used more efficiently in alleviating direct needs (health, education, water and sanitation) of populations in Third World countries (Lomberg 2001). The French philosopher Luc Ferry (particularly in Ferry 1992) has even accused (radical) ecologists of anti-democratic and anti-humanistic sentiments – thereby receiving his share of media attention in (for the largest part) the French press. For other examples of dreaded infringements of environmental policy on the principle of liberal neutrality, see Coglianese (1998). 26 Concerns about the fruitfulness and the utility of the concept of sustainable development have for instance been aptly summarised by Cohen et al. (1998, pp. 352-354) who suggest that the concept is “…vague, attractive to hypocrites and conducive to fostering illusions…”.

20

Chapter 1

disparity in (political or economic) power between the parties involved27. This raises the question whether we should resign to O’Riordan’s (1988, pp. 39, 48) early pessimistic assessment that ... Sustainability is not regarded seriously by those who really count, namely those at the top of political structures and those who control the flows of national and international capital. (...) One can only be cautious about an effective future role for the concept of sustainability. It is probably going to languish as a ‘good idea’ which cannot be put into practice – like ‘democracy’ and ‘accountability’...?

Criticisms of this kind indeed cast serious doubt on the feasibility of any scientific attempt (including ours) at evaluating and providing guidance on energy policy within a sustainability framework. If we cannot agree on the very meaning or scope of a concept such as ‘sustainable energy’, how can we ever assess progress and/or guide policy towards that goal? In the following sections, we argue that this feeling of dejection is not justified a priori28. In an attempt to explain or even transcend the apparent impasse, we will defend the position that – even considering the essentially intersubjective and political character of the notion of sustainable development (and there is no point in denying this) and the problems this causes in representing sustainable development as a legitimacy principle29 – it is still possible to refer to sustainable development as a ‘legitimacy compromise in the making’(in Boltanski and Thévenot’s terminology – cf. Section 3.1) for policy making, which only acquires its specific meanings and justifications in specific policy situations and contexts. With Fischer (1980, pp. 111-115) and Van de Graaf and Hoppe (1996, pp. 66-74), we consider that policy analysis can involve argumentation on two levels30: First order judgments: • Technical verification: concerned with empirically checking whether certain given policy goals are really attained by the policy measures which were introduced in order to solve or alleviate a problematic situation (‘do we get what was promised?’); 27 This point of view is for instance made abundantly clear in the “Johannesburg Memo” (Sachs 2002). This critical memorandum was published just before the Johannesburg World Summit on Sustainable Development (a follow-up conference ten years after the ‘Rio conference’, therefore also called ‘Rio+10’, 26 Aug. – 4 Sept. 2002) by a number of critical development thinkers originating from the academic world, NGO’s, politics and business. 28 However, this position does not deny the danger of the appropriation of the concept by powerful elites, bureaucracies in search of new legitimacy, firms in search of marketing arguments, etc. 29 In order to found a new and fair ‘commonwealth’ in Boltanski and Thévenot’s terminology (cf. Section 3.1). 30 Other authors (e.g. Grin 1997, Jacobs 1999) have introduced similar (that is, on a formal level) two-level structures of policy analysis. Jacobs (1999, pp. 25-26) for instance notes that the concept of sustainable development operates at the same time on a top level of ‘general intuitive comprehension’ (where, according to him, a general political consensus is likely to exist – ‘nobody can be against democracy, freedom and sustainable development’) and at the same time on a lower level of ‘technical verification’ (where, according to him, the practical interpretation and implementation of sustainability is most likely to reveal tensions among the different policy views). Grin (1996, pp. 38-39) talks of ‘first order notions’ (opinions about the effectiveness of policy solutions, problem definitions in a given policy context) and ‘second order notions’ (empirical and normative background theories, final value preferences). We will return to this issue and defend our choice for Fischer’s scheme in Chapter 2, where we will outline different ideal-typical governance theories for sustainable development.


21

• Situational justification: concerned with justifying or criticising the stated policy goals, given the application context of the proposed policy measures (‘what can we propose as legitimate goals for policy, given the circumstances?’); Second order judgments: • Institutional support31: concerned with justifying or criticising the stated policy goals, keeping in mind the contribution of these goals to the maintenance of the established order (‘what can we propose as legitimate policy goals, given a certain way of life, (political) culture, dominant ideology, social order, etc.?’); • Reasonable choice32: concerned with justifying or criticising policy goals or political institutions in the light of a certain value system (possibly independent of prevailing societal values) (‘which values do we want to embody in our political community, and which institutions are needed to realise this political identity?’)33. In this dissertation, we embrace an encompassing policy-analytical view of sustainable development oriented towards empirical observation. This means that, rather than advancing our own substantive normative interpretation of sustainability (at the ‘reasonable choice’ level), we want to clarify how this concept ‘works’ in practice – in the context of the debate on nuclear energy as part of a wider strategy for sustainable (energy) development in Belgium – by showing its multiple attachments to ethical theories, scientific knowledge, existing power structures, situational justifications, etc.; and to reveal possible tensions, omissions or problems on these different levels, thereby indicating a potential for improvement. Both the substantive (concerning the nature of the knowledge and arguments used in the debate on sustainability and nuclear energy) and the procedural (concerning the processes leading to a decision) sides of the policy-making equation draw our attention. The present chapter however is mainly concerned with laying out the (meta-) theoretical foundations of our work. This (meta-)theoretical stance is necessary, precisely because we want to explain how different actors, possibly drawing upon different theoretical insights, justify policy choices in light of different ultimate principles. At the same time, we are concerned with the question of the role of the policy analyst when

31

Van de Graaf and Hoppe use the term ‘system support’ to denote this level of policy analysis. We however prefer the term ‘institutional support’ instead of referring to a ‘system’ – a concept with overt overtones of determinism, order, regularity, etc. With ‘institution’, we roughly indicate a rule or set of rules guiding people’s actions and the accompanying views that provide the rule with meaning and with a relevant context of application. This very broad view of institutions as generalised rules of coordination concurs with Deblonde (2001, pp. 48-56) who describes them as both enabling and restricting historical entities, often being the unintended result of intentional interaction – hence the impossibility of ‘locking up’ institutional rationality within the boundaries of system rationality. We will expand on this further in this chapter (Section 3.4). 32 Here too we deviate slightly from Van de Graaf and Hoppe, since they use the term ‘rational choice’ to denote this level. We prefer to use a less restrictive specification of ‘rationality’, reserving its application not only to the domain of justification in the light of ultimate values, but also e.g. to reasoning about means-ends relationships, strategic reasoning, etc. Of course, the same can be said of a restricted use of ‘reason’ or ‘reasonable’, but nevertheless ‘reasonable’ carries stronger connotations of fairness and/or righteousness than ‘rational’ (synonyms of ‘reasonable’ in a dictionary are ‘rational’, but also ‘fair’, ‘moderate’, etc.; whereas ‘rational’ only means ‘well-thought out’, ‘well-considered’, ‘logical’, ‘systematic’, etc.) 33 Rawls’ defence of ‘justice as fairness’ as a set of principles for checking the basis structures and institutions of society is a good example of this level of argumentation (Rawls 1971).

22

Chapter 1

studying processes of justification. Can the analyst assume a neutral position, or is he/she inevitably drawn into the ‘muddy waters’ of normative judgment? In our view, the recognition of the rationales and justifications (in plural) that govern ‘sustainability thinking’ is an important step towards better support for decision making. We realise that the following sections deal with sustainability on a rather abstract, even ethereal level. Nevertheless, even in writing this chapter we have kept our specific research issue and policy setting – that is, the question of (nuclear) energy policy in Belgium – at the back of our mind34. Moreover, the following chapters will progressively describe the implications of the concept of sustainable development on a more practicallyoriented level, and the closer we get to this practical level, the more we will be able to focus on our specific research theme by including more down-to-earth examples taken from (nuclear) energy policy35. These chapters are therefore on the one hand based on the foundations laid out here, but – in a reflexive twist – also provide further support for our (meta-)theoretical choices. But let us not get ahead of ourselves: firstly, we should at least gain some more insight in the cluster of concepts frequently associated with sustainability.

2 Characteristics of sustainable development We agree with a number of observers that sustainable development in effect owes its staying power for a large part to its intuitive appeal on a ‘naïve’ level of understanding. It is this ‘naïve’ conception of sustainability that we will now try to explore further, before entering into the more ideologically coloured domain of integrated definitions. Drawing on this most fundamental level of basic understanding seems necessary to us, taking into account the fact that over the years, sustainable development has become a highly politicised principle. In light of our main objective in this chapter – namely to present an ensemble of conceptions, while exercising care not to present one in a more favourable light than the other – taking a step back might be a good idea. However, before proceeding to the results of this analysis, some remarks seem relevant here. A fairly reasonable objection could be that there will always be values at work in selecting the information that underpins our list of characteristics, and that the ‘description’ will therefore also contain a possibly insidious ‘prescription’. While on a fundamental level it is true that any kind of selection will be value-laden, we believe that the dangers of being overly prescriptive are seriously moderated by the fact that we are of course not the first to attempt to provide a 34 Resulting of course from the fact that the (meta-)theoretical framework has evolved during the course of our investigation, in a process of constant ‘dialogue’ with the empirical findings. Thus, while in the final result of the investigation the (meta-)theoretical framework is logically presented at the beginning of our exposé, this was – chronologically speaking – certainly not the case. 35 One consequence of this particular policy setting is for instance that although the problem of international distributive justice is of central importance in the global discourse on sustainability (e.g. the principle of ‘common but differentiated responsibilities’ enshrined in the UNFCCC (UNCED 1992)), this particular question will occupy a less central position in our work – e.g. the UNFCCC and the greenhouse gas emission reduction obligations agreed in the Kyoto protocol (-7.5% compared to the 1990 levels for Belgium in the period 2008-2012) will be accepted as given, without further need for justification within this context.


23

more or less exhaustive list of fundamental constituents for sustainable development (see e.g. Lélé 1991; Godard 1994; Dobson 1996a; Gimeno et al. 1996; Nieto 1996; Peeters 1997; Harribey 1998; Tilman 1998; Zaccaï 1999, 2002; Gibson et al. 2005)36. Therefore, it is safe to assume that a diligent survey of the material produced hitherto can reveal a ‘complete’ and ‘unbiased’ (to the maximum extent possible) selection of dimensions. Furthermore, the concept of sustainable development has been in use for quite some time already now, and an enormous amount has been said and written about it since the early beginnings, so that it is very likely that the major contours of the concept have by now been established. Lastly, we have also been guided by the practical applicability of the conceptions to our research topic, thereby focusing on the most frequent questions actually raised in the context of energy policy, rather than on any possible theoretic question37.

2.1

Appeals to sustainability

Following the work of Godlovitch (1998, pp. 294-301), Paskaleva (2000) and Zaccaï (2002, pp. 23-27), we can introduce some basic analytical distinctions in making appeals to sustainability38. At the most general level39 (and at the risk of being too obvious), the interpretation of the term ‘sustainable’ causes no great difficulties, it roughly means ‘enduring’, ‘lasting’ or ‘to keep into being’ (Pearce 1998, p. 69); in the words of Holland (1994, p. 169) …on any account of sustainability, something or other is supposed to be kept going, or at any rate not allowed to decline, over time…

Thus, at this level, it is not relevant whether sustainability refers to a given state of affairs or a process; nor does it matter much which states of affairs or processes are sustained: the plasticity of the concept is such that it can be applied to a number of contexts – i.e. one encounters ‘sustainable consumption patterns’, ‘sustainable agricultural practices’, and indeed, ‘sustainable energy systems’. Of course, while there are no ‘hard’ limits to this plasticity, the genealogy of sustainability roughly reserves its invocation to the domain of collective choices in situations that pose potentially significant hazards for the environment or human health, and/or have a dimension of (economic, environmental or

36 Variably called ‘core ideas’, ‘characteristics’, ‘objectives’, ‘concepts’, ‘criteria’, ‘dimensions’ etc. For the time being, we accept all of these denominations as signifying more or less the same thing. 37 For a more ‘exotic’ metaphysical exploration of sustainable development, see e.g. Godlovitch (1998), tracing the roots of the concept back to the Parmenidian philosophical tradition in Western thought (defending a deepseated orderliness), in contrast with the Heraclitian outlook (defending chance and change). While mentioning this example might seem a bit anecdotic, we include it here in order to underline that the ramifications of the sustainability discourse are in principle unlimited, and that any judgments on the practicability (or indeed ‘exoticism’) of one perspective over the other can only be defended contextually, remaining essentially open to revision. 38 The relevance of these distinctions will become clearer in chapter 2, where we will argue that problems regarding sustainability are not merely given, but require an active (and political) construction effort in order to make these problems manageable at the level of policy intervention. 39 ‘General’ of course being limited to some extent by our cultural ‘Eurocentric’ background (e.g. the fact that we write in English). The (undoubtedly more) fundamental problems of advancing sustainable development as an interculturally accepted paradigm are not dealt with here.

24

Chapter 1

social) intra- or intergenerational equity40. What matters most here is that sustainability refers to some extent to the notion of permanence, to steadiness over time. But this in itself is not enough: steadiness over time has to be generally desirable or legitimate – which sets the term ‘sustainable’ apart from the more neutral term ‘sustained’. To give an example, ‘sustained’ economic growth is in itself not equivalent to ‘sustainable’ economic growth, without further justification41. Sustainability always refers to a set of principles, to something which we value. Also implied is some form of causal relation of maintenance, captured in the skeletal phrase X sustains (maintains) Y42, with the emphasis on causation by human agency, or being amenable to human agency43. Simple as it might seem, there is however clearly something problematic with this ‘continuous causation’, and that is the nature, presence and frequency of countervailing factors – further aggravated of course the longer we want to sustain something into the future (Godlovitch 1998, p. 296) 44: …Regarding countervailing factors, apart from the usual disclaimers about the known defeating conditions of any continuous causal process, there are ghostly legions of unknown defeating conditions, the full extent of which can never be known. But this problem is just part of the standing uncertainty of science; hence, the ever-present shadow of the ceteris paribus – “all things being equal” – rider behind any generalisation…

40

To give an example, in the Belgian federal report on sustainable development (FPB 2002) the following thematic issues are taken into account (although other selection criteria also played a role, such as sufficient expertise on the topic, or falling within the competences of federal authority): production strategies of firms, ethical financing of firms, social economy, use of information and communication technologies, fishery and marine biological diversity, use of genetically modified plants, energy production and consumption, mobility and the transport of people, health at the workplace and tobacco consumption. Paskaleva (2000, p. 17), presenting an overview of national policy frameworks from different countries, lists the following recurrent problem fields: critical global ecological problems (e.g. global warming, forest depletion, reduced biological diversity), energy and resource consumption, economic and regional inequality, unemployment, poverty, population growth, the widening gap between rich and poor, the erosion of trust in central power and large institutions, and unprecedented levels of growth of industrialised countries. 41 This ‘semantic slipperiness’ can of course provoke some (intentional or not?) ‘misunderstandings’, for instance in the official UN discourse since the Brundtland report, where the ‘sustainable’ and ‘sustained’ growth denominations are both – seemingly interchangeable – in use (Pallemaerts 1995). 42 Y being the state of affairs or the process we value, and X being the causal factors which are necessary and/or sufficient to continuously bring about Y. 43 This can go very far. In an overview article on conceptions of sustainability, Dobson (1996a) presents a ‘diagnostic package’ for the causes of unsustainability he encountered in literature, and this includes ‘cause’/’cure’ couples such as ‘western science’/’ontological shift’, ‘western technology’/’appropriate technology’, ‘disempowerment’/’empowerment’, next to the more tangible (at least for policy makers) ‘unsustainable resource use’/’command and control’ or ‘trade’/’protectionism’ – and also, as a testimony of the divergence of opinions involved – ‘protectionism’/’trade’ (Dobson 1996a, p. 408). Chawla (1991) even designates the fundamental characteristics of Indo-European languages (namely their inclination towards expressing invariable states and emphasis on individuality) as one of the root causes of the environmental crisis. ‘Amenable’ should therefore be understood very broadly, and certainly not be limited to conscious human interventions. 44 Some economists (most notably Georgescu-Roegen 1971) have argued that according to the entropy law, economic growth is not only ‘unsustainable’ from a moral point of view but even impossible. While we do not want to argue with this view on theoretical grounds, we simply point out the fact that, when talking of sustainability, one should also have an idea of the timescales involved (‘how long do you want to sustain whatever it is you would want to sustain?’); and this timeframe will be more of an indication of our normative space rather than of the timeframes involved in some fundamental scientific research areas. Indeed, on the very long term (e.g. the timescales studied in evolutionary biology or geology) it makes no sense to talk of any ‘sustainability’ whatsoever.


25

This observation seems particularly relevant, since ‘proof of causation’ can be especially difficult in the controversial problem contexts where the principle of sustainability is usually invoked45. The crux here comes down to the conventional wisdom that the more complex a situation is, the more likely it is it will get out of hand; or put differently, our best efforts at sustainability work best only in the most isolated of circumstances (e.g. in a laboratory)46. Now, in view of the commonsense logic behind the above statements, it seems quite legitimate to share Stengers’s (1999, pp. 32-33) feelings of bemusement concerning the reason why sustainable development has acquired such (political) importance: …La nécessité, affirmée par les discours sur le développement durable, d’avoir à “prendre en compte” la durabilité a ceci de très intéressant qu’elle pose comme un nouveau défi ce qui aurait dû aller de soi…

On a similar note Latour (2003a) draws the attention towards the ‘striking banality’ of the precautionary principle (one of the main subsidiary principles of sustainable development)47, as he maintains that it is not the growing importance of this principle in public decision making which should somehow be explained, but rather the fact that precaution was excluded for such a long time from public decision making in the name of ‘scientific certainty’48! It is certainly worth quoting him at some length here: … Pour les domaines à fondement scientifiques (par exemple le nucléaire civil, l’amiante, les innovations techniques), et pour eux seulement, on avait inventé de renverser le cours ordinaire des choses et supposé l’existence d’une forme d’action rationnelle qui nous paraît aujourd’hui de plus en plus fantasmatique : les experts s’étant réunis, l’information pertinente ayant été amassée, la cause étant entendue, l’action suit par voie de conséquence. La réalisation pratique, dans cette vision, n’ajoute rien d’essentiel au savoir; une fois la décision prise, la vigilance n’est plus nécessaire sinon pour les détails d’application. Ce modèle positiviste de l’action rationnelle, s’oppose, on le voit, à ce que l’on pourrait appeler le modèle expérimental de l’action, soit celui que nous suivons dans toutes les autres formes de décision, et que le principe de précaution va nous obliger bientôt à rappliquer dans les mêmes termes aux décisions prises jusqu’ici au nom de l’expertise savante. Ce modèle nouveau n’est pas moins rationnel que l’autre, j’espère le démontrer, mais il n’obéit pas aux mêmes définitions de la raison, de l’expérience, de l’information et de l’action que l’ancien modèle positiviste…

45

We will return to this issue later in more detail. For now, Godard’s (2003, p. 4) description of what he calls “…des univers controversés…” can serve as an indication of what we mean by ‘controversial problem contexts’, namely problems which (a) cannot be perceived immediately, but require the intervention of scientists and/or other social actors in order to put the problem on the agenda (e.g. global warming); (b) are subject to uncertainty and scientific controversy; (c) require the representation of ‘absent voices’ in the decision-making process (e.g. nature, future generations); (d) could lead to irreversible changes, therefore making it impossible to ‘experiment’ in real-life conditions (e.g. the safety of nuclear installations). 46 One of the central tenets of Beck’s theory of ‘risk society’ (cf. Section 3.2) is precisely that the borders between ‘society’ and the ‘laboratory’ are being erased in contemporary (western) societies. Beck’s (1992, p. 69) claim on the development and application of environmental sciences in the age of ‘risk society’ is that “…Society is becoming a laboratory. (…) A permanent experiment is being conducted, so to speak, in which people serving as laboratory animals in a self-help movement have to collect and report data on their own toxic symptoms…”. 47 Compare to Stirling (1999a), who, in an overview paper prepared for the European Commission, uses old English sayings (e.g. ‘a stitch in time saves nine’, ‘don’t put all of your eggs in one basket’, ‘don’t burn your bridges’, etc.) as a framework for discussing the broad implications of the precautionary principle. 48 Even in the face of scientific warnings and evidence abouts hazards – see the very revealing historical casestudies (in the period 1896-2000) in the “Late Lessons from Early Warnings” report commissioned by the ‘European Environment Agency’ (EEA) (Harremoës et al. 2002).

26

Chapter 1

The Belgian federal planning bureau (FPB 1999, 2002) adds to this that the novelty of sustainable development also resides in the fact that with the advent of the concept as a policy principle, ‘development’ also became a task of the ‘developed’ countries – while, broadly speaking, this task was previously (in official discourse) reserved for the ‘developing countries’ (as ‘something they had to do in order to get richer’)49. In order to stress this ‘novelty’ of something which, on all accounts, should not be ‘novel’ at all, Zaccaï (2002, p. 39) proposes to take up the innovative character of sustainable development (or, more accurately, the strange combination of continuity and disruption that it represents in the flow of the development discourse) in his own list of ‘key characteristics’, at least as a point of debate. Drawing upon these basics of common understanding, we can now distinguish three possible ways in which an appeal is made to the principle of sustainability. When reading the following paragraphs, one has to keep in mind that these separate categories are mainly deployed for analytical purposes – a specific appeal to sustainability can never be fixed on one level, there will always exist inevitable instabilities and fluctuations. At the same time, we will witness the emergence of some key concepts, which appear to form the lexicon of the discourse on sustainability; and the first outline of core issues or challenges which will be taken up as central problems in the remainder of this dissertation. 2.1.1

Sustainability as a ‘manifest image’ of a form of life

‘Manifest’ is used here in order to indicate something which is accepted at face value; something which is, without too much reflection, picked up from the natural and social context50. It is not something which has to be ‘uncovered’ by scientific (philosophical,

49

Of course, the concept of ‘development’ has a long history of interpretations and contestations in international politics, which we cannot document here. For further reading, see e.g. Rist (1996) – generally considered as a reference work on the subject matter – and Zaccaï (2002, pp. 73-116). A more or less problemfree definition of ‘social development’ would be “…the process of change concerning the life circumstances in society (i.e. human, ecological and economical – our addition), including the possibilities to take decisions and propose action…” (FPB 2002, p. 4, our translation). 50 The term ‘manifest image’ is derived from the philosopher of science Sellars (1963, pp. 127-196). We realise that on this level of generality, we venture into a deeply philosophical territory inhabited by such famous characters as Wittgenstein (with his notions of life form and language game), Habermas (with his notions of Lebenswelt and Lebensform), Geertz (with his notion of common sense), etc. (for an overview, see van Brakel (1998, pp. 61-81)). In the context of this dissertation, it is not necessary to enter into a discussion about the similarities, differences, strengths and weaknesses of these concepts as a support for a general philosophy of science or technology (for a good introduction to these subjects, see Weiler and Holemans (1993, 1994)). We ‘simply’ want to point at a level of understanding prior to scientific understanding; the elusive background of all justifications; a point of reference which is inescapable, even for scientists, e.g. when agreeing on ‘sufficient’ proof of causation, designating ‘categories’ for their research, motivating the importance of their research, etc . Korthals (1994, pp. 22-26) talks about the ‘continuity between the sciences and the everyday experience’ (“…we can thus state that science is an institutionalised way to continue the everyday discussions about and experiences with knowledge claims in modern society...” (p. 24, our translation)), while van Brakel (1998, pp. 73-81) convincingly argues that scientific reasoning is always grounded in the manifest worldview, with ample examples taken from the theory of natural kinds, logic, quantum physics, epistemology, etc.


27

anthropological, sociological and so on) inquiry – e.g. in terms of deep-seated cultural structures, ‘the order of things’, etc. ‘Image’ is used here in order to evoke the largely metaphorical character of this everyday understanding of sustainability. We are thinking here of phrases such as ‘our Common Future’, ‘from one Earth to one World’, ‘the Limits to Growth’, ‘spaceship Earth’, etc. that seem to capture the ‘essence’ of sustainability in a few simple, basic ideas and mental models (for an overview, see Zaccaï 2002, pp. 64-70)51. As an integrated and holistic view, the manifest image has to be easily applicable to very divergent life experiences and contexts; hence its general and diffuse character. Van Brakel (1998, p. 71, our translation) says about ‘manifest’, ‘everyday’ worldviews that …on the everyday, experienced level everything is real; relative to this level is the reflective level, where everything is constructed…

It is precisely this global distinction that we also wish to introduce here in the context of sustainability: a ‘manifest’, ‘everyday’ understanding, where everything is intrinsically connected to everything, where a concrete event is easily integrated as a specific example of the more general outlook, where no attempt is made at conscious systematisation or exhaustive causal explanation; and a more ‘expert’ level of understanding, be it ‘philosophical’ or ‘hard’ scientific, where this is no longer the case. Before acquiring a positive content the demand for sustainability is, on this level, often experienced as a ripple in the otherwise smooth everyday existence. An appeal to sustainability is then mostly made as a form of denunciation: ‘your development model – be it with regard to industry, demography, agriculture, energy etc. – is not sustainable!’ (Godard 2003, p. 4). This of course does not represent a disinterested statement; it implies a demand to redress the balance – thus it is inherently moral52. Manifest images are, by definition, not accessible to drawing up an exhaustive scientific inventory; but just to give an indication of what has

All of this should however not make us blind to a growing interpenetration of the scientific and manifest worldview, one of the central tenets of Giddens (1991) and Beck (1992) (cf. Section 3.2), who point out that in late-modern societies (such as ours), people become increasingly dependent on ‘mediated knowledge’ (originating from newspapers, popularised science, self-help books, etc.) in order to ‘design’ their life plans in the vacuum left by the decline of more traditional ways of life. 51 Kempf (1994) and Hajer (1995) present a critical view (already apparent from the derisory title of Kempf’s book : “La baleine qui cache la forêt ”) on the ‘manifest image’ and its metaphorical nature, using terms such as ‘dramatisations’ and ‘mythologies’ (“…le fait de présenter une tendance comme à la fois globalisante (elle constitue une dimension partout à l’oeuvre) et porteuse de catastrophe…” (Zaccaï 2002, p. 64)) to denounce the use of these ‘mythologies’ as shrouds to cover up real social (local) conflicts over (the use of) nature. We believe that, although the sustainability discourse indeed often takes on the form of a certain ‘dramatisation’, and although there always exists a danger of a colonisation of the manifest image by politically inspired visions (but where to draw the line between these two?), these authors are presuming too much. Therefore, we prefer the more neutral denomination of ‘image’ rather than ‘dramatisation’ or ‘mythology’. 52 Compare to Mormont (1999, 2000), who suggests a reading of sustainable development as a concept “…qui naît du croisement de divisions qui sont devenues insupportables et d’autres qui sont devenues impossibles…”. He goes on to enumerate these ‘divisions’: between the rich and the poor (in one country, and between countries); between those who undergo the risk of (technological) development and those who reap the advantages; and between science and the social utility and/or relevance of scientific knowledge – and he suggests that ‘the environment’ represents ‘the canvas’ on which these tensions are painted out. To this, other divisions can be added at will, for instance the difference in political struggles over the environment in developed and developing countries, the tension between the prevailing dominance of the market logic at the global level and the need for more international agreements, etc. (Redclift 1987).

28

Chapter 1

become, over the years, a reasonably well-established list of development problems and/or challenges, we can perhaps follow Godard (2003, pp. 1, 4) in his enumeration53: …Il faut d’abord compter avec la conviction partagée par beaucoup à travers la planète que les processus de transformation physique et sociale à l’oeuvre, pour partie représentés comme processus de développement économique, sont incompatible avec la perpétuation du monde qui leur est familier et auquel ils tiennent, même pour ce qui était considéré jusqu’alors comme le plus sûr et, à l’échelle d’une vie humaine, le plus stable des éléments, comme le climat ; la conviction également que ces transformations deviennent insupportables à un nombre important et croissant d’êtres plongés non seulement dans la misère matérielle mais aussi dans l’exclusion sociale et la destruction culturelle. (…) Les générations présentes découvrent en la planète un monde fini et fragile dont elles épuisent les ressources et menacent les équilibres écologiques, provoquant une rareté dont le règne s’étendra à l’ensemble des générations à venir; ce pouvoir nouveau acquis par l’homme lui donne une responsabilité nouvelle envers les générations futures: appartenant à toutes les générations, la terre et ses ressources sont un patrimoine commun de l’humanité qui doit être préservé et géré pour le bien de tous ; il en va de la survie de l’espèce humaine, bien commun suprême ; en dépit de la complexité de notre monde qui empêche encore la science de toujours faire valoir ses certitudes, l’irréversibilité des évolutions majeures qui sont en train d’affecter la planète commande d’agir sans attendre selon un principe de précaution …

From the point of view of policy making, an appeal to sustainability on this level can and, most likely, will be problematic. There is an inherent tension between the manifest image and the more ‘expert’ interpretations which are required in order to issue policy guidelines. Indeed, in order to issue reliable policy measures, one needs to have a good idea of the ‘true’ cause of unsustainability – the FPB (2002, p. 4) for instance clearly stresses …the need of knowledge in general, and scientific knowledge in particular, in order to describe the possibilities to steer development…

But it might very well be that upon further investigation the ‘true’ cause of the perceived sustainability problem cannot be found54; the ‘true’ cause might be counterintuitive; and/or some causes for the perceived unsustainability of the situation might not open to control by policy measures (in principle or because of the limits of the mandate of the policy actor involved) – e.g. when the root cause of unsustainability is thought to be international trade, and a demand is addressed at the national policy level to solve the problem.

53

Another indication (bearing in mind the methodological limitations of opinion surveys as constructions of social reality) can be found in the results of a quantitative survey on sustainable development in Belgium, carried out by Bruyer et al. (2002). They found that although the concept of sustainable development was not well-known in itself (only 35% of the people addressed recognised the term), the problems usually related to the concept were all represented in the top five of most pressing problems for society (thus also lending credence to our observation that prior to ‘sustainability’ as a positive idea, the ‘unsustainability’ of present situations holds sway over public opinion) : health problems, poverty, unemployment, environmental problems and criminality / violence. Those who did know the concept related it to ‘environmental problems’ (27%), ‘a long-term vision’ (28%) and ‘an international dimension, concerning economic or social affairs’ (19%). People generally felt ‘concerned’ about environmental problems (93%) and were generally of the opinion that things had gotten worse over the last ten years (concerning poverty (75%), the gap between the rich and the poor in Belgium (75%), the state of the environment in Belgium / the world (41% / 73%)). People also indicate that when choosing between ‘the environment’ and ‘the economy’, the former is more important to them. 54 In complex problem contexts, one usually deals with different interwoven cause-effect chains, so that each cause can be in its turn an effect, and so on – so one has to apply a ‘stop’ criterion.


29

A word of caution seems in place here: we are not saying that ‘good’ policy measures can only be the result of a suppression of the manifest image by a more ‘scientific’ problem investigation. Neither are we implying that the manifest image should just be accepted without further reflection. What we are saying is that there is a tension between both the ‘manifest’ and the ‘scientific’ worldview; that this tension is – at least partly – irreducible55; and that public policy making (as one sphere of human agency) has to find good ways of dealing with this tension. In fact, this is one of the key issues we wish to explore further in this dissertation. 2.1.2

Sustainability as a ‘vision’

It is also possible that an actor, when referring to the principle of sustainable development, already has a positive image or idea in mind (described in terms of ‘criteria’, ‘ultimate goals’, etc.)56. Sustainability then refers to a ‘gap’ between a perception of an existing situation and a (positive) conception of a desired future situation (hence, a gap between ‘is’ and ‘ought’). The principal difference with the sustainability as a ‘manifest image’ is that in this case a positive direction of change is implicitly or explicitly included in a certain actor’s definition of the problem, as opposed to simply indicating a tension or an undesired situation (which have to be cancelled out by manipulating the causes deemed to be relevant). Visions have two major features: they are a mental image of an attainable future shared by a collection of actors; and they guide the actions of and interactions between these actors (hence the essentially intersubjective character of visions)57. They are rooted in an actor’s assessment of past experiences and expectations of the future, and they delimit a range of possibly attainable futures (Grin and Grunwald 2000, p. 11). Visions in many ways occupy the middle ground between the ‘non-reflective’ reference to sustainability as a manifest image, and the ‘entirely reflective’ reference to sustainability as a policy target (cf. Section 2.1.3). They are also metaphorical in character, but the metaphors used here are already more ‘detailed’, ‘precise’; it is easier to discuss visions and their implications on a ‘rational’ level58. At the same time, at the level of the visions, we are drawn into a political battle over the concrete interpretation and meaning of the manifest image. Visions are, on the one hand, dependent on certain context conditions 55

The problem we signalise here is but one aspect of what is generally known among philosophers as the ‘frame problem’ or the ‘problem of complete description’ (van Brakel 1998, p. 18). 56 The Belgian ‘federal plan on sustainable development’ (ICDO 2000, 2004) accepts three ‘ultimate goals’ of sustainable development (paraphrased here, see §64-66 for the official version): a) the economic goal of fulfilling the needs of present generations (including health, sanitation, clothing, food, energy, water, housing, etc.; but also less tangible needs, e.g. dignity, personal development, social integration), without making it impossible for future generations to fulfil their needs (this implies e.g. changing patterns of consumption and production towards less environmentally damaging alternatives); b) the social goal of equity (i.e. equal chances in life, e.g. concerning financial means, natural resources, cultural integration) in our society and between societies, with special regard for the minimum wage earners; c) the environmental goal of respecting the limits of natural resources (in their ability to provide base materials or as sinks for wastes). 57 As such, they are related to the French concept ‘futuribles’ (a term coined by de Jouvenel in 1963 (Grin and Grunwald 2000, p. 11)) and the German ‘Leitbild’ (introduced by Mambrey et al. 1995). 58 Hence, they lend themselves to scientific investigation by certain branches in technology assessment, e.g. ‘vision assessment’ (Grin and Grunwald 2000) and ‘metaphor assessment’ (Mambrey et al. 1995).

30

Chapter 1

(e.g. constellations of actors, historical backgrounds) – and have to be assessed with respect to these contextual issues (Grin and Grunwald 2000, p. 49) – but on the other hand they cannot be restricted to specific contexts – they have a certain integrative power which makes them suitable to integrate different contexts. They indicate a course without tracing it in detail59. But visions do not only open up to perspectives on certain aspects, they can also downplay or obscure others. Hence the dangers of relying on only one perspective for policy development; not to mention the impossibility of steering a (technological) development towards that vision in modern pluralistic societies, in the sense that the steering process would yield exactly and guaranteedly the results intended (Collingridge 1980)60. Hence, investigating and mediating between the different existing visions on sustainability (in our context of energy policy) in an effort to construct or identify possibilities for shared understanding and action, next to areas of contestation, seems to be a second major challenge for our work. 2.1.3

Sustainability as a ‘policy target’ or ‘goal’

It is here that we encounter sustainability as a ‘fully equipped’ concept – clearly defined by its application to clear-cut problem fields (which clearly fall under the field of application implied in the concept of sustainability); by scientific analysis of the causes of unsustainability and relevant trends in these fields; by definition of quantifiable ‘policy targets’ (for achieving a desired level of sustainability in the problem fields), ‘intermediary targets’, ‘indicators’ (needed to measure progress towards the targets), policy measures, etc. It is obvious that this kind of reference to sustainability heavily draws on an expert point of view, supported by a large political consensus (in a perfect marriage of knowledge and power). The inherent danger here is that sustainability is redefined in terms of what is already known and open to technical mastery. Hence the third challenge for our work is to navigate between the undeniable advantages of having a clear view of sustainability in term of (a) policy target(s), and the dangers of the premature closure of a more fundamental political debate according to this model. In order to avoid misunderstandings, a last (again seemingly obvious, but nevertheless necessary) remark seems in place. While it is difficult (and dangerous) to make any definitive statement on ‘widely accepted’, ‘everyday’ context-free views on sustainability, we believe (almost) everyone can agree that the ‘ultimate’ (and ultimately unattainable) goal of sustainability in the energy field is to devise a set of energy technologies which can meet human needs on an indefinite basis without producing environmental effects, 59 A good example of a vision (taken from Grin and Grunwald 2000, p. 47) would be ‘the information highway’ – opening up to a field of questions about ownership, access, possible ‘traffic jams’, possible payment for use of the highway, etc. A well-known example from energy policy is ‘the centralised energy system’ vs. ‘the decentralised energy system’ with the associated notions of siting, access to networks, the network manager, the users of the network, etc. 60 That governments nowadays face the combined challenge of steering and legitimacy problems in technology policy has almost become a commonplace, cited in almost any introductory chapter of contemporary books or articles dedicated to technology assessment (see e.g. Joss and Bellucci 2002). Societal processes, including the development and use of technology, are essentially institutionally embedded multi-actor processes, without any single actor being able to determine what is going to happen.


31

especially effects that are considered to be irreversible (slightly adapted from Elliott (2003, p. 46)). From the first part of this simple definition, it is clear that technologies using finite resources (coal, oil, gas and uranium) do not qualify as ‘sustainable technologies’ as such61. Also, everyone will agree that war does not contribute to a healthy development model, especially nuclear warfare; and that on these grounds, nuclear power again cannot qualify as a ‘completely sustainable’ technology, since there will always be a (however small) chance of proliferation of nuclear material. Besides the proliferation issue, we believe that other aspects of the use of nuclear power (e.g. the issue of radioactive waste and the long timescales involved, possible impacts of radiation on the human genome, the remote chance of a catastrophic accident) make this technology very hard to ‘imagine’ on the level of everyday understanding; and therefore nuclear energy will, according to us, have to rely more than other energy technologies on expert interpretations and rationalisations (not to mention the possibility that even the ‘experts’ might disagree on the matter). Hence, the tensions between both ‘worlds’ will perhaps always be bigger in the case of nuclear energy, as witnessed for instance by the ‘proliferation’ of risk perception and communication studies over the last few decades – trying to explore these tensions and, if possible, alleviate them (for a convenient overview of the major paradigms in this research field, see Löfstedt and Frewer (1998)). All of this would thus imply that the only viable long-term energy source is renewable energy62. While that may be true, the real political issue is of course the timescale involved, and the inevitable handling of tensions as long as the ideal state is not reached. For the near future – let us say over the next few decades at least – there seems to be no choice but to rely on fossil fuel resources for the bulk of power provision, and some would argue that we must continue to rely on nuclear power. Hence, the question is not (or rather, should not be): “Is nuclear energy a sustainable technology?”, but “Can nuclear energy contribute to a development process leading to sustainability – if so, under which conditions; and how can this question be answered?”63. Even when putting aside the above

61

Breeder reactors would increase the time of use considerably (to a thousand/a few thousand years), but compared to the ‘conventional’ thermal reactors operating in a ‘once-through’ mode, breeder reactors show disadvantages in terms of costs, fuel cycle safety, short-term waste issues and proliferation risks (MIT 2003). 62 Of course, these technologies are not perfect either. Most of them have some impacts on the local environment during functioning or related to the production of the energy devices (e.g. solar cells, wind power, tidal power, wave power), while the widespread use of energy crops might cause more significant problems, depending on the agricultural techniques and practices (e.g. impacts on biodiversity, pesticide and herbicide use, transport of crops to the site of energy production, etc.). Large-scale hydropower is also a more contested topic, because of possible loss of biodiversity, land use conflicts, expropriations, dislocated people, etc. (Elliott 2003). 63 A (political) appeal to ‘common understanding’ – “Nuclear energy is not sustainable, so why bother to even talk about it ?” – can serve to hide the often difficult choices that have to made (even when the nuclear option is abandoned). Of course, the same critique holds for the reverse position: “Nuclear energy is sustainable (or part of a sustainable energy mix), so why bother to even talk about it ?”. Both positions – which, as will become clear from our reconstruction of the debate on the Belgian nuclear phase-out decision, are less ‘academic’ than it might seem at first sight – serve to deny the authentic appeal to sustainability, arising from the different tensions explained above: the former by obscuring the tensions with the promise of paradise just around the corner (lying within immediate reach once nuclear energy will be phased out), the latter by assuring that these tensions are already adequately dealt with within existing structures.

32

Chapter 1

comments, the former questioning leads to a definitional approach of sustainability, prompting a substantive, largely context-free answer (e.g. yes/no/maybe, providing conditions x, y, z are met – see e.g. NEA (2000) and WWF (2000) for contrasting evaluations along these lines). Only the latter formulation provides the – for our purposes – necessary openness towards a broader context-dependent understanding of all of the issues involved. Perhaps we should make a distinction between ‘sustainable development’ (implying movement, change, etc. towards some desired end state) and ‘sustainability’ (implying the preservation of something we value), and with Meadowcroft (1997, p. 430) even accept the view that sustainable development does not necessarily depend upon the comprehensive adoption of practices which can themselves be sustained indefinitely64. As will become clear in chapter 5, our answer involves a move away from a substantive answer towards a more procedural one (without however completely abandoning the substantive side of the equation) – i.e. how to set up a proper process of collective learning that could lead to a substantive answer.

2.2

Defining sustainable development?

From the above, it will already be clear that the so-called definitional approach to sustainable development, where commentators attempt to sum up what sustainable development is by encapsulating its meaning in a definition, leads to some severe problems. Each definition necessarily advantages one aspect of sustainability over another, thereby reflecting the perspective of one particular actor or group of actors, and (Dobson 1996a, p. 402) …with something like 300 definitions available, seekers after enlightenment are often left as confused in the end of their search as at the beginning…

Semantic discussions on definitions of sustainability seem to have become a favourite pastime of certain commentators, in which we however will not engage. Let us just mention, with Zaccaï (2002, pp. 27-31), the distinction between three broad groups of definitions, mainly with regard to their position on economic growth and its limits65. The most famous and widely known definition of sustainability (found in the Brundtland report) is representative of the first group (WCED 1987, p. 43):

64 Subject to the following constraints however: the activity must still contribute in some way to ‘development’ (i.e. genuine improvement) and not foreclose the possibilities for future development. Thus the challenge of sustainable development can be understood as one of ensuring that society evolves in such way as to avoid historical pathways which constitute a deterioration (hence, the crucial importance of learning from past mistakes – see e.g. Harremoës et al. (2001) in the context of environmental and/or health risks and Laes et al. (2004c) for a historic view on (nuclear) energy policy making in the Belgian context) and pathways which inexorably lead to a point from which no further progress is possible. 65 This (perhaps most fundamental) strand in the sustainability debate goes back to the late ‘60s when the first criticisms on the incompatibility of economic growth and environmental protection could be heard (e.g. already in 1962, Rachel Carson’s book “Silent Spring” on the worldwide consequences of the use of DDT and other pesticides in Western agricultural practices was a strong charge against a naive technological optimism). This approach to conservation of nature and economic growth as competing activities culminated in the first report of the Club of Rome called “The Limits to Growth” (1972), which called for an end to economic growth in order to protect the environment.


33

… Sustainable development is a development that meets the needs of the present generation without compromising the ability of future generations to meet their own needs. It contains within it two key concepts: the concept of needs, in particular the essential needs of the world’s poor, to which overriding priority should be given; and the idea of limitations imposed by the state of technology and social organisation on the environment’s ability to meet present and future needs…

But other definition-like statements can be found in the same report (WCED 1987, p. 55): …Sustainable development is not a fixed state of harmony, but rather a process of change in which the exploitation of resources, the direction of investments, the orientation of technological development and institutional change are made consistent with future as well as with present needs…

Whereas most commentators (see e.g. Pallemaerts (1995, p. 74); Verburg and Wiegel (1997, p. 250)) agree that the Brundtland report contains an urgent plea for economic growth in the ‘developing’ countries (in view of the ‘overriding priority’ attributed to fulfilling the essential needs of the poor), its position on the ‘developed’ countries is less clear: the first definition suggests the existence of ‘limits’; while in the second no more reference is made to ‘the environment’ (and the limits it imposes on economic growth) – mentioning only a ‘harmonised’ process of social change in view of present and future needs. Other passages in the report also promote a concept of ‘qualitative growth’ – i.e. another model of economic growth than the one we were used to, but growth nevertheless. The other two groups are more definite on the issue: one is the ‘environmentalist’ orientation, which clearly demands respect for the limits of the ‘carrying capacity of ecosystems’; the other is the ‘economist’ orientation, with its emphasis on the intergenerational aspects of sustainability, demanding a ‘non-declining level of welfare’ over time. In the context of Belgian federal policy, the ICDO (2000, §66, our translation) produces a very ambiguous phrasing, reminiscent of the Brundtland definition: …The environmental goals of a sustainable development are defined in a way that respects the limits of natural resources through their management, by taking into account technological development and institutional structures. (…) At the same time, these goals take into account the fact that the environment has only a limited adaptive ability, as well as a provider of energy and resources, as in its capacity to absorb wastes and toxic pollution…

This situation of semantic confusion should not leave us in despair, nor should it even come as a surprise. Žižek (1998) points out that overarching political ideologies can only be successful precisely because they succeed in uniting contradictory perspectives on a symbolic level66. For Žižek, the truly political importance of such overarching ideologies lies in the recognition of demands of ‘entities’ which were left out of the official political discourse before (e.g., in the case of sustainable development, ‘ecosystems’, ‘future 66

Žižek gives the example of the ‘Solidarity movement’ in Poland (uniting ideologically very divergent groups such as leftist intellectuals, farmers, the Catholic church, etc.), but it is clear that sustainable development falls – at least conceptually – into the same category of overarching ideologies (uniting, or trying to unite, economic growth and environmental protection, ‘developed’ and ‘developing’ countries, nature and social aspirations, etc. in a common purpose). Lélé (1991, p. 613) colourfully describes this as follows: “…Sustainable development is a ‘metafix’ that will unite everybody from the profit-minded industrialist and risk-minimising subsistence farmer to the equity-seeking social worker, the pollution-concerned or wildlife-loving First Worlder, the growth-maximalising policy maker, the goal-oriented bureaucrat, and therefore, the vote-counting politician…”.

34

Chapter 1

generations’, ‘the poor’, etc.). Of course this makes such ideologies ‘essentially contestable’ and open to political struggle once the terms under discussion become subjected to more precise interpretations (e.g. ‘what are the needs of the poor?’; ‘what is qualitative growth?’; etc.). Hajer (1995) concurs with this when he approaches sustainable development as a political ‘story line’, able to absorb the shock of the erstwhile radical critique of ecological movement, but at the cost of a certain ambiguity or even ambivalence. Moreover, Van de Graaf and Hoppe (1996, p. 64) observe that political judgment is hardly ever characterised as the simple application of a rock-solid principle to a particular case. Compare to Mormont (2002, p. 12): …La plupart des concepts politiques qui ont servi de référentiel aux politiques publiques et aux mobilisations de nos sociétés contemporaines sont en fait des concepts de ce genre. Le mot d’ordre de modernisation de nos sociétés européennes au sortir de la seconde guerre n’a acquis de la force qu’en devenant un concept ambiguë qui promettait en même temps le progrès économique et la protection sociale. Appliqué à un secteur comme l’agriculture, il se transformait en projet de l’agriculture familiale moderne, terme très ambigu qui porte en même temps un mouvement de dépassement de la paysannerie et de préservation de sa forme…

On the contrary, these authors argue that the concepts used to express political values or principles do not have an objective, ‘truth-seeking’ theoretical status, but rather serve a more practically oriented, ‘expressive’ function of mobilisation. This also implies a ‘fuzzy’ designation of the application context for the principle, as it is impossible to define a priori ‘objective’ criteria which would justify the reference to the political principle in the specific context in question. The point is that considerations of the above kind have led to the general demise of the definitional approach to sustainable development in scientific literature. The definitional approach is at fault on a fundamental level: it seeks to make the plural singular. Definitions are only interesting to us on account of their ultimate failure – that is, they remain definitions, in plural! Thus, if we want to keep the relation of this plural discourse on sustainability to (energy) policy open; if we wish to treat the failure of the definitional approach as a starting point for speech rather than as an obstacle, then perhaps we must erase the ‘proper names’ by which the concept has become known, and preserve the infinity of the task. Perhaps it is through the medium of the more anonymous language of common understanding, always taking meticulous care in order to avoid premature closure, that the concept of sustainability may, little by little, reveal its illuminations.

2.3

An operational list of characteristics

What we might retain from the preceding sections is that however vague, sustainable development nevertheless refers to a more or less stable ensemble of characteristics. Some speak of an ‘intellectual landscape’ (Godard 1994), others of ‘the mainstream’ (Lélé 1991), still others of ‘core ideas’ (Jacobs 1998) or ‘attributes’ (Crabbé 1997). This ensemble forms what we might call a ‘thin conception’ of sustainability, i.e. a limiting case – located perhaps somewhere in between the ‘manifest image’ and ‘visions’ of sustainability – in the sense that somebody who justifies a line of action with an appeal to the principle of


35

sustainable development can hardly do so without referring to the characteristics discussed here. Moreover, the specificity of the concept lies in the fact that all of these characteristics have to be referred to somehow; none can gain an a priori ascendancy over the others. We are well aware that a number of these characteristics have been defined more clearly in conventions, guidelines, declarations, or legislation67. Our move should not be interpreted as an implicit questioning of the validity of these official documents. It is mainly inspired by methodological considerations: the list of characteristics is meant to lay out the (minimal) common ground for us, in our role of researcher, serving as a guideline for a further questioning on how reference is made to sustainability in practice – even if this turns out to be at odds with ‘official’ views. Thus, as an operational reference for our work, we accept the following list of characteristics (chiefly based on Crabbé (1997) and the FPB (2002)): • Focus on the sustainability of development, not just economic growth: development refers to an improvement of ‘life circumstances’, ‘quality of life’, ‘standard of living’, etc. – including the capacity to change these circumstances; in any case this includes more than just a growth in gross domestic product (GDP) or the per capita income levels68. On the policy level, this has inspired a search for new indicators of development, ranging from ‘objective’ (e.g. energy consumption rates) to ‘subjective’ (e.g. welfare); • Inter- and intragenerational equity: this focus on development has to include equity considerations within this generation (thus strengthening the more ‘classical’ principle of social equity) and between generations. In essence, the intergenerational issue boils down to three questions: ‘What is a generation (e.g. is it a whole or an aggregate of individuals)?’; ‘Do we owe anything to future generations?’; and, ‘If we do have obligations, how do we define them?’ (Gosseries 2001, p. 295); • Prospective analysis: the concern for intergenerational equity considerably lengthens the horizon of the impact evaluation of proposed policy measures. The time involved depends on the decision context and is an indication of the extent of the moral horizon. Typically, in energy policy the ‘long term’ deals with the next 30, 50 or maximum 100 years (depending on the decision context and the technology involved); 67

This is especially true for the precautionary principle, which figured in a great number of international agreements, conventions, etc. (see Godard (1997) for an overview). The precautionary principle is incorporated in the ‘Treaty on the European Union’ as an overarching framework for environmental policy (Art. 174), and the European Commission has issued specific guidelines for application (EC 2000). Jurisprudence of European courts also adds to the evolving interpretation of precaution. Also, the principle of participation has been legally enshrined in the UNECE ‘Convention on access to information, public participation in decision making and access to justice in environmental matters’ (signed in Aarhus, 25 June 1998, entered into force on 30 Oct. 2001, and ratified by Belgium on 21 Jan. 2003). The complete text is available at ). 68 The limitations of GDP or GDP per capita are well-known: no information about the distribution of income (between societies and in society) is given; damages to the environment are not included (the so-called ‘externalities’); social, political and cultural development is not included, etc. Quality of life is however not entirely unrelated to GDP per capita: the Human Development Index (HDI, a measure of development used by the UN in the “Human Development Report” series, calculated on the basis of adult literacy rates, life expectancy at birth, and skewness of distribution of income, next to absolute levels of GDP per capita) is closely positively correlated to the GDP per capita.

36

Chapter 1

• Integration: public policy has to take into account the fact that causes for unsustainability can arise as well in the economic, social, ecological as the institutional spheres of society – including in the interactions and feedbacks between these spheres; and before proposing policy measures they should be evaluated according to their impacts on all of these domains; • Increased value of the environment: in recognition of the genealogy of sustainability, environmental problems or the impacts of public policy measures on the environment or human health have to be in the centre of public attention. This is also based on the realisation that economic systems are an integral part of larger enveloping ecological systems (e.g. as source of resources or ‘sinks’ for waste), and therefore, that the environment cannot be entirely reduced to a commodity in economic production processes; • Global responsibility: in recognition of the fact that the principle of sustainable development tries to establish a renewed sense of common responsibility for worldwide (environmental) problems. It is also recognised that this responsibility is more stringent for the ‘developed’ countries, in view of their historic share of pollution and of their capabilities (e.g. technological, financial, etc.) to combat global problems (the principle of ‘common but differentiated responsibilities’); • Precaution: public policy should take into account all (scientific) uncertainties involved; and particularly in the field of environmental or public health policy, the threat of serious or irreversible damage should lead to protective measures. It implies a break from practices where persistent scientific dissent could be used as an excuse not to take action at all; • Participation: public policy on sustainable development should reserve the opportunity for a consultation and involvement of a wider (concerned) public69.

3 Sustainable development as a principle of justification So far, so good. In our sweeping overview of the territory occupied by the sustainability debates, we have established that sustainability is either used as a vision, as a form of justification of a proposed and concrete action, or as a form of denunciation of existing trends, policy measures, developments, etc. – and all of this taking place in a large (and not a priori demarcated) field of application, extending from technological and economic development to demography, natural resource use, environmental protection, etc. Furthermore, in the course of this overview, a number of subsidiary concepts have emerged as likely candidates to fulfil this role as new principles of justification: ‘ecosystem health’, ‘our common world’, ‘intergenerational equity’, ‘fundamental needs’, ‘the precautionary principle’ (to name but a few). Also, at first sight this ensemble of candidate principles

69

The Brundtland report sees participation as a necessary condition for sustainable development. Also, principle 10 of the Rio Declaration reads clearly: “Environmental issues are best handled with the participation of all concerned citizens, at the relevant level.” In addition, participation is omnipresent in Agenda 21, as the concept is woven through the 40 chapters.


37

seems to have sufficient consistency in order to transform sustainable development into a screening principle for development choices, as witnessed for instance by the efforts of international, national and local governments to produce programmes, declarations and/or doctrines on the subject. In what follows, we will put this capacity to serve as a stable basis for justification to the test. This test will consist of different steps. Firstly, we advance Boltanski and Thévenot’s theory of justification (the ‘commonwealth model’, Boltanski and Thévenot 1991), as a potentially suitable (meta-)theoretical framework for our purposes (Section 3.1). In particular, the framework will be used by us throughout this dissertation in order to attain some conceptual clarity in the sustainability hodgepodge. However, as the commonwealth model rests on some potentially debatable epistemological assumptions (with potentially far-reaching methodological implications), we think it could be useful to bring these assumptions clearly out into the open. This will be done through a confrontation with two fundamental theoretical challenges70. The first one, Beck’s theory of risk society (Beck 1992), claims that an analysis of sustainability in terms of processes of justification at the policy level is misguided, or rather irrelevant. What matters for Beck is that contemporary society has become increasingly vulnerable to self-produced (technological) risks, and that this fact sets in motion a structural transition from ‘modernity’ towards ‘reflexive modernity’ which will eventually determine the meaning of sustainability71, but in any case excludes reliance on particular ‘risky’ technologies (including genetically modified organisms (GMO’s), nuclear power, etc.). Beck represents what we might call the position of ‘descriptive (structural) essentialism’ – i.e. processes of change can be explained by referring to the factual conclusion of ‘objective’ forces at work. As an observer, all one can do is simply record these changes (Section 3.2). The second challenge, coming from sociological theories of technological change, goes further. The claim is here that processes of technological change can and should never be explained by referring to justifications, which are, according to this school of thought, at best a posteriori mystifications or a legitimation of processes which are entirely determined by a particular contextual setting and a particular constellation of (social) actors. This second position is one of ‘descriptive and normative relativism’ – i.e. processes of technological change are entirely situation-specific, and furthermore, analysts studying these processes should refrain from taking any normative position on them (Section 3.3). After these theoretical elaborations on the strengths and weaknesses of the different

70 Other theories falling within the broad categories of ‘essentialism’ or ‘relativism’ could have been chosen. Again, we concentrate here on the theoretical outlooks which seem most relevant to our particular topic, i.e. the broad field of technological choices addressing the questions which are raised by an attention for sustainable development. Each theory in a sense looks at the complex social reality through a different lens. Revealing the contrasts between these different theoretical lenses will also help us in clarifying the specificities of our position. 71 Admittedly, Beck never discusses sustainability in his theory of risk society. The link between both concepts is nevertheless clear, in the sense that sustainability, through the inclusion of precaution as a subsidiary principle, encompasses discussions about acceptable risks for society. Beck, in advancing risk as the organising concept in (late) modern societies, therefore implicitly downplays other aspects of sustainability.

38

Chapter 1

outlooks, we conclude with a statement on a suitable (meta-)theoretical outlook for our purposes (Section 3.4). In a last section, we provide a preliminary exploration of the possibilities and limitations of the ‘commonwealth model’ when applied to the questions raised by sustainability (Section 3.5).

3.1

3.1.1

The ‘commonwealth model’ of Boltanski and Thévenot

Framing the problem of justification

In their book “De la justification. Les économies de grandeur” the French sociologist Luc Boltanski and economist Laurent Thévenot present a general sociological theory of processes of agreement and rupture, as alternatives to violence72. This theory is quite ambitious, as it attempts to bridge the traditionally sharp distinctions between different social sciences, most notably between sociology and economics. One of the most intriguing and original aspects of the theory is the assertion that sociological research is firmly grounded in political philosophy. This allows the authors to deal explicitly with some very persistent epistemological problems in social sciences, such as the position of the ‘researcher’ in relation to the ‘research object’, the problematic relation between the ‘particular’ and the ‘universal’, and the tension between different ‘action logics’ (lying at the base of disciplinary distinctions in social sciences). As such, the work of Boltanski and Thévenot is of interest to a large research field, perhaps most adequately described as ‘sciences of management’ (in a broad sense, including e.g. public policy). This pluridisciplinary character of the theory certainly makes it a very interesting prospect for studies in the field of sustainable development. Briefly outlined, their approach is the following. Boltanski and Thévenot’s key theoretical focus is on the forms of evaluation that social actors use to coordinate their actions. Primarily, the possibility of coordination among actors rests, according to the authors, on shared systems of equivalence, called ‘grandeurs’, functioning as a beacon for the actors by permitting them to characterise the relationships between themselves, and between them and the objects deployed in order to stabilise these relationships in the situation encountered (Boltanski and Thévenot 1991, p. 11) : … Pauvres en groupes, en individus ou en personnages, cet ouvrage regorge en revanche d’une multitude d’êtres qui, tantôt humains, tantôt choses, n’apparaissent jamais sans que soit qualifié en même temps l’état dans lequel ils interviennent. C’est la relation entre ces états-personnes et ces états-choses, constitutive de ce que nous appellerons une situation, qui fait l’objet du livre…

This focus on a ‘situational’ logic represents a first break with some strong traditions in sociological research, as Boltanski and Thévenot claim that the different forms of 72

In our presentation of Boltanski and Thévenot’s theory, we follow a recent summary given by one of the authors himself (Thévenot 2002). We also benefited from the excellent analytical work done by Duyvendak et al. (1992) and Isaac (1996). Isaac relates the ideas represented in “De la justification” (which, in a sense, was a synthesis of a decade of research) to previous and later work by both authors, which enables a more profound understanding of the core issues.


39

justification people refer to when they have to coordinate even distant actions with anonymous others, are not relative to people (or social groups), as cultural values are, but are adjusted to the situation encountered. But notwithstanding this adjustment to circumstances, justifications also always aim at going beyond contingencies and claim a general validity. Hence, a second rupture, this time with sociological research aiming at an explanation of justifications in light of the interests of the different actors involved, for instance in conflicts over (the use of) nature (Lafaye and Thévenot 1993, p. 496) : …N’y a-t-il qu’un nouvel habillage masquant des intérêts bien établis: des chasseurs défendent leur terrain de chasse sous couvert de “maintenir l’équilibre d’un écosystème”; des pêcheurs à la ligne assurent la pérennité de leur passe-temps en réclamant de haut et fort “qu’un droit à l’environnement prenne place parmi les droits de l’homme”? On pourra reconnaître les intérêts de groupes sociaux plus larges, ou de groupes institutionnels, les ressources de la nature étant exploitées dans le cadre des luttes qui les opposent…

That is not to say that explanation in light of personal or group interests does not play any role at all in Boltanski and Thévenot’s theory. The role of interests in the coordination of action is however limited to what they call ‘an arrangement’ (cf. Section 3.1.4) between actors, i.e. a situation where the actors are not challenged or generally do not feel the need to step outside the boundaries of the particular situation they are involved in order to justify the proposed action towards a broader constituency. However, based on previous research (e.g. Boltanski 1982, Boltanski et al. 1984, Boltanski and Thévenot 1989), the authors observe that there exists more than one form of making an appeal to general validity. Thus, the fundamental theoretical question Boltanski and Thévenot pose themselves becomes: ‘How can justification be valid and coordination be effective if several underlying principles of justifiable action are available and each is supposed to be universally valid?’. They give the following answer. People engaged in public dispute and critique refer to different ‘regimes’ or ‘worlds of justification’, each with their own criteria of validity and internal consistency. These regimes are called ‘commonwealths’73. Such regimes of justification make it possible for situated actors to engage in disputes with others on the ‘common good’. The commonwealths establish different registers of ‘grandeur’ and of denunciation to be employed in disputes. They are shared as well by the people who invoke them as by the observer standing outside of the dispute. They do not imply a search for consensus. Consensus is possible only within a given regime of justification – across different regimes, only compromise is achievable. Just for illustrative purposes, the different commonwealths as described by Boltanski and Thévenot are briefly outlined here (see also Table 2 for the transposition of the ‘commonwealths’ into ‘common worlds’):

73 Cité in French. Common English translations of the concept include ‘orders of worth’ (proposed by the authors themselves – see e.g. Thévenot 2002), ‘cities’, ‘value systems’ (Laurans 2001), etc. We prefer the term ‘commonwealth’ (as coined by Jacobs 1996) as a very elegant translation, conveying in one word the idea of a community of beings (‘common’) and grandeur (‘wealth’) – two of the key concepts in Boltanski and Thévenot’s theoretical apparatus.

40

Chapter 1

• The inspiration commonwealth is characterised by making reference, in action and justification, to transcendent values; • The domestic commonwealth involves reference to heritage, communal identity, relations of closeness, familiarity and habits; • The opinion commonwealth is built upon the basis of the degree of notoriety or fame. Famous people or objects are the usual references in justificational discourses. The stature of a person depends on the opinion of others, people are assessed according to their public audience; • The civic commonwealth is built on notions of membership of a political community, with (more or less) equality in membership and access. Individual compliance with the general will is the basis of stature, based on an innate capacity of people to ‘have access’ to this general will. The general will defines rights and duties through forms of free political association; • In the market commonwealth, action is motivated by the desire for gaining wealth or advantage through commerce. Order and social coordination arise through the market. Expensiveness is a sign of ‘grandeur’, dignity is based on a capacity for self-interested behaviour and a desire for personal property; • In the industrial commonwealth, argumentation refers to technical performance, and places emphasis on scientific and technical expertise as a basis for achieving excellence in ‘system’ management and design. We will first further investigate the authors’ theoretical conception of justification wihin the confines of the commonwealths (Section 3.1.2), before turning to the practical implications of the theory as a guide for the coordination of social action (Section 3.1.3 and 3.1.4). In section 3.1.5, we will investigate the epistemological and methodological consequences of the theory. 3.1.2

The ‘common grammar’ of the commonwealths

The theoretical foundation of the commonwealth model is found in classical accounts of political philosophy. At first sight, this might seem odd for a sociological theory, but as the authors explain (Boltanski and Thévenot 1991, p. 26) …les disputes n’amènent pas pour autant à un exposé systématique qui permettrait de remonter aux principes de grandeur qui fondent l’évaluation. Or c’est précisément à de telles exigences de systématisation et de remontée aux principes que doivent satisfaire les philosophes politiques qui sont sommés, pour convaincre, de faire la démonstration du caractère bien fondé des définitions du bien commun associés à ces grandeurs…

The authors continue with an enumeration of certain criteria for the selection of the philosophical treatises serving as a base for their model. Their focus on establishing legitimate orders leads them to discard political philosophies based on domination or power. Besides that, the selection of texts also has to be practically applicable – serving as a guide for action (thus leaving aside the utopical tradition), generally known, and have served as a basis for the design of ‘political technologies’ (Boltanski and Thévenot 1991, pp. 60-61) – i.e. instruments which allow the constitution of broad classes of equivalence.


41

The resulting selection of texts is then submitted to a ‘grammatical’ investigation, in order to establish the fundamental categories needed to build a legitimate political order74. The fundamental categories are called the axioms of the (legitimate) commonwealths. Each of the selected political philosophies proposes a ‘common superior principle’ (principe supérieur commun), orienting action like a distant point of light. The first axiom states that every human must have access to the commonwealth and is considered as someone having to profit from that particular common good. No-one can be a priori excluded from the commonwealth. This ‘principle of the human community’ represents a principle of ‘simple equality’ (A1 – principe de commune humanité). But in order not to describe the trivial case of a utopia of permanent agreement between all, there has to be a principle of differentiation. People can occupy different positions in the commonwealth, in different classes of equivalence (A2 – principe de dissemblance). However, the second axiom cannot conflict with the first one, so all members of the commonwealth have to have equal power of access to these different classes. One has to assume and assure that people have certain basic capacities (‘dignity’) in order to accede to the different levels of grandeur; they have to dispose of equal ‘trump cards’ at the outset – i.e. a principle of ‘equity’ (A3 – principe de commune dignité). The fourth axiom states that the different positions people can occupy in a commonwealth can be ordered or ‘measured’ according to a value scale. People can be classified based on the way they relate to the superior common principle, thus acquiring a certain ‘grandeur’ (A4 – principe de l’ordre de grandeur). Those who live by the principles of a commonwealth, have to give up their egoism (immediate pursuit of pleasure) in order to assure the establishment of the common good. They are by virtue of the distinctive classification scheme ‘rewarded’ for this with ‘grandeur’: they are considered and can consider themselves as ‘great’ (A5 – principe d’investissement). The last axiom states that, in order to establish a legitimate order, the benefits enjoyed by the ‘great’ in the commonwealth should also advantage the ‘small’ in some way – i.e. the grandeur of those who realise the superior principle must ‘radiate’ down towards the less fortunate (A6 – principe du bien commun). Models of justification which do not have a principle of common good and which block access to the commonwealth for certain groups of people (for instance by denying that they are human or human enough), can not be considered to be ‘genuine’ commonwealths. They cannot be seen as one of the broad types of argumentation which bring agreement without resorting to violence. As an example of such an illegitimate order, the authors refer to the model of eugenetics and national-socialism (Boltanski and Thévenot 1991, pp. 103-106).

74

“The City of God” by Saint Augustine; “La politique” by Bossuet; “The Leviathan” by Hobbes; “Le contrat social” by Rousseau ; “The Wealth of Nations” by Smith ; and “Le système industriel” by Saint Simon. Boltanski and Thévenot justify their selection of six spheres of legitimacy or commonwealths with a reference to their previous empirical work – i.e. these six forms of claiming legitimacy suffice to adequately explain their empirical findings. However, they leave open the possibility of the existence of more commonwealths, and the authors have indeed studied the genesis of new commonwealths in subsequent work (the ‘information’ commonwealth (Thévenot 1997), the ‘connectionist’ commonwealth (Boltanski and Chiapello 1999)). Criticisms have been levelled against this selection procedure (Negri 1993).

42

Chapter 1

Thus, summing up, Boltanski and Thévenot claim that their commonwealth model is able to meet the following conditions (required by the authors’ expressed ambitions of explaining any form of social coordination): 1. the principles are justified according to political theories seeking to establish legitimate orders; 2. the principles cover general norms in the population which are broadly accepted; 3. the principles are prima facie; in the situation concerned, a principle from another commonwealth can have more weight or greater stringency (certain considerations have decisive force until another principle outweigh those considerations); 4. the principles are universal, committing all moral agents to all moral subjects; 5. a set of principles can cover any ethical dimension to a specific problem; 6. the principles are overarching, to which specified norms can be subtracted. Admittedly, this formulation of different commonwealths could mislead one into thinking that the each particular commonwealth has the status of an ideal, rooted in universal presuppositions of argumentation, and able to be approximately realised. Indeed, Boltanski and Thévenot claim that the different commonwealths open up a perspective allowing people to go beyond local practices of justification and to transcend their particular spatio-temporal contexts. But in making context-transcending validity claims, people themselves are not transported into an ideal realm of ‘noumenal’ beings. Rather than presenting an ideal, in light of which deviations can be measured, the commonwealth model represents a methodological fiction, in order to make visible certain processes of social interaction. The commonwealths have to be transformed to common worlds through an active communicative effort, firmly situated in time and space. 3.1.3

From the commonwealth(s) to the common world(s)

So far, we have outlined six different ‘regimes of justification’ or commonwealths, rooted in firmly established traditions of political philosophy. These commonwealths are constitutive of value or ‘grandeur’75; they provide the grammar of calculation and rationality. As principles of evaluation they involve systematic associations of ideas – and thus they have some similarities to culturalist notions76 – but they go beyond that similarity

75

We will use both terms as synonyms throughout the text. Some theoretical perspectives on sustainable development make use of the ‘cultural theory of risk’, as developed by Douglas and Wildavsky (1982) (see also Thompson et al. (1990) and Wildavsky and Dake (1990)). This theory has been used in the sustainability debate for the construction of broadly conceived policy scenarios (see e.g. WRR 1994; de Vries 1996; Laes and Meskens 2001a; FPB 2002 – although not all of these authors are faithful to the letter of the ‘cultural theory of risk’, they are nevertheless obviously inspired by it). Put very briefly, the ‘cultural theory of risk’ claims that there is a causal link between forms of social organisation, the different rationalities they determine, and finally, as a result, the way people select and handle risks in society or the way they view nature. Furthermore, these cultural perspectives (known as ‘the hierarchist’, the ‘individualist’, ‘the fatalist’ and ‘the egalitarian’) are said to be a constant throughout human history. However, in spite of these far-reaching claims, the empirical basis of the theory has been shown to be very weak, leading Adams (1995, p. 201) to the conclusion that “…cultural theory, like the myths of nature it embodies, remains an abstraction beyond conclusive empirical verification…” (for similar critiques, see Boholm (1996) and Sjöberg (1997)). 76


43

to show how each of the multiple principles of evaluation entails discrete metrics and ‘measuring instruments’ for classifying people. All of this takes place on the rather tenuous level of the ‘grammar of the commonwealths’, but tells us nothing about agreement in ‘the real world’. In order to explain coordinated action in the real world (after all, this was the authors’ main concern), Boltanski and Thévenot need to make the transition from a purely grammatical level of investigation to a hermeneutic and semiotic one. The key to understanding how this transition works lies in the axioms outlined above. We recall that, in order to describe a legitimate commonwealth, people could not be fixed forever in a certain statute of grandeur within the commonwealth (i.e. axiom A3); therefore, in principle, proofs of worth have to be renewed each time persons are engaged in common action. People’s stature in a commonwealth is always surrounded by uncertainty, and needs to be stabilised by in objects in the ‘real’ world (Boltanski and Thévenot 1991, p. 165): …L’attribution d’un état – qui suppose une équivalence générale – à une personne particulière est une opération soumise au paradoxe de codage. Le code ou la catégorie étant une forme d’équivalence dépassant, par définition, les particularités d’un être, comment peut-on relier cette forme à ces particularités, ce que suppose l’opération même du codage ? (…) Ainsi, la preuve de grandeur d’une personne ne peut reposer simplement sur une propriété intrinsèque, ce qui supposerait déjà en amont une forme d’équivalence suivant cette propriété. Elle doit prendre appui sur les objets extérieurs aux personnes, qui serviront en quelque sorte d’instruments ou appareils de grandeur. (…) La référence à des choses qualifiées entraîne donc une extension du cadre de cohérence par laquelle les cités se déploient dans des mondes communs…

With this fundamental insight that the regimes of justification and the orders of grandeur which are derived from them are not related to different groups but to different situations, Boltanski and Thévenot break with a strong tradition in sociology77. The focus on situational justification also excludes – or rather, takes issue with – theories that limit justification to a mere ‘battle of ideas’ or an exchange of arguments (Boltanski and Thévenot 1991, p. 166) …L’épreuve de grandeur ne se réduit pas à un débat d’idées, elle engage des personnes, avec leur corporéité, dans un monde de choses qui leur servent d’appui, en l’absence desquelles la dispute ne trouverait pas matière à s’arrêter dans une épreuve…

Furthermore, Boltanski and Thévenot are not only interested in knowing what is happening within a single regime of justification, but also in situations in which different regimes clash or compromise with one another. With this notion of a limited set of regimes of justification, they try to find a middle ground between a formal universalism and an unlimited pluralism. Each ‘world’ makes possible the appearance of other ‘objects’ and requires a different kind of ‘proof of grandeur’78. Hence, each world is characterised by different elements, which serve as distinguishing marks for people to recognise which world(s) they are confronted with in a specific situation (Table 2).

77 Most notably the notion of ‘habitus’ (developed by Boltanski and Thévenot’s teacher, the French sociologist Bourdieu), which constrains (or even determines) the actions of certain groups in a prescribed way. 78 This also implies that the authors are not interested in an ontology of things ‘as such’ (i.e. the modality of their existence) – they are only interested in objects insofar as they are deployed by people in situations which require justification (état-choses).

44

Chapter 1

Inspiration

Domestic

Opinion

Civic

Market

Industrial

Superior

Flash of

Hierarchy,

Opinion of

Primacy of the

Competition

Scientific

Common

inspiration

tradition

others

collective

method,

Principle

efficacy, performance

State of

Spontaneous,

Benevolent,

Reputation,

Official

grandeur

defying

cautious

fame

representative

Passion,

Common

Desire for

Political self-

Private

Capacity for

creation

sense

recognition

determination,

interest

work

Merchandise

Means

Businessmen,

Experts,

clients

operators

Opportunism

Investment

Wealth

Functional, operational

reason Dignity

freedom from egoism Repertory

Spirit, body

Gifts

of objects Repertory

Children,

Superiors,

of subjects

artist

ancestors

Necessary

Personal risk

Duty

investment

Names,

Laws,

brands

decrees

Celebrities

Collectivities

Renouncing

Renouncing

privacy

particularism

of time, money

Relations

Singularity

of

Subordination,

Identification

honour

‘grandeur’ Persuading

Representaton,

Possession

Mastery

delegation of

implies res-

interests

ponsibility

‘Natural’

Dreaming,

Educating,

Mobilising

Conducting

Functioning

relations

imagining

reproducing

State of

Imaginary

Family

Audience

Republic

Market

System

Typical

Interior

Ceremony

Event

Manifestation,

Transaction

Test

proof

adventure

Expression

Flash of

Price

Effective,

of

insight

business

harmony

vote Appreciation

Public

Collective

opinion

decision

judgment

correct

making

Evidence

Certainty of

Example

inspiration Position of ‘smallness’

Routine

Success,

Law

fame

Money,

Measure

benefits

Vulgar,

Unknown,

Isolation,

shameless

trite

division

Loser

Inefficacy

Table 2. The six worlds in Boltanski and Thévenot’s theory of justification (Source: Boltanski and Thévenot 1991, pp. 200-262; table translated from Isaac 1996, p. 12)


3.1.4

45

Criticism, conflict, clarification, bargain and compromise

Thus, Boltanski and Thévenot paint us a world where social coordination and action is reigned by a fundamental form of uncertainty (Boltanski and Thévenot 1991, p. 171): …Soulignons qu’aucune situation, aussi pure soit-elle, ne peut éliminer à jamais la diversité des contingences dont le bruissement se maintient aux confins de ce qui est en ordre. La permanence de ce tohu-bohu fait peser une incertitude sur les grandeurs. La situation risque toujours d’échapper et d’amener à reconduire l’épreuve, comme le jet du dé ou le tirage d’une carte relançant la partie. En l’absence du bruit extérieur, prévaudrait un jugement dernier justifiant une distribution harmonieuse des états qu’aucun élément nouveau ne remettrait en cause. Ainsi le bruit du monde, que l’épreuve fait provisoirement taire, est ce qui le meut. Chacun des mondes dans lesquels se réalise le modèle d’une cité et qui, pris en lui-même, possède un caractère de complétude et d’autosuffisance, porte la trace, par ce tohu-bohu, de la possibilité d’autres mondes…

Under these conditions, it becomes an open question how social actors can ever come to an agreement. In answering this question, Boltanski and Thévenot, adopting an ethnomethodological attitude, stress that actors use and construct for themselves one of these commonwealths as the legitimate one in several situations, based on concrete ‘tests’. Such a ‘test’ can take on different forms: actors can either claim that the ‘proof of grandeur’ within one commonwealth is not valid, because it lacks sufficient support by objects (défaillance des objects/personnes) – e.g. a machine/operator does not perform according to specifications/quality procedures; or that the ‘proof of grandeur’ within one commonwealth is based on a grandeur from another commonwealth (transport de grandeur) – e.g. political power resting on personal wealth; or they can argue that another commonwealth should come into play in order to judge the situation by its ‘real’ merits (conflit des mondes) – e.g. the conditions of work in a factory do not allow for creativity; or it is possible that the conditions within one commonwealth are such that no valid proof is possible (within another commonwealth) (transport de misère) – e.g. structural poverty limits people’s chances of political participation, access to good education and work, etc.The order in which we represented these possible ‘tests’ is not arbitrary, as the conflict potential rises when we go from a ‘defect’ to a situation of ‘misery’. In this process of going from a ‘test’ towards a possible solution, actors try to mobilise objects to indicate that they have a ‘sense of common order’, or in other words, to prove that they refer to the proper commonwealth apt to the situation. Hence, the relative openness in Boltanski and Thévenot’s theory: in trying to arrive at ‘solid’ agreements, people act as creative ‘tinkerers’, using for the purposes of construction of agreement almost anything that comes at hand, particularly existing constructions, or the remnants of former constructions. Depending on the degree of conflict, Boltanski and Thévenot indicate three possible ways out (without having to resort to violence or power): 1. Clarification within one world: the solution to the difference of opinion is found within the limits of the world which seems most relevant for the situation in question (e.g. in the industrial world, when an operator fails to perform as requested, he could receive additional training or be replaced by another operator). The fact that an agreement can be reached within the limits of one world reinforces the relevance of that particular

46

Chapter 1

world (and its specific way of defining proof, relationships between people, etc.) towards other possible action logics; 2. A local bargain or arrangement: here, the actors involved agree to settle down to a deal – i.e. a temporal agreement that only holds good for the people involved. The stability of the local bargain cannot be assured, as no appeal to general validity is made; 3. A compromise: if actors cannot agree on a legitimate ‘test’ of their environment, because they keep referring persistently to different commonwealths, a temporary compromise might help to solve that problem for the moment. They then need to agree on a common good, without however letting prevail a specific grandeur of one of the commonwealths of the different actors in conflict. They do so by extracting elements and objects from the different worlds in question in order to arrange them in a specific disposition (e.g. an employee’s council in a company, trying to reconcile productive and social demands; or a socio-economic council trying to reconcile industrial and civic demands). As a consequence, the compromise is always vague and fragile. It includes an orientation towards the common good (this is what sets the compromise apart from the local arrangement), but this common good is not further specified in terms of the principles underlying the different commonwealths. An ‘intrinsically’ stable compromise is only possible when the creation of a new commonwealth (with a corresponding superior common principle) is undertaken, which is a very difficult endeavour. On this process of creation, the authors only say the following (Boltanski and Thévenot 1991, pp. 345-346): …L’indétermination du bien commun visé par le compromis devient de plus en plus problématique quand, avec la prolifération des objets composites, se constitue l’ébauche d’un nouveau monde et que se multiplient du même coup les épreuves dans lesquelles ces objets se trouvent engagés. Leur fragilité, la facilité avec laquelle ils peuvent être dénoncés conduit à un renouvellement très rapide d’épreuves qui ne sont pas jugées assez probantes pour arrêter la controverse. Les différends auxquels ces épreuves donnent lieu sont particulièrement favorables au travail d’explicitation pouvant conduire à la mise en place de nouveaux principes d’équivalences et à la clarification du bien commun visé. (…) La philosophie politique opère une mise en forme systématique de ces débats, qu’elle soumet à des critères de cohérence interne et de compatibilité avec les conventions admises par ailleurs. (…) Le travail philosophique constitue ainsi un moment fondamental du procès de généralisation qui, consacrant au bien commun des qualités ordonnées jusque-là à des fins particulières, achève l’universalisation des valeurs…

Regarding this last point, the most important question for our purposes is whether sustainable development can already take up its role as a ‘superior common principle’, or is it (for the moment) relegated to the status of a ‘compromise’ (cf. Section 3.5) ? Most of the remainder of Boltanski and Thévenot’s book is devoted to describing possible criticisms from one commonwealth to another, and possible compromises between commonwealths. At the end of the book (in a post face), in response to criticisms on an earlier version of their theory (published in 1989), the authors reflect on action without (the need for) justification – either because the actors involved are not subjected to criticism, because there is no crisis situation, etc. – but mostly by indicating some questions they


47

wish to explore further79. Furthermore, Boltanski and Thévenot’s model is by definition blind to forms of political action and commitment that, without making an appeal to an explicit political philosophy, nevertheless might invent new forms of political cooperation. This relative inattention to ‘action’ (instead of situations of reflection, where the action is suspended) provides an entry point for a different school of sociological thinking (the ‘(social) construction of technology’) presented in section 3.3. 3.1.5

Theoretical implications and evaluation

In view of the explicit desire expressed in “De la justification” to liberate themselves from traditional thinking patterns in sociology80, we will give a brief overview of the theory’s epistemological and methodological consequences81. Firstly, with regard to their ‘subject matter’ (the social actors involved in processes of dispute and formation of agreement), their model not only supposes that it is possible to construct a system of constraints bearing upon legitimate agreements; it also supposes that people have a capacity for reason, which allows them to recognise situations for what they are (or what they should be), and act accordingly. This is not a purely subjective capacity (in the form of an innate mental scheme), but it also lies enclosed, somewhat hidden from the immediate view, in a certain preconceived order of the world (e.g. institutions, prescriptions for proper behaviour, etc.), only to be fully revealed in a ‘test’ requiring proof. Thus, the conventions regarding proper conduct in a situation lie as well enclosed within the actors themselves (the subjects) as in the environment (the relevant objects). Therefore, Boltanski and Thévenot’s work also throws an entirely new light on the concept of ‘bounded rationality’. Whereas this concept usually refers to the cognitive limits on the rationality of an actor, in Boltanski and Thévenot’s work rationality is only possible insofar as it takes place within the boundaries and through the grammar of particular commonwealths. In this latter sense one should probably speak of ‘bounded rationalities’. The authors suggest (though they do not develop the thought any further) that people learn to reason using the elements of commonwealths by going from one ‘test’ situation requiring proof to another (Boltanski and Thévenot 1991, p. 185) 82: …Ainsi, c’est en allant au garage ou au supermarché, et non en bibliothèque pour lire Adam Smith dans le texte, qu’il acquerra la capacité à s’engager dans des situations fondées sur un principe de justification de nature marchande…

79

For a concise summary of the most important of these criticisms, see e.g. Negri (1993): “…Que veulent donc ces auteurs? Leur façon de prendre au sérieux ce que font et disent les gens me paraissent très importants. Mais après? Tout ceci ne finit-il pas par se réduire à un jeu de paroles, à un jeu transcendantal (qui feint d’être un ‘empirisme des idées’), d’autant plus incontinent et prodigue dans le gaspillage du patrimoine méthodologique (non seulement de la sociologie, mais aussi de l’histoire des doctrines politiques) qu’il est avare dans la production des descriptions concrètes ?... “. 80 Isaac (1996) speaks of a ‘third way’ in social sciences. 81 We do recall however that the authors are ‘only’ presenting a theory about how people arrive at justifiable agreements; their aim is not to develop a general epistemology of all forms of human knowledge. 82 Thereby also sidestepping (age-old) fundamental philosophical questions about how this link between ‘object’/’structure’/’substance’ etc. (on the ‘objective’ side) and ‘justification’/’action’/ ‘intentionality’ etc. (on the ‘subjective’ side) can be conceived.

48

Chapter 1

As a further epistemological consequence of Boltanski and Thévenot’s theory, there are no grounds to make a strict distinction between a moral judgment (le sense du juste) and a factual judgment (le sense de la justesse). In judging a situation, people indiscernibly mobilise a superior common principle stemming from one commonwealth, as well as the typical factual proof required within that commonwealth. They have to be ‘metaphysical’ and ‘practical’ at the same time. Given these epistemological insights, strong methodological consequences emerge. Boltanski and Thévenot insist that the sociologist’s job does not consist any more in revealing hidden genuine social mechanisms, behind unconscious individual or collective behaviours (usually strained relations or inherited values). They clearly want to break up with the tradition of ‘critical sociology’ with its clear distinction between scientific knowledge (the privileged domain of the critical sociologist) and common knowledge (the domain of the actors being observed) (Boltanski and Thévenot 1991, p. 172): …Notre mouvement diffère aussi de l’opération critique par laquelle, en se situant dans un autre monde, on a les yeux dessillés et l’on voit le premier comme artificiel et comme le produit d’une illusion, d’une “naturalisation”. Notre démarche de description, de l’intérieur de chaque monde, exige donc du lecteur qu’il suspende la critique qui découle, comme on le verra, de la connaissance de plusieurs mondes, pour se plonger dans chacun d’eux comme il le ferait s’il était pris dans une situation où la sincérité de son adhésion aux principes serait une condition de la justification de son action…

Sociologists from the critical school are accused of never revealing their own normative stance, by hiding behind a claimed scientific neutrality (see also Dosse 1995, p. 59)83: ...Ceci est absurde. Soit on est axiologiquement neutre et l’on n’est pas critique, soit on est critique et on a une position normative…

In fact, according to Boltanski and Thévenot, social actors are not blind but they close eyes to follow logical rules, the grammar of their actions in private, public, industrial, mercantile, religious or aesthetic life. They legitimate actions with arguments shared by everybody for their efficiency and are able to differentiate two kinds of arguments or arrangements, just as well as the critical sociologist: the legitimate ones which can be generally justified and universally agreed when criticised, and the illegitimate ones which can be used by actors in favour of some parties but which fail to justify or support agreements for the common welfare. Boltanski and Thévenot are aware however that the capacity for justice is not always available. Justification follows or goes with criticism except if the actor(s) involved answers violently or hides behind urgency. Thus, on the one hand, Boltanski and Thévenot reject a fundamental rupture between scientific (sociological) and common knowledge; but on the other hand they also refuse a purely descriptive discourse (Boltanski and Thévenot 1991, p. 11):

83 Jacobs (1996), with some sense of humour, compares the position of the critical sociologist with the elevated spot of the two old men on the balcony in the Muppet Show, commenting on the ongoing programme with the actors ‘down there’.


49

…Point non plus de ces personnages grandeur nature que les formes les plus littéraires transportent dans l’espace du savoir scientifique, au travers de témoignages souvent très semblables à ceux que recueillent les journalistes ou que mettent en scène les romanciers…

Hence, when applying the theory of the commonwealths, one has to steer a delicate middle course between those two extremes. For Boltanski and Thévenot, understanding a situation means having access to the representations people give of the situation, as the ‘material’ basis of their investigation. This means interrogating people in situations of crisis, when the need for justification and coordination in order to restore the action potential is clearly felt. In other words, moments of interruption of action, crisis, and subsequent search for compromise are the privileged entry points for sociological observation; it is here that the different commonwealths involved, including the mobilisation of objects destined to perpetuate the compromise between different commonwealths in time, will reveal themselves. The sociologist’s role is then to carefully record this process as closely as possible – an approach made possible precisely because the sociologist shares knowledge of the different commonwealths with the actors, the only difference perhaps being that the sociologist’s knowledge is more profound and reflexive (i.e. through the knowledge of the political philosophies underlying the different commonwealths). Having arrived at this point in our account, and before we give a first round-up of the methodological strengths and weaknesses of the commonwealth model when applied in the context of the sustainability debate, we think it could be useful to present some other very influential theoretical outlooks on processes of (technological) change in contemporary society, and their implications for the possibility of social coordination based on justification. This confrontation hopefully allows to bring into focus the fundamental theoretical choices at stake, and to formulate challenges for strengthening our own theoretical and methodological position.

3.2

Beck’s theory of risk society

Most commentators of contemporary society would agree that society nowadays is confronted with risks in a fundamentally different way than traditional societies used to be. Besides nuclear power, large-scale chemical facilities, hydroelectric dams, global warming, mad cow disease (BSE), persistent organic pollutants, endocrine disruptors, genetically modified organisms (GMO’s), etc. all figure prominently in recent risk research literature as paradigms of these ‘new’ risks (Klinke and Renn 1999). What sets these risks apart from more ‘traditional’ ones (e.g. floods, fires, earthquakes, etc.) is that they have (or can have) global consequences; they are not accessible to direct sensory perception (i.e. without a certain expert point of view, the risk would not ‘exist’ in the first place – they are ‘constructed’); and they are the result of industrialisation itself (i.e. ‘self-produced’, as opposed to ‘natural’ risks). As a result of the constructed character of the ‘new’ risks, they become open to political and economic struggle: the management of risks has become ‘big business’ (e.g. through insurance firms); social risk positions (‘Who benefits from the risk?’; ‘Who suffers the potential consequences?’; ‘Who is assigned the role of risk

50

Chapter 1

manager?’; etc.) are defined through scientific inquiry; and therefore knowledge about risks has gained particular political significance. To the German sociologist and philosopher Ulrich Beck, this state of affairs is symptomatic for what he calls a society in a ‘permanent state of emergency’ (or ‘risk society’), undergoing a transitional phase of ‘reflexive modernisation’. The theory of risk society has become quite influential and popular over the last decade, and has given rise to a large literature (see e.g. Giddens 1991, 1992; Beck 1992, 1995, 1997, 1999; Beck et al. 1994; Lash et al. 1996; Adam et al. 2000)84. We will briefly summarise some of the defining characteristics of risk society, with special attention for the elements of this theory which seem particularly relevant to our topic (i.e. policy making in view of sustainability), and we will end this section with the resulting theoretical challenges to our framework. 3.2.1

A brief guide to risk society

In his book “Risk Society”, Beck (1992) asserts that contemporary risks can no longer be defined as ‘undesired side effects’ of the modernisation process, in the sense that previously, industrial risks used to pass by largely unnoticed in public or political debates85. Beck talks of the ‘residual risk society’: a society where the consensus on the necessity of the risky activities was sufficiently large to legitimise the resulting risks as ‘side effects’. In other words, in the era of the ‘residual risk society’ both government and industry were able to calculate risks and relate solutions to unambiguous outcomes. They were able to do so because the premises of social order in the society of that time were relatively unambiguous. To give a concrete example taken from energy policy, one might argue (with some sense of simplification) that in many industrialised countries, the energy sector has been shaped in the past (roughly before 1990) by the dominant importance accorded to strategic issues and economic competitiveness: energy supply was made as diverse as possible in order to minimise geopolitical risks of dependence (e.g. oil deliveries) and measures were provided to guarantee a secure and continuous supply of energy at a reasonably competitive price for all concerned. At the core lies the vision of the post-war ‘social pact’: representatives of both employers and employees recognised the need for a growing economic output in order to maximise welfare, and direct state intervention was often encouraged. In many countries (including Belgium), this meant making an appeal to nuclear power. The transition to a risk society – as a new phase in the history of modernity – is then initiated by the growing conviction that some industrial and social risks have become simply unacceptable. But although the temporal and spatial aspects of contemporary risks such as nuclear power or human induced climate change are indeed unprecedented in history, it would be wrong to assert that these physical parameters are solely responsible for the above-mentioned transition. Beck’s main interest lies not in a ‘realistic’ discussion concerning ‘modern vs. traditional risks’ (i.e. ‘Is modern society riskier than traditional

84

For an easily accessible introduction to Beck’s theory, see also Weiler and Van Ooteghem (1999). Beck’s magnum opus was already available in German since 1986 under the title “Risikogesellschaft – Auf dem Weg in eine andere Moderne”. 85


51

society?’), but rather in the consequences of new risks in the domain of (social, cultural, scientific, etc.) representations. Furthermore, in Beck’s theory, the new technological risks go hand in hand with ongoing processes of individualisation and loss of tradition (e.g. regarding the educational role of the nuclear family, gender roles, structural unemployment, erosion of lifetime job security, etc.), leading to new risks in the social domain. Both developments strengthen each other, leading to a spiral of fundamental uncertainty: risk society. In risk society, the ‘traditional’ inequalities and conflicts of industrial society (e.g. between social classes) of course do not simply disappear, but, according to Beck, they are redefined in terms of risk in the field of representation (Beck names this the ‘progressive individualisation of social inequity’). The focus of the ongoing social debates slowly changes from the ‘distribution of goods’ (centred on the equity theme – i.e. ‘solidarity out of need’) to the ‘distribution of bads’ (centred on the safety theme – i.e. ‘solidarity out of anxiety’). In the following paragraphs, we will however chiefly focus on the role of technologically induced risks in Beck’s theory, as this carries the greater relevance for our purposes. Beck’s central claim is that as the new technological risks emerge, the institutional instruments of rational control in industrial society – more technology, more government, more market opportunities – no longer seem to be able to cope. According to Beck, the side effects of industrial successes have already undermined the foundations of the society out of which they have emerged. Modern institutions (e.g. law) are founded on the premises of accountability, responsibility and manageability (Holemans 1999). They can no longer live up to these expectations, because the new risks are no longer calculable in their consequences. Furthermore, through processes of social amplification, technological accidents not only have direct health or environmental consequences, but can also have large impacts in the economic, social and political sphere. Hence, according to Beck, risks management institutions (and governments in general) of industrial societies are faced with a severe legitimacy crisis – a diagnosis that has, since the publication of “Risk Society”, received ample empirical support (see e.g. Slovic 1993, EC 1999b, Rosa and Clark 1999). With his notion of ‘organised irresponsibility’, Beck denounces the loose coalition of administrations, government, industrial firms, experts, political parties, interest groups, etc., who, in the name of ‘progress’, have defended the interest of the industrial system without regard for the wider consequences for the environment, health or future generations (Beck 1992, p. 214): …Faith in progress replaces voting. Furthermore: it is a substitute for questions, a type of consent in advance for objectives and consequences that remain unknown and unmentioned. Progress is a blank page as a political program, to which wholesale agreement is demanded, democracy has been turned on its head by the model of progress. Officially, one is dealing with something quite different and always the same – economic priorities, competition in the global market and jobs. Social change takes place only in displaced form. Progress is the inversion of rational action as a ‘rationalisation process’. It is the continuing changing of society into the unknown, without a program or a vote. We assume things will go well, that in the end everything we have brought down upon ourselves can be turned back into progressiveness. But even asking why or wherefore has something heretical about it. Consent without knowledge of wherefore is the prerequisite. Everything else is heresy...

52

Chapter 1

Thus, the inability to calculate risk has made institutional monitors blind to the transformation of the pattern of risk in industrial society. Beck’s book is larded with examples of this so-called ‘organised irresponsibility’, but perhaps one particularly illustrative example for our topic might suffice to understand the general idea: insurance policy. Systems of insurance have become increasingly sophisticated and widespread with the rise of industrial capitalism (de Meere 1996). This is no coincidence, since the existence of insurance makes it possible for modernity to advance without facing continuous battles over guilt and responsibility (Beck 1992, p. 99): …it permits a type of ‘technological moralisation’ which no longer need employ moral and ethical imperatives directly. To give an example, the place of the ‘categorical imperative’ is taken by the mortality rates under certain conditions of air pollution. In this sense, one could say that the calculus of risk exemplifies a type of ethics without morality…

According to Beck, the principal significance of insurance lies in the fact that it allows the industrial system to deal with the future; through a network of insurance, providing protection against the ‘dark side’ of development, a consensus on the way forward becomes imaginable. But the consensus is always unstable (Beck 1992, p. 100): …a norm system of rules for social accountability, compensations and precautions, always very controversial in its details, creates present security in the face of an open uncertain future. Modernity, which brings uncertainty to every niche of existence, finds its counter-principle in a social compact against industrially produced hazards and damages, stitched together out of public and private insurance agreements…

Beck’s thesis is then that, with the advent of new risks such as nuclear power or GMO’s, the limits of this consensus have been transgressed: the (global, long-term and irreversible) consequences of possible accidents are of such magnitude that no form of adequate compensation for losses seems to be possible anymore (e.g. Chernobyl). Full private insurance operates (and can only operate) where losses have some stochastic or probabilistic character. Paradoxically, modernity has created risks which can no longer be processed by the modern instruments of calculation86. Nevertheless (and perhaps surprisingly), Beck is not fundamentally pessimistic about the future. According to him, the dangers of the present situation have opened up new and enlarged possibilities for political action. The increasing dependence of central governments on the global economic evolutions and, in their struggle for legitimacy, on expert knowledge has led to an important delegation of power to technical experts, research institutions, management boards, industrial laboratories etc. Thus, the traditional distinction between politics, science and economy (as expressed by the existence of a ‘political system’, a ‘system of science’ and an ‘economic system’) have become blurred to

86

Another well-known example Beck uses is the possibility of synergistic effects caused by the thousands of chemicals released into the environment. The point here is that although these substances have been tested and approved on an individual basis, the interaction between them, and the possible environmental and health consequences thereof, clearly lie beyond the scope of established scientific and policy routines.


53

the point where one can no longer tell whether it is still ‘science’ being practiced between the walls of a laboratory, or the design of a particular way of life (e.g. genetics). Beck calls this new form of politics, ‘sub-politics’. But the evolution towards sub-politics is a very ambiguous one (Beck 1992, p. 157): …This is a development of great ambivalence. It contains the opportunity to emancipate social practice from science through science; on the other hand it immunises socially prevailing ideologies and interested standpoints against enlightened scientific claims, and throws the door open to a feudalisation of scientific knowledge practice through economic and political interests and ‘new dogmas’…

Beck places great faith in the liberating potential of other forms of sub-political action. Through the influence of various centres of sub-politics – media publicity, citizens’ initiatives, new social movements, critical engineers and judges – operational decisions and production methods can be publicly denounced immediately, and firms can be forced with the cudgel of lost market shares to give a non-economic, discursive justification of their measures (Beck 1992, p. 120): …That is precisely where the future of democracy is being decided: Are we dependent in all the details of life and death issues on the judgments of experts, or will we win back the competence to make our own judgment through a culturally created perceptibility of the hazards. Is the only alternative still an authoritarian technocracy? Or is there a way to counter the incapacitation and expropriation of daily life in the civilisation of threat?...

In a sense, even a visit to the supermarket can then become a political act: through their purchasing behaviour, consumers can send strong signals on the acceptability of certain products. The belief in sub-politics as a seed of fundamental changes follows from Beck’s belief that minimising ecological and health hazards is the key to ameliorating a general feeling of insecurity, ‘existential insecurity’, that pervades current Western society87. Subpolitics then takes on the form of ‘life-politics’. Beck states that life-politics are an evident part of a number of new social movements. As collective disenchantment exhausts other sources of meaning, each of these movements strives to introduce new sources of meaning to life itself, thereby replacing ‘traditional’ ones like faith in technical progress and class consciousness. Life-politics and the new meanings it sustains, rooted in localism and in the political thrust of civil society, are able to reshape society from below, argues Beck, in a reflexive process of ‘modernisation undercutting modernisation’. Although life-politics are therefore clearly rooted in processes of individualisation, Beck looks upon these processes as something primarily positive: he argues that life-politics will enable individuals, in their drive towards finding and inventing new certainties against those of the risk society, to stage their own biographies through social networks. For Beck, the exciting and challenging aspect of the risk society is that it represents a whole new stage in the development of modernity; one which allows for the opening up of previously depoliticised realms of decision making to democratic scrutiny. States and critics of the risk society might want to return to a more comfortable and predictable pattern of politics of the (first) industrial society, but Beck leaves us with the strong impression that they will fail, even

87

This seems particularly close to our discussion of an appeal to sustainability at the level of the manifest image as a form of denunciation (cf. Section 2.1.1).

54

Chapter 1

threatening with the spectre of the ‘still undigested past of Germany’ (Beck 1992, pp. 223228): …A society attuned to minimising the problems is without preparation when the ‘future shock’ hits it. Under the influence of that shock, political apathy and cynicism can grow rapidly in the populace, and the already existing gap between the social structure and politics, or the political parties and the electorate, can widen rapidly. (…) The old coalition between insecurity and radicalism would be revived. The call for political leadership would once again resound ominously. (…) The hunger for order and reliability would revive the spectres of the past. (…) Ultimately, it could not be ruled out that the still undigested past of Germany might become a possible developmental option for the future although in a different form…

3.2.2


It will be clear from this brief summary that Beck’s thesis offers some potentially challenging elements to policy analysis with regard to sustainability. It is however not our intention here to provide a full-fledged commentary on this large body of sociological thought88. In our discussion, we will concentrate on the concepts of risk, justification, the role of the state apparatus in policy making (i.e. the ‘political system’ – parliament, government, political parties, administrations, etc.) and the role of scientific risk assessment in the theory of risk society. The concept of risk functions as a cornerstone of Beck’s theoretical construction. In Beck’s work, modernity is pictured as a huge structured beam of light that shoots through time and space – separating ‘enlightened’ areas from the ‘dark’ ones89 – suddenly hitting a mirror (a function reserved to the notion of ‘risk’) and becoming diffused over space, thus ‘enlightening’ the objects previously left in the dark90. The fundamental dynamic behind reflexivity is the process of modernisation (and the risks it produces) itself – when we make interventions into it, it exercises its own agency and reflects back on us – causing us to reflect, question, and become alienated from those institutions which were supposed to control these processes. Thus, in a nutshell, the logic is as follows: in the second phase of industrial society (risk society) individuals are freed from their unselfconscious immersion in traditional group determinations and are challenged to come to terms self-consciously

88

We will most notably not discuss the neo-Marxist critique that Beck’s focus on modern science and technology as risk-producers understates the realities of underlying economic processes (see e.g. O’Connor 1994, Pepper 1998). This critique in effect replaces one all-encompassing structural explanation of change (modern science and technology) by a functionalist one (the subsumption of society under capital). 89 Think of classical divisions between expert/layman; rational/irrational; subject/object; nature/culture; science/politics; fundamental research/applied research etc. 90 Beck therefore fundamentally rejects any speculation about ‘post’-modernity, ‘post’-industrial society, etc.; on the contrary, risk society to him represents a radicalisation of the project of modernity. To give an example: according to Beck, it is not the failure but the success of science that has dethroned the sciences; the critique that was previously reserved to those inside the scientific system, and which assured its unique position as science (the methodical scepticism institutionalised for instance in peer review procedures) is now being extended to other social arenas.


55

with their unmediated relation to society91. At the same time, the formerly latent dimensions of producer-driven industrial risks are brought into the open and can be apprehended for what they ‘really’ are, i.e. problems that are constructed in scientific terms, and at the same time as a new source of conflict and social formation. In all of this, the concept of risk is like a probe which permits over and over again to investigate the entire construction plan of modernity, as well as every individual speck of cement in the structure, for the potential of self-endangerment. Several questions can be raised by this approach. Admittedly, Beck was well ahead of his time in calling attention to the importance of the concept of risk and the practice of risk management as essential features of modern society. Developments since he first completed his work have confirmed his view of that importance, and there is little doubt that in the future the centrality of risk debates will steadily increase. But Beck of course wants to do much more when promoting risk as the central driving force behind the process of modernisation. Even leaving aside the question whether this truly is the case in an empirical sense, it does not seem a priori justified to us to narrow down the debate on technological choices to the risks they impose on society, let alone narrow the debate on technological risks to the terrifying ‘mega-risks’ (with potentially unforeseeable, irreversible consequences)92. According to Beck, these ‘mega-risks’ quite simply cannot be justified on any account; their expulsion from modernity is only more than logical. He therefore provides us with only one criterion to judge technological developments. This point of view omits several other criteria – an omission Beck nevertheless never deems worthy of further justification. What about the benefits of a risky activity? Can a ‘megarisk’ (no matter how improbable) never be justified, even if it would imply large benefits? Or what if two (or more) ‘mega-risks’ have to be balanced against each other (e.g. global warming vs. risks of nuclear power)? What about natural risks (e.g. some groups promote the future use of nuclear reactors for the desalination of sea water, in order to minimise the risks of water shortage in dry areas)? Even more fundamentally: why do discussions about ‘the good way of life’ or ‘the sustainable society’ necessarily have to be framed exclusively in terms of risk? Other questions can be raised, but the point we want to make here is that sustainability includes a much wider range of possible criteria for the judgment of technological developments, in different ‘worlds of justification’. As Beck’s selective account of the disrupting effects of potentially terrifying ‘mega-risks’ on modernity does not seem to be based on empirical findings, this leads us to speculate on other motivations,

91

Beck convincingly shows that the achievements of industrial society (the first phase of modernisation) fundamentally rested on ‘traditional’ forms of social organisation (the nuclear family with the embedded ‘standard’ biographies of men and women, standardisation of labour, etc.). Therefore, problems of industrial society where problems of modernity applied in a not-yet-modern society; whereas the problems of risk society arise out of modernity applied in a thoroughly modern society… 92 Simplified, Beck’s argument is as follows: a) society produces risks with unknown consequences; b) people, as a result of the built-in reflexivity of modernisation, will become aware of these risks (as they become carriers of a ‘logical’ programme of further modernisation); and c) people will reject these risks through sub-political action (a possibility which is opened up again by the inherent structure of the modernisation process). De Meere (1996) however takes issue with this ‘logical’ sequence, showing empirically that different interpretations of ‘mega-risks’ compete in public consciousness, and that therefore the outcome of this battle of interpretations is nowhere as determined as Beck suggests.

56

Chapter 1

most notably the need to promote his own substantive ethical ideal of how people ought to take up social responsibilities. We do not deny that incurring ‘mega-risks’ in light of a higher goal puts a particular strain on processes of justification93. But if society carefully (i.e. in a well-considered procedure) balances sustainability criteria, we see no reason why ‘mega-risks’ should be excluded a priori from these considerations. Returning to the issue of insurance as developed above, we found a good illustration of this point of view in the concluding section of a study on the evolution of civil responsibility and insurance schemes for nuclear installations (Degros 2004, pp. 170-172): …Le moteur principal de la création du régime spécifique de réparation des dommages nucléaires dans les années ‘60 nous semble avoir été le soutien du développement de la technologie nucléaire perçue comme un élément participant incontestablement au bien commun. Le souci de protéger les victimes, présent dès l’origine des régimes de responsabilité civile spécifique, a pris au fil du temps et des accidents une importance grandissante. (…) Cependant, le régime de réparation des dommages nucléaires reste fondamentalement un compromis entre les intérêts des exploitants et des victimes. Présenter ce régime, dans sa version révisé ou non, comme protégeant uniquement les victimes nous semble aller trop loin. (…) Un accord politique à long terme est nécessaire. Cette position devrait reposer sur un débat public profond et un processus de décision transparent…

Regarding the political implications of risk society, Beck’s analysis runs parallel with the broader theme of reflexive modernity. Since state regulation and modernity are inextricably linked it follows that the crisis of modernity will be reflected in a crisis of the institutions. Beck first notes some global social changes that contradict ‘classical’ conceptions of state and society. Using a form of intellectual stenography, these changes can be represented as follows. First and foremost, a polycentric society of large organisations has emerged, in which influence and political power pass into the hands of collective actors and can be acquired and exerted less and less by individuals. In addition, competing interest groups have multiplied, making impartial will-formation difficult. Furthermore, the growth of state bureaucracies and their functions fosters the domination of experts. Finally, apathetic masses have become alienated from the elites, who become independent oligarchies that paternalise voiceless citizens. All of this leads to a situation where public demands for higher levels of protection from risks have grown steadily, yet the (fiscal, ideological, etc.) capacity of the state apparatus to intervene decisively has diminished. By now, this diagnosis of contemporary society has become widely accepted by political analysts (see e.g. Habermas 1996; Holemans 1999; Gandy 1999 – Bobbio (1987) seems to have been another forerunner) and thus seems to provide a solid basis for further analysis. But Beck does not leave it at that and goes on to speculate about what this all means for the longer-term economic and political development of our society. He investigates three broad scenarios for the future (Beck 1992, pp. 224-235) – ‘back to industrial society’ (where the role of the state is first and foremost the promotion of capital accumulation by private enterprise in order to ensure that the economic cake is big enough before deciding on issues of fair distribution); ‘the democratisation of techno-economic development’ (an ecological variant of the welfare state) and ‘differential politics’ – with, as we have shown, a distinct preference for the last scenario. Again, Beck is ambiguous 93

One of the points we want to make in this dissertation is precisely that this type of risk should set into motion a particular, more stringent decision-making procedure.


57

when he presents these conclusions as if they merely arise from an observation of broad tendencies. In any case, it seems to us that the level of detail provided by him does not warrant the strong conclusions he is willing to draw. Furthermore, Beck seems to be among a slew of other ecology-oriented authors with not only a sceptical view of the role of states in the destruction of ecosystems within and beyond their borders, but also of statecentric analyses94. While acknowledging the limitations of state-centred analyses, we maintain with Eckersley (2004) that, given that the state is very likely here to stay for some time ‘whether we like it or not’, the increasing importance of structures of global governance should not prohibit an inquiry into the potential of states to promote sustainability. Thus, for our analysis, the explicit recognition of a series of structural constraints (pointed out by Beck and others) facing the state apparatus in late-modern societies seems to be an inescapable starting point. This however does not imply the determinism Beck seems ready to accept. It does imply that we should carefully investigate in greater detail, in the specific setting of this dissertation (i.e. energy policy in Belgium), the extent to which the public institutions, in their capacity for seeking legitimation and inducing action, are bound by structural requirements; and from this empirical observation, derive a desirable yet realistic form of sustainability governance. Beck has provided us with stimulating impulses to this reflection, sketching out the broad lines within which the discussion might evolve, but these do not allow us to jump to the conclusions. A final theme of risk society, the role of science, which pops up here and there in the text and receives one full chapter (“Science beyond Truth and Enlightenment?”), has perhaps so far in our discussion received less attention than its treatment by Beck warrants. For the sake of continuity in his chief thread of argumentation Beck is required to contend that science also becomes reflexive, because its monopoly on ‘truth’ is challenged. In line with the central tenet of risk society, this is not the result of a development exterior to the scientific system, but on the contrary of a ‘natural’ evolution of science in a very advanced stage of hyper-complexity. Again, Beck does not stand alone in his diagnosis that traditional scientific approaches to risk assessment face increasing difficulties when applied to the complex (i.e. global, long-term, potentially catastrophic, etc.) problems95. In response to these complex problems, new approaches to the production and utilisation of scientific knowledge have been advocated. Whether this plea for a new science goes under the name of ‘mode 2-science’ (Gibbons et al. 1994), ‘precautionary science’ (Stirling 1999a), or ‘post-normal science’ (Funtowicz 1993), all these new currents share the insight that scientific knowledge is, in essence, a social construct, and therefore the attention is increasingly directed towards the context(s) of application of scientific knowledge, rather

94

For an overview, see the introduction in Eckersley (2004, pp. 1-17). Many variants exist, but with ‘traditional’ risk assessment we mean a form of risk assessment rooted in positivistic methodologies, ranging from the application of statistical interference and probability based estimates to the use of complex models to predict the behaviour of systems. The ‘traditional’ approach is based on three fundamental presuppositions: the possibility of a realistic identification of all possible outcomes of a ‘risky’ decision, the possibility of an estimation of the magnitude of these outcomes, and the possibility of an estimation of the probabilities of these outcomes. 95

58

Chapter 1

than to its ‘truth’ in an absolute sense (a more profound discussion of constructivism is undertaken in section 3.3). This observation is particularly salient in a context where scientific rationality is confronted with political rationality, as it is more difficult to reach closure (on the validity of scientific results, the selection of relevant disciplines, etc.) in these circumstances (Vranckx 2000). But here also, Beck is willing to draw conclusions based on sweeping generalisations and a selective account of a few technologically induced risks. His remedy to the problems experienced by the ‘traditional’ scientific risk assessment is a science that 1. is focussed on fighting the causes of risks rather than the symptoms (e.g. replacing risk-producing industries by other less risky activities); 2. allows learning from practice (thus excluding developments with potential irreversible consequences, or making demands of infallibility on the people who have to put the science into practice – nuclear power is again excluded on this account, given its ‘unforgiving nature’ to human error); 3. avoids the risks of overspecialisation (the argument being that, the more science becomes subdivided in separate disciplines, the more each discipline will produce ‘side-effects’ which remain out of sight for that particular discipline, in turn feeding the demand for more science explicitly dealing with these ‘side-effects’, etc.). Again, the relevance for our research lies in the questioning of scientific practice in the context we wish to examine along the lines suggested by constructivist insights – thereby empirically checking to what extent Beck’s (and others) diagnosis holds true, and which lessons can be drawn from this analysis. While not denying that the solutions Beck suggest could be valid, or at least part of a wider strategy reconceiving the role of science in attaining sustainability, we should not accept them a priori, without careful empirical analysis. Thus, we retain from Beck’s analysis the attention for the structural changes modern society is undergoing. Our research therefore should take into account the strong points of his analysis, namely the attention for the structural strains on modern institutions responsible for risk management (including science, governments, administration, etc.), and the impact this has on the justification of technological choices, as starting points for further investigation. However, our contention is that the result of this process of justification in light of sustainability cannot and should not simply be predicted in a trivial way from structural changes.

3.3

(Social) constructivist studies of science and technology

Perhaps the best way of summarising our critique on the theory of risk society is by saying that, although Beck is correct in pointing out broad structural changes concerning ‘modernity’, ‘society’, ‘science’ and ‘technology’, the theory nevertheless remains rather abstract – i.e. it says little about particular political settings, particular technologies or


59

particular controversies over technological choices. Worse still is that the theory advances a strong ethical position (i.e. the rejection of ‘mega-risks’, the promotion of a particular scientific and political practice) as the result of a ‘simple’ empirical observation of structural macro-changes, thereby hiding the ‘construction work’ done in order to arrive at these conclusions (i.e. the singular focus on ‘mega-risks’, and hence the neglect of other sustainability issues). Boltanski and Thévenot’s ‘commonwealth model’ on the other hand transgresses the ‘dictate of structure’ by opening up the possibility of multiple schemes of justification to arrive at a conception of sustainability; but, as we have mentioned, the practical applicability of their theory also remains somewhat in suspense, particularly concerning the relationship between the need for justification and action in the ‘real’ world. After all, critics would ask, is a theoretical model primarily based on the search for consensus (within one commonwealth) and compromise (between commonwealths) as a source of stability in contemporary society (where, ranging from the decisions of expert commissions to the individual consumer in the supermarket, ‘politics’ have entered every aspect of social life) not a rather pious fiction? It is precisely on this point – empirically explaining (technological) choices without having recourse to transcendent schemes of justification – that a distinct theoretical perspective enters into view: the (social) construction of technology. In what follows, we summarise the conceptual framework underlying this perspective, and survey some of the main methodological and explanatory difficulties that arise from this approach. Again, our interest is not just criticism for the sake of criticism, but we also look for the lessons to be learned. 3.3.1

A brief guide to (social) constructivist studies of technology

Constructivist approaches are currently not only influential in both science studies and technology studies96, but in many other domains (for an overview, see Hacking 1999, p.1)97. The label ‘(social) constructivism of technology’ is used to refer to a variety of related, predominantly sociological approaches, most of which have applied ideas and concepts originally developed in the ‘Strong Programme’ of the sociology of knowledge (see e.g. Bloor 1976; Barnes 1977)98. The starting point of social constructivist technology studies

96

For literature reviews, see Vinck (1995), Williams and Edge (1996), and Pinch (1996). Hacking (1999, p. 48) makes a rigorous distinction between ‘constructivism’ (a theory in mathematics), ‘constructionalism’ (a philosophical project undertaken by Russell, Carnap, Quine, and others) and ‘constructionism’ (sociological and historical studies aiming to explore the social interaction or causal routes that led to the establishment of some ‘entity’). Hence, according to Hacking, we should rather be talking of ‘social constructionist studies of science and technology’. However, since Hacking (1999, p. 48) also notes that “…constructivists, constructionists and constructionalists live in different intellectual milieus…”, we believe there is no harm done in retaining the now widely-used ‘constructivism’ denomination also for the sociological studies in science and technology that we aim to explore. 98 The ‘Strong Programme’ (influenced by Wittgenstein’s notion of ‘language game’) tried to describe science and knowledge as a social convention, very much dependent on institutions and institutionalised practices for its continuation, including for instance the socio-political and religious context, the social regard for scientists, prevailing ideologies, etc. Proponents of the ‘Strong Programme’ did not claim (as is sometimes suggested by their critics) that ‘hard’ scientific facts could be explained or even predicted from a ‘soft’ constellation of social factors. The ‘strength’ of the ‘Strong Programme’ does not refer to the ‘hardness’ of causal explanation, but rather to the (at that time) rather radical novelty of the approach – i.e. questioning the common-sense interpretation of scientific inquiry as a ‘disinterested’ activity – and hence provoking strong reactions (Vranckx 2000). 97

60

Chapter 1

can be traced back to the mid-’80s (see Pinch and Bijker 1984, 1987; Bijker et al. 1987). From this seminal work has flowed a body of work that is both rich and diverse: since then, dozens of books and hundreds of articles, most of them socio-historical case studies of technological innovation and technological change, have appeared99. The term ‘social constructivism’ is sometimes used in a narrow sense, to refer to the ‘Social Construction of Technology’ (SCOT) approach that was outlined originally in Pinch and Bijker (1984, 1987) and Bijker et al. (1987) (for a related approach, see Woolgar 1991). In a broader sense, which will be used throughout this section, the term also includes the so-called ‘social shaping’ approaches (see e.g. MacKenzie and Wajcman 1985; MacKenzie 1990), and the ‘actor-network’ approach of Bruno Latour, Michel Callon, John Law, and their followers – see e.g. the seminal works by Callon (1987) and Latour (1987). Although there are different approaches in (social) constructivism, they all share a certain family resemblance. In what follows, we will try to summarise the basic concepts (indicated in italics), taking care however not to paint an overly simplified picture of the diversity of existing approaches100. Firstly, (social) constructivism includes a conception of technological development as a contingent process, involving heterogeneous factors. Accordingly, technological change cannot be analysed as following a fixed, unidirectional path, and cannot be explained by reference to economic laws or some inner technological ‘logic’. Rather, technological change is explained by reference to a number of technological controversies, disagreements, and difficulties, that involve different actors (individuals or groups that are capable of acting) or relevant social groups, which are groups of actors that share a common conceptual framework and common interests concerning the technology in question. Relevant social groups should be understood broadly, as any group who has played a role in the development of the technology in question (Bijker 1995, p. 45). These actors or groups engage in strategies to win from the opposition and to shape technology according to their own plan. Secondly, (social) constructivist approaches typically employ a principle of methodological symmetry, or methodological relativism (Pinch and Bijker 1987; Pels 1996). This principle, in its most common form, implies that the analyst remains impartial as to the ‘real’ properties of his/her object of analysis, e.g. technology. This implies, among other things, that the analyst does not evaluate any of the knowledge claims made by different social groups about the ‘real’ properties of the technology under study. This principle was originally formulated in the sociology of knowledge (Bloor 1976), where it was motivated by the idea that in a sociological explanation of claims to (scientific) knowledge, it is both possible and desirable to remain agnostic about any role of ‘reality’ (in the sense encountered in a common interpretation of empirical science – i.e. ‘the objective world out there’) in settling scientific controversies. Instead, the analyst should

99 To name but a few: MacKenzie’s (1990) account of missile accuracy; Bijker’s (1987, 1992, 1995) work on early bicycles, bakelite and fluorescent lighting; and Misa’s (1992) investigation of the manufacture of steel. 100 Our enumeration is mainly inspired by Brey (1997) and Klein and Kleinman (2002).


61

analyse putatively true and false claims symmetrically, interpreting them by reference to similar (social) factors. Such agnosticism is held to be desirable, and indeed possible, because the analyst is claimed to have no independent access to ‘reality’, and hence no independent way of evaluating knowledge claims of scientists and others. All one can do is to ‘follow the actors’: discerning the meaning of their self-understanding (taken at facevalue) revealed through their actions, whilst, as a social scientist, staying ‘impartial’101. Central to this technique is the idea that the only categories and lines of demarcation of importance are those consciously recognised by the actors, thereby avoiding the pitfall of ‘retrospective distortion’ (Klein and Kleinman 2002, p. 32). As a consequence of this principle, when applied to technology, the analyst will generally avoid making claims about the true nature of technology, including claims about the (in)operational character of artefacts, technological (in)efficiency, success or failure in technical change, the (ir)rationality of technological choices and procedures, technological progress, the real function or purpose of an artefact, and/or intrinsic effects of technology (Brey 1997). Because the analyst avoids reference to real properties of a technology, moreover, such properties cannot be invoked to explain technological change. For example, no reference should be made to the actual properties of an artefact in explaining its success, or indeed its sustainability. The outcome of the process of controversy and strategy mapping that surrounds technical change is the stabilisation of a technology, together with concomitant (‘coproduced’) social relations. Stabilisation of a technology implies that its contents are ‘black-boxed’, and are no longer a site for controversy. Its stabilised properties come to determine the way that the technology functions in society. Most (social) constructivists, including SCOT scholars, attribute the stabilisation of an artefact to an agreement or settlement between different social groups, which arrive at a similar interpretation of a technology, as the result of a series of controversies and negotiations. Technology is claimed by these (social) constructivists to have interpretive flexibility: it has no objective, fixed properties, but allows for different interpretations, not only of its functional and socio-cultural properties but also of its technical content, that is, the way it works. Technological artefacts are supposed to be sufficiently underdetermined to allow for multiple possible designs – so whatever the result of technological change, it could have been different. Facts about a technology are hence not objectively given by the technology itself, but are determined by the interpretations of relevant social groups. The rhetorical process of agreement on the true nature of a technology as the outcome of negotiation and social action is called closure. Technology is hence socially shaped or socially constructed: its properties are largely if not exclusively determined by the interpretive frameworks and negotiations of relevant social groups.

101

Technology is sometimes metaphorically described as a ‘text’ that is ‘read’ by different actors in different ways, and the task of the analyst is to analyse how the text of technology is ‘written’ by different actors, and how particular ‘readings’ of it come to prevail (Woolgar 1991; Grint and Woolgar 1995). The task of the analyst is not to select a particular ‘reading’ of a technology (i.e. his/her own reading), and present it as the ‘correct’ reading.

62

Chapter 1

The above broad characterisation of (social) constructivist technology studies obscures the fact that a variety of approaches exists, between which there are important differences. There have been various attempts at classifying different approaches within (social) constructivism (e.g. Woolgar 1991; Sismondo 1993; Grint and Woolgar 1995; Brey 1997; Bijker 2001). The following taxonomy of three (broad) approaches is based on these attempts. The most characteristic variety of (social) constructivism in technology studies may be called strong social constructivism. This approach is the one aligned most closely with the sociology of scientific knowledge, and includes the SCOT-approach. It vigorously upholds the principle of symmetry, and hence avoids all reference to the actual character of technology in its analyses. Strong social constructivism corresponds to what Grint and Woolgar (1995) call ‘post-essentialism’ and ‘the constitutive variant of anti-essentialism’. Technological change is to be explained by reference to social practices, particularly by reference to processes of interpretation, negotiation, and closure by different actors and social groups. Technology is a genuine social construction, that is, a stabilised technology can only be explained by reference to the social elements (including other socially constructed entities) that have produced its stabilisation. No ‘properties’, ‘powers’, or ‘effects’ can be attributed to technologies themselves. Mild social constructivism is the label that will be used to characterise more moderate approaches, that sometimes go under the name of ‘social shaping’ approaches (e.g. MacKenzie and Wajcman 1985; MacKenzie 1990). Social shaping approaches retain conventional distinctions, between the social and the natural, and between the social and the technical, and study the way in which social factors shape technology. They do not reject a role for non-social factors in technological change, and are also willing to attribute properties and effects to technology, although these properties and effects are usually claimed to be defined relative to a particular social context. Because technologies are socially shaped, these properties and effects are in large part social properties and social effects, that can be attributed to social biases or politics ‘built into’ or ‘embodied by’ these technologies. Actor-network theory, sometimes simply called ‘constructivism’ (without the ‘social’), is a third influential approach102. It studies stabilisation processes of technical and scientific objects as a result of the building of actor networks, which are networks of human actors and natural and technical phenomena. Actor-network theorists employ a principle of generalised symmetry, according to which any element (social, natural, or technical) in a heterogeneous network of entities that participate in the stabilisation of a technology has a similar explanatory role (Callon 1987; Latour 1987). Strong social constructivism is

102

Actually, in 1986, Latour dropped the adjective ‘social’ in the second edition of his book (first published in 1979, co-authored with Woolgar) “Laboratory Life. The Social Construction of Scientific Facts” (Latour 2003b). This marked the turn in his thinking towards the ‘actor-network’ model. Therefore, when referring to the overarching category of (social) constructivist studies (including the actor-network approach), the ‘social’ denominator has been put between brackets.


63

criticised for giving special preference to social elements, such as social groups and interpretation processes, on which its explanations are based; whereas natural or technical elements, such as natural forces and technical devices, are prohibited from being explanatory elements in explanations. Actor-network theory also allows for technical devices and natural forces to be actors (or ‘actants’) in networks through which technical or scientific objects are stabilised. The network allows for making links between present, localised interactions and distant and/or historical ones – e.g. a household appliance is accompanied by instructions on how to use it (incorporating an image of the ‘normal user’), drawn up in some engineering firm far away from the present circumstances of use, but nevertheless able to ‘act’ through the appliance. By an analysis of actor networks, any entity can be shown to be a post hoc construction, but entities are not normally socially constructed, because stabilisation is not the result only of social factors. 3.3.2


Should we, as Negri (1993) seems prepared to do in his critique of Boltanski and Thévenot’s ‘commonwealth model’, rejoice in this ‘radically materialistic’ turn in the approach to processes of (technological) change? Negri’s enthusiasm can be explained by his assessment of Boltanksi and Thévenot’s theory of justification as being caught in an ambiguous (and even impossible) middle position between a ‘transcendental’ and a ‘materialistic’ point of view – thereby, in his view, rendering the theory utterly ineffective (“…ça ne fonctionne pas…”, he exclaims empathically). According to him, one has to choose sides; and in contemporary society, where (sub-)politics and (the potential for) the exertion of power are everywhere, pretending that consensus and stability in society can still be guaranteed by making reference to ‘transcendental orders of justification’ is, quite simply, an empty pretension. The only alternative Negri (1993) sees is the perspective of the actor-network theory, which he approvingly describes as follows103: …Cette direction de recherche nous l’avons qualifiée comme franchement matérialiste. Ce que je veux dire par là, c’est que dans la théorie des réseaux (surtout telle qu’elle est définie théoriquement et développée au niveau de l’enquête empirique de la sociologie de Callon et de Latour), le transcendantal n’a pas sa place. Il n’a vraiment pas de place. Le relativisme est absolu. (…) Le problème de la légitimation s’évanouit donc: ces acteurs qui interfèrent de façon machiavélique dans la construction de la vérité ne sont pas des imposteurs; ils mettent en place des rapports de force. La vérité est fruit de la lutte. Aucune vérité n’est pourvue de charisme objectif. La vérité est totalement subordonnée aux rapports, aux tensions, aux tendances qui déterminent les acteurs subjectifs (individuels et/ou collectifs, toujours singuliers). Nous existons à l’intérieur de cette relation, nous existons collectivement, nous existons de manière complexe, nous produisons donc nous-mêmes les objets auxquels nous nous confrontons. Il n’y a donc rien, et surtout rien de transcendant, qui soit au-dessus ou simplement en dehors de l’activité constructive humaine, qui se déroule entre réseaux…104

103

This should probably not be seen as a choice among the different variants of (social) constructivism outlined in section 3.3, but rather as a result of a greater familiarity with French literature (Latour, Callon, etc.). 104 Based on Latour’s recent work (e.g. 2000, 2003a & b, 2004a) it is doubtful that he would agree with this assessment of ‘radical’ materialism. However, it remains true that Latour, and more generally the actornetwork perspective, is the most materialistic representative of the general class of (social) constructivist theories of science and technology.

64

Chapter 1

A rather dim prospect indeed for a notion such as ‘sustainability’, which, whatever else is added in terms of interpretation, at the very least implies the stability over time of something we value105… But should we really choose sides (as Negri suggests)? In what follows, we will argue that this is not the case. We will start with an investigation of the methodological and explanatory flaws of (social) constructivist theories of technology – arguing that when one seriously thinks through the methodological prescriptions outlined above, the (social) construction of technology can equally be accused of theoretical ambiguities and impossibilities. A singular ‘essentialist’ focus on explaining ‘action’ only in the terms of those participating in the action (e.g. in the various forms of (social) constructivism) is as self-defeating as a singular focus on ‘ideal spheres of justification’ (e.g. in the commonwealth model) or ‘structure’ (e.g. in the theory of risk society). On a more ‘constructive’ note then, we consider in the next section a first possible reconciliation of the different perspectives we encountered so far, and give a first outline of our own broadly conceived constructivist position (Section 3.4). A suitable entry point for showing the contradictions of theories of (social) construction is the programmatic ‘principle of symmetry’. As stated, Latour (1987) has extended this principle by abolishing the difference between ‘human’ and ‘non-human’ actors altogether: both are ‘actants’ (they provoke change) linked through ‘hybrid’ networks (hybrid, because traditional distinctions between nature/culture, nature/artefact, etc. no longer hold). Latour (1993, p. 113) claims that this position is “…more modest, but more empirical…”, which can be read as an explicit criticism on the SCOT approach with its focus on the ‘social’ factors. Let us, perhaps a bit ill-willed, interpret this claim as a pretence to the exclusive rights of the correct way of conducting empirically-oriented social studies of science or technology106. Since there is no a priori limit to this extension of symmetry as a methodological requirement, one can always add another ‘hyper-symmetry’ to an already existing one (van Brakel 1998, p. 59)107. The principle of symmetry is advanced precisely because one wants to explain action as truthfully as possible – i.e. by erasing the ‘traces of contamination’ left by the observer as remnants of his/her own conceptual scheme. 105

Although we have not found any examples in the literature we surveyed, we suspect that the notion of sustainability would, due to its ‘interpretative plasticity’, easily lend itself to processes of ‘rhetorical closure’, described in Bijker et al. (1987) as one of the possible mechanisms of reaching agreement among actors. 106 Latour himself has later explicitly stated that actor-network theory and its methodological admonitions should be interpreted ‘negatively’, i.e. they do no tell us what to do, but rather what not to do (Latour 2004b). We will clarify this further in section 3.4. Our motivation here is mainly to present the inherent tension in its utmost consequences, i.e. when interpreting them in a ‘positive’ sense. We are well-aware that (social) constructivist scholars over the years have sought to lessen the tension (thus demonstrating that they are aware of it) through a continuous critical re-interpretation or reworking of their original basic assumptions. However, it is precisely by bringing the inherent tension very sharply into focus, that we can assess whether (social) constructivist scholars have been successful in these attempts, and to what extent we can accept some of their concepts for the purposes of our theoretical investigation. 107 The ‘politics of symmetry’ are nicely described by Pels (1996, p. 279): “…Without doubt, it is this democratic accessibility and transferability of the method of symmetry which defines its paramount attractions. (…) Its radicalism is methodical, and can therefore be extended indefinitely; if properly socialised, everybody can play the game. This turns symmetry and impartiality into tools-for-all-users, while anything which can be construed as a form of asymmetry – any kind of opposition, polarity, demarcation, dualism or categorical boundary-line – is in principle vulnerable to symmetrical deconstruction…”.


65

However, when claiming that an extended principle of symmetry is ‘more empirical’, the observer has to have some implicit measure of objectivity towards his/her own research, perhaps acquired through previous research experiences, but in any case not stemming from the phenomenon he/she is investigating at that moment. Thinking through the consequences of a generalised methodological principle of symmetry, (social) constructivists are required to make an ultimate and ultimately undesirable recourse to either value-free ‘pre-modern’ language or value-free ‘positivistic’ research of social facts (without any form of interpretation), in terms of ‘superior interests’, ‘force’, ‘power’, etc. attributed to all the ‘actants’ (Boltanski and Thévenot 1992, p. 416): …Le recours à un équivalent général est nécessaire parce que le relativisme le plus conséquent ne peut se soustraire aux contraintes qui pèsent sur la relativisation sans tomber dans un nihilisme radical et autodestructeur, puisqu’il se condamnerait lui-même au silence, qui n’est jamais complètement réalisé dans le nihilisme philosophique ou politique…

This inherent tension inhabiting (social) constructivist studies of science and technology is not just a matter of academic fiddling, but can be demonstrated by a close reading of some key texts. The tendency towards ‘pre-modern’ language becomes apparent for instance in the commitment towards a position of so-called ‘co-construction’ – a variant found in SCOT literature (e.g. Bijker 1995, Fujimura 1996). Proponents of ‘coconstruction’ hold that one should take the common evolution of society and technology (metaphorically described as a ‘seamless web’) as a unit of analysis, in order to move beyond the traditional analytical difficulties of separating the ‘contextual factors’ from the ‘internal factors’. But of course, such a position in its utmost consequences makes disaggregation and, in the end, analysis in itself impossible. This is not to deny the idea that society and technology are engaged in a reciprocal process of change, but quite simply that, in order to perform any analysis at all, the analyst is required to somehow artificially stop the process of interaction and ascertain, at least tentatively, what is affecting what. Thus, the methodological admonition that society and technology should be studied as a ‘seamless web’ in our view does not add anything in terms of a positive methodological guideline, but rather serves as a statutory reminder to the reader of (social) constructivist accounts of technological change that ‘the map is not the territory’. At the other extreme – the tendency towards ‘positivistic’ explanation – (social) constructivism has, perhaps not surprisingly, been submitted to what we might call a ‘structural critique’ (Klein and Kleinman 2002) and a ‘critique of justification’ (Winner 1993)108. Over the years, these criticisms on (social) constructivist theories of technology have crystallised around the following major theoretical issues: the relevant social groups, the principle of symmetry, the role of interpretation in technological controversies, stabilisation and/or closure of technological controversies, and the wider social context (political influence, power struggles, deep-seated cultural values, etc.). In what follows, we focus on strong social constructivism (the SCOT approach), as we will try to demonstrate that this ‘strongest’ variant of social constructivism, with its most strict anti-essentialist appeal, cannot do without the notions of ‘structure’ and ‘justification’ in order to explain technological 108

Both are of course ultimately related, but for the purposes of clarity, we will try to separate them somewhat artificially.

66

Chapter 1

change adequately109. SCOT scholars have attempted to add notions of structure and justification to their analyses, sometimes by adapting previous methodologies; but, in our view, these responses have only succeeded in causing confusion. Again, it is not our intention here to engage in a discussion involving all the themes mentioned above; for our purposes, a few selected examples will suffice110. The ‘structural critique’ can be resumed by Klein and Kleinman’s (2002, p. 35) claim that: …an adequate understanding of the limits of interpretive flexibility, stabilisation, and closure requires attention to power asymmetry. (…) Societies are structures around power asymmetries and the bases of these can be found at a level above that of the interactions of the groups themselves. The capacities of relevant social groups and actors within them, furthermore, are shaped by their structural characteristics…

In exploring how SCOT scholars have responded to this critique, it is useful to make a distinction between those parts of the structural critique that address the scope of most SCOT studies (i.e. these studies explain technological change at the ‘micro-level’ of particular technological designs) and those that take issue with the adequacy and explanatory power of the methodological basic assumptions themselves. An example of the former form of critique includes the neglect for ‘absent’ social groups (including the (structural) reasons of their absence – i.e. ‘the power to exclude’), rules of interaction in forums of design and decision making, or the question of how the relevant social groups with their shared meanings came into being in the first place. We agree with Brey (1997) that these criticisms can be fended off reasonably well by referring to the necessary delineations of micro-action research, and furthermore, that there is no inherent (methodological) reason why SCOT concepts could not be applied to the interpretive frameworks of groups which did not have an influence on the shaping of the technology in question. The second form of critique goes deeper, arguing that SCOT cannot explain technological change adequately, at least not with their preferred ‘toolbox’. Klein and Kleinman (2002) for instance rightly notice that SCOT scholars not always restrict themselves to the actors’ categories in presenting their empirical research, and as a consequence accuse them of either ignoring concepts like ‘power’ or deploying them in an ad hoc fashion111. Remember that, on account of a strict interpretation of the social constructivist programme, power is something that is to be explained as a result of social action (the explanandum), rather than an explanation in itself (the explanans). To be fair, the original outline of the SCOT approach (Pinch and Bijker 1984, 1987) already acknowledged the existence of a certain ‘structural background’, although at that time it

109

The other approaches (‘social shaping’ and ‘actor-network’) have less difficulties in accommodating ‘structural’ elements, since they do not grant exclusive rights to social factors in the explanation of (technological) change. However, for actor-network theorists, ‘structures’ can never be invoked as an explanatory variable exclusively by the analyst; they have to be revealed through the actions of the ‘actants’. 110 For this, we refer the reader to Klein and Kleinman (2002). 111 They give the following examples taken from Bijker (1995): use of the ‘capital intensive’ and ‘oligopolistic market’ as an explanatory variable in the discussion of the development of fluorescent light (p. 200); a recourse to ‘a strong position owing to the possession of patents’ (p. 201); ‘economic power’ of a group (without defining the term) (p. 200); etc. Furthermore, Bijker (1995, p. 49) himself admits some of the categories he uses are exclusively analysts’ categories.


67

remained rather invisible in analyses (Klein and Kleinman 2002, p. 30). Since the original presentation, one major systematic attempt to counter the ‘structural critique’ has been made. To the four foundational concepts, Bijker (1995) added the concept of the ‘technological frame’112. This is defined as the shared cognitive frame between different relevant social groups and constituting members’ common interpretation of an artefact. Like a paradigm – in the sense employed by Kuhn (1970) – a technological frame can include goals, key problems, current theories, rules of thumb, testing procedures, and exemplary artefacts that, tacitly or explicitly, structure group members’ thinking, problem solving, strategy formation and design activities (Bijker 1995, p. 192). Bijker (1995, p. 263) correspondingly argues for a semiotic conception of power, according to which power is …the apparent order of taken-for-granted categories of existence, as they are fixed and represented in technological frames…

But of course, by stating this, Bijker is attempting to ascend a very slippery slope indeed. What started out as an explicitly agency-centred approach risks evolving towards a more structural form of analysis, without however fully admitting that this is the case. Social actors now suddenly act (at least partly) according to the guidelines laid down in a specific technological frame (which they themselves might not even recognise as such), which stabilises the interpretation of technological change in semiotic structures or categories. Notice however that this reference to semiotic structure opens a wide door towards functionalist or structuralist accounts of technological change, which is probably not what Bijker has in mind. Through all of this conceptual juggling, he quite literally risks losing the ‘subject’ of his research… Perhaps anticipating the critique offered here, Bijker (2002, p. 368) – in a recent defence of the SCOT approach – urges researchers to take his conceptual framework in the right spirit: …SCOT is not a recipe or ‘a simplistic rulebook’. A theoretical approach can never be a fool-proof datamining machine: it will never make the researcher’s craft superfluous, nor guard the researcher against errors of using wrong sources, or of not finding the right sources. The implication is that theoretical frameworks typically are not used without modification and adaptation by the researcher (perhaps with the exception of brief essays and exercises in undergraduate teaching)…

Indeed a good piece of advice, which by the way applies to the deployment of any set of concepts to draw up a portrait of an essentially infinite world. The point is not that concepts should be used carefully (which seems evident to us); but rather that concepts should serve the purpose they are destined for – i.e. better explanation and interpretation. It seems to us that Bijker, by stretching his concepts to an almost tautological point, is more concerned with saving his original conceptual framework rather than explanation/

112

Other examples of SCOT-related analyses that show a much larger attention for structural considerations include Hughes (1989) analysis of ‘large technological systems’ and Feenberg’s (1999) attempt to develop a ‘critical theory of technology’, focusing on power asymmetries and macro-sociological concepts such as class and culture.

68

Chapter 1

interpretation of technological change113. If ‘sensitive use’ of the SCOT concepts is required, it is not clear why we might not choose to abandon (part of) them altogether, provided that an alternative set of concepts assures equally good or better empirical results. Therefore, for reasons of analytic clarity, it seems more useful to us to keep the notion of ‘structure’ as an explanatory variable114. Essentially the same critique can be aimed at the difficulties experienced by SCOT scholars to include any form of justification or legitimation (by the social actors as well as by the observer) in their work. To SCOT scholars, technology development is a process in which multiple groups, each embodying a specific interpretation of an artefact, ‘negotiate’ over its design (with different social groups ‘seeing’ quite different objects). Development and negotiation continues until all involved groups come to a consensus that the common artefact ‘works’. Thus, somehow a final decision – or at least a cessation of further decision – occurs. This is called the ‘closure’, and Pinch and Bijker (1987) give us some examples of what they see as closure mechanisms: a ‘rhetorical closure’ (a simple declaration that no further problems exist and no additional design is necessary); ‘closure by redefinition’ (when unresolved problems are redefined so that they no longer pose a problem); or ‘consensus’ (an agreement on the working of the artefact). But in the end, SCOT scholars are simply unable to explain what makes a certain form of closure acceptable to the actors involved. From the conclusions of typical SCOT accounts, one understands that the technology ‘worked’ for the relevant social groups. But what is lacking is an explanation of how differences were resolved (or suppressed). If closure is merely ‘rhetorical’, why do actors recognising this closure as such (remember: the analyst can only use actors’ categories) nevertheless accept it? If closure takes on the form of a ‘consensus’, on what grounds is this consensus then accepted as ‘fair’ by the relevant actors? Furthermore, how is the SCOT analyst able to distinguish a mere ‘rhetorical’ closure from a closure by ‘consensus’? Again, it seems to us that the SCOT approach can only be saved by ‘structural’ considerations in the case of ‘rhetorical closure’ (e.g. the ‘powerful’ imposition of one group’s interpretation on others with only a slight rhetorical repainting, the ‘force’ of cultural presuppositions bearing on the process of closure, etc.), or by considerations of ‘justification’ in the case of consensus (e.g. Boltanski and Thévenot’s account of bargaining and compromise under section 3.1.4). The ‘critique of justification’ has also been aimed at the role of the analyst in SCOT theory, with the philosopher of technology Winner (1993) as one of the most influential exponents. According to Winner, one of the most important weaknesses of (social) constructivist studies of technology is its lack of concern about the evaluative 113

If interpretive frameworks of actors are expanded to incorporate the notion of structure (somehow working through the subconscious of the social group), the concept loses its original analytical powers of distinction and risks becoming too protean to be of any further (analytic) value. 114 Keeping in mind however the ‘dual nature’ of structures: social actors can provoke change by drawing upon resources provided by structures (e.g. interpretive frameworks, power, material means, legitimacy, etc.), but on the other hand, structures can only be maintained because they are reproduced constantly through social action (Giddens 1979, Vandenabeele 1999).


69

consequences of its own analyses. He maintains that one of the more important aims of social studies of science and technology should be to provide a ‘standpoint’ or ‘core of moral concerns’ from which to criticise or oppose any particular pattern of technological or scientific development (Winner 1993)115. The philosopher Giere (1993, p. 109) has also concluded that SCOT scholars avoid these larger issues: …Social constructivists have had precious little to say about science or technology policy. I suspect this silence is not an accident. If one takes seriously the position that science and technology are social constructs, the only policy advice one can give is to improve one’s use of the rhetoric of science and technology to persuade others of one’s point of view and to build cohesive social networks…

The ‘critique of justification’ has been countered by SCOT scholars by stating that the plea for relativism in social studies of science and technology should be understood as a methodological form of relativism, and therefore does not imply a normative or political form of relativism (Aibar 1996). Thus, according to Aibar, methodological relativism does not imply that all socio-technical ensembles are equally right to everyone, or that one cannot discriminate between different ensembles in regard to specific goals. Many SCOT scholars argue that their analyses can result in a kind of politics, even when they obey the principle of symmetry. Such a politics does not require that the analyst make evaluative statements or prescribe courses of action. Because the subject matter of SCOT studies are rather unique events, where the (non-reproducible) interplay of scientific insights, material means and social circumstances have enabled a certain breakthrough in science or technology, an extrapolation towards political conclusion is in the eyes of SCOT scholars even ‘risky’. If anything, the political agenda of SCOT studies should be to show (Bijker 1995, p. 280) …the malleability of technology, the possibility for choice, the basic insight that things could have been otherwise…

In this regard, it is also interesting to mention the response of Elam (1994) to Winner’s criticism. Elam argues that adherence to ‘the truth’, even in the name of a ‘politically correct’ analysis, goes against the foundations of liberal politics. According to Elam, liberal politics requires one to refrain from enforcing one’s view on others, and hence, from presenting one’s view as the only possible ‘true’ interpretation116. Thus, Elam holds that the very denial that any view qualifies as ‘true’ or as superior to other views is in the interest of protecting fundamental liberties. Even setting aside the fact that there is a difference between presenting a statement as true and a dogmatic adherence to that statement to the point of refusing any discussion (only the latter position could, in our view, conflict with fundamental liberties), this discussion clearly reveals that the adherence to the ‘interpretative flexibility’ of technology, so vigorously defended by SCOT scholars,

115

Winner is particularly known for his thesis that ‘artefacts have politics’. This thesis was first expounded in an influential article called “Do artefacts have politics?” (Winner 1980), where he provides an account of a technological design process (in this case bridges to Long Island, New York) that promoted the embodiment of the values of the dominant and privileged social classes (the bridges were designed so that buses – the transport of the lower classes – could not get through the overpasses). Winner’s example has however instigated a profound academic controversy, see e.g. Joerges (1999) and Woolgar and Cooper (1999) for criticisms. 116 For a similar argument, see Woolgar (1993).

70

Chapter 1

in fact reflects a clearly politically motivated choice, open to justification in light of (a particular interpretation of) the higher principle of political liberalism. Recently, the philosopher of science Hacking (1999) has done a great deal of clarification in this muddled topic. Hacking understood that the reason these disputes around constructivism trigger so much passion is that they are fundamentally political: they seem to talk about epistemology but they are really about how people should go about living together (i.e. politics in a broad sense). For purposes of clarification, Hacking (1999, p. 6) uses a simple scale to classify the interpretation of various enunciations of the sentence ‘X should be taken as constructed’117: (0) In the present state of affairs, X is taken for granted; X appears to be inevitable. Actually, this level is no assumption or presupposition concerning X; it rather states a precondition for a constructivist thesis about X. For instance in the above introduction to (social) constructivist theories of technology, we mentioned how these theories were developed in the first place to counter the prevalent model of technological determinism. Technological determinism thus occupies this levelzero position for constructivist theories of technology to take issue with, (1) X need not have existed, or need not be at all as it is. X, or X as it is as present, is not determined by the nature of things, it is not inevitable. This is the position taken by many SCOT scholars described above when explaining the political relevance of their studies. In this case, the political commitment of the researcher is limited to a ‘historical’ stance: X could have been otherwise – X is not ‘good’ or ‘bad’, but simply contingent; (2) X is quite bad as it is. In the above discussion, this position is taken by Winner, who argues for a break with a methodological principle of symmetry upheld by SCOT scholars on three accounts: by privileging some of the effects of technologies over other because they are considered to be more politically relevant (most of the time the effects of the technology in question on the most disadvantaged groups in society), by relating these effects to a definite cause in the design history of the technology, and by making evaluative statements about the political significance of these effects. This position reflects a political commitment of ‘unmasking’ (undermining a particular technological development by showing its ‘true’ – e.g. ideological – purpose) or ‘reforming’ (actually presenting an alternative, a modification of certain aspects of the technology – typically by subscribing to the emancipatory project of the most disadvantaged social groups); (3) We would be much better off if X were done away with, or at least radically transformed. Although we have not explicitly dealt with this position in our overview so far, the neo-Marxist view of the ecological crisis as an exponent of the 117

Hacking does not focus only on technology; instead, his work is based on a broad literature review of all articles with ‘the social construction of…’ or ‘constructing…’ in their title. This alphabetic list spans a wide field of study, ranging from ‘authorship’, over ‘nature’, ‘quarks’, ‘technological systems’, all the way to ‘Zulu nationalism’ – and much more (see Hacking 1999, p. 1).


71

second fundamental contradiction of capitalism (see e.g. O’Connor 1994, 1998)118, leading to its inevitable downfall119, falls into this last category. This position can only lead to a ‘rebellious’ or a ‘revolutionary’ commitment, the difference between the former and the latter being the willingness to discuss ideas further or rather take recourse to action… To this, Latour (2003b, 2004a) adds the politics of those who he calls ‘naturalists’, namely those for whom X exists as a permanent fixture of ‘nature’ (‘X is the way it is, period’). This is meant to include all those who claim the indisputable nature of something in order to define a common world: it is already ‘made’, and remains off limits for any political process. According to our reading, Beck’s theory of risk society would fall into this category, albeit in the somewhat different sense of politics being determined by a powerful undercurrent of reflexive ecological modernisation. However, as we have shown, it is not always clear whether Beck has written an empirical study rather than an accusatory pamphlet (to put things a bit bluntly – in section 3.2 we have clarified our point of view with somewhat more nuance). Therefore, once this objective cloak is discarded – doing a bit of ‘unmasking’ ourselves – Beck can just as well be classified with Hacking’s level-2 constructivists.

3.4

Constructivism as a (meta-)theoretical framework

Perhaps the time has come for some form of preliminary conclusion to these (meta-) theoretical reflections – which, after all, are only made in the margins of the research project we set out for ourselves. So far, we have only been concerned with possible ways of interpreting (technological) change, ways of conceptualisation, classification, etc., but not however with the ‘reality’ of the debate on nuclear power and its role in sustainable development. In doing so, we have untied some conceptual knots and are left with some basic threads of thinking; our aim is now to start the difficult work of knotting some of these threads back together in order to provide the weft of the richer fabric we aim to weave in the remainder of this dissertation. We would like to describe this basic weft as constructivist, both in a ‘broad’ and more ‘narrow’ sense. By this we mean that 118

In O’Connor’s ecosocialist theory, the first contradiction of capitalism refers to the ‘classical’ contradiction between social production and private appropriation (on the demand side), while the second refers to ‘the conditions of production’ (on the supply side), which O’Connor takes to be nature, labour and infrastructure. Given the expansionary dynamic of capitalism and the limited supply character of the conditions of production, he reasons that we can expect the costs of production to increase over time. This is exacerbated by the demands of labour, environment and welfare movements to improve working conditions, protect the environment and improve social infrastructure. Thus, for O’Connor, the ‘limits to growth’ will not appear as physical shortages but rather as higher costs. The contradiction then arises from capital’s standard response to the risk of losing profits: to externalise costs. Yet such a response only serves to further reduce or undermine the profitability of the conditions of production and thereby raise the average costs of production. 119 This is the logical consequence of such highly functionalist analyses of ties between the state, society and the economy, tending towards an overly deterministic understanding of state/economy relations. That is, functionalist claims can be very difficult to prove or disprove, since any policy output can be explained as promoting the accumulation imperative of capitalist society. Short of catastrophic collapse of the entire system, there seem to be no other explanatory variables available in such system analytic frameworks (Eckersley 2004, pp. 60-61).

72

Chapter 1

constructivist insights apply to both the ‘meta-theoretical’ level (i.e. ‘theorising about theories’ – where the largest part of our discussion has taken place so far) as to the ‘theoretical’ level (i.e. the actual conceptual framework or theory we consider most adequate to describe the ‘reality’ of the subject matter of our research). Let us clarify things a bit further… Of the meta-theoretical level, we will not say too much, since that would take us into a hotly debated domain of epistemology. It would be feckless to address such topics here, since our general purpose is not to further the development of a theory of knowledge. We will limit ourselves to some reflections that we deem to be absolutely necessary for a better understanding of our position120. On this level, constructivism should (at least) be understood as an attitude of ‘de-essentialisation’. De-essentialisation does not lead to methodological anti-essentialism or relativism, in the sense that SCOT scholars employ these terms. De-essentialisation simply means that one should refrain from positing concepts as ‘universally valid’, ‘a-historical’, etc. by laying down their cognitive content in exact context-free criteria. In our opinion, this fundamental insight enables to take the sting out of a lot (but not all) of the inter-theoretical clashes we have tried to expose above. Let us demonstrate this with two examples taken from the above sections. We already gave the example of the ‘principle of symmetry’, which, when interpreted in an essentialist way, leads to fundamental difficulties. However, if the ‘principle of symmetry’ is – on a less highbrow note – taken to mean that, when studying technological, scientific or other developments, we should not a priori limit our search for determining factors only to the ‘internal’ ones (e.g. performance, efficiency, cost, truth, etc.), or, conversely, limit our attention to the ‘social factors’ (e.g. power, institutions, etc.), then we believe that a lot more observers (including us) could agree to this ‘de-essentialised’ version of the principle of symmetry. A symmetrical treatment of technological change thus implies that ‘nature’ (or ‘objects’, ‘matters of fact’, etc.) and ‘culture’ (or ‘values’, ‘social habits’, etc.) are entangled in numerous imbroglios, which should not be a priori disentangled by an analyst/observer into ‘neat’ prefabricated categories. Essentially the same observation holds for the alleged ‘transcendentalism’ of Boltanski and Thévenot’s commonwealth model121. Let us again attempt to give a less highbrow approach to the apparently irreconcilable theoretical positions. A commonsense interpretation of ‘transcendence’ holds that something is transcendent when it ‘exceeds’ something else, e.g. in expressions like ‘he transcends himself’ or ‘this transcends the human mind’. Transcendence manifests itself in Boltanski and Thévenot’s ‘commonwealth model’ through the search for ‘the

120

Our meta-theoretical stance is largely influenced by van Brakel (1998) – an essay we can warmly recommend. Van Brakel (1998, p. 13) by the way calls himself a ‘realist-with-a-small-r’ (as opposed to ‘Realism-with-a-big-R’, which states that sciences produces knowledge that corresponds with how ‘the world out there’ truly is), mainly because he does not want his position to be confused with “…social constructionism or anti-realism as understood by alarmed scientists involved in the science wars…” (i.e. roughly meaning that there is no such thing as ‘objectivity’, that ‘everything is social’). Our point of view upon reading van Brakel (1998) is that his position might as well be called ‘constructivist’, without doing injustice to the above disclaimer he himself adds. For an enlightening discussion of the ‘realism vs. anti-realism’ debate in the philosophy of science, see also Burms and De Dijn (1990, pp. 61-79). 121 E.g. by Negri (1993); accusations of ‘a-historicism’ can also be found in Dodier (1993) and Jacobs (1996).


73

common world’ in situated interactions between people, which is driven by something they cannot fully grasp (i.e. the superior common principles). As we have explained in section 3.1, this should not only be understood as something positive; transcendence also creates a fundamental uncertainty – e.g. an agreement within one ‘world’ can always be denounced by an orientation towards a superior common principle from another ‘world’. Furthermore, people can simply refuse or be forced to deny the transcendent call from other ‘worlds’, by sticking to familiar arrangements, strictly abiding to rules laid down within one ‘world’, etc. Keeping this in mind, a close reading of Negri’s (1993) critique shows there is something odd about his arguments. Now it becomes quite apparent that he does not object to transcendence as such, but rather that he believes that Boltanski and Thévenot’s model does not allow for enough transcendence (the same can be said of Latour)122! Indeed, Negri’s fundamental objection to the commonwealth model lies in the fact that people are limited in their potential for action to the six commonwealths and the order they impose. The point is that, according to Negri, transcendence in the ‘commonwealth model’ is too well-defined: …Boltanski et Thévenot, en effet, ne peuvent comprendre qu’une cité désobéisse que dans la mesure où son projet est passé par le compromis. Mais supposons qu’une cité qui désobéit ne veuille rien savoir du processus de compromis. Supposons qu’ils appellent ‘valeur’ ou encore ‘grandeur’ le fait de ne pas se soumettre au processus de compromis. Qu’est-ce qui peut nous empêcher de considérer que la négation absolue de la valeur et du transcendantal du compromis (n’est ce pas le transcendantal même de la politique ?) constitue une valeur ? (...) Au terme d’une lecture de l’ouvrage de Boltanski et Thévenot, (…) un problème reste posé : celui de la définition d’une cité de désobéissance…

Hence, authors such as Negri and Latour do not object to ‘transcendence’ as such, but rather to a form of ‘transcendental philosophy’ that attempts to lay down a definite version of the a priori conditions for the formation of concepts (in this case concerning human interaction in view of reaching agreement). Our position meets these critics halfway. ‘Transcendence’ should in any case also be ‘de-essentialised’. This point of view (still concurring with van Brakel (1998)) rests on epistemological insights concerning the inseparability of a priori conceptualisation and a posteriori evidence. Furthermore, as van Brakel (1998) points out, each attempt to accede to a level of a priori understanding is in its turn unavoidably funded in another conceptual scheme, and as a limiting case, the most basic of notions are funded in the ‘manifest image’ (as introduced in section 2.1.1 in the particular context of appeals to sustainability)123. We realise that these epistemological insights would deserve a much more intensive treatment than we can afford in the present circumstances; but our main purpose here was to show how the conclusions of epistemological inquiries support our view that a meta-theoretical dialogue between the 122

Latour – throughout his publications on actor-network theory – does not cease to stress the fundamental uncertainties of the results of ‘action shared with numerous actants’. For a specific critique on the ‘commonwealth model’, see Latour (2004c) – that is at least if we can identify one of the partners in the fictitious dialogue represented in this publication as the spokeswoman for Latour’s theoretical insights.

74

Chapter 1

different theories presented in the above sections is at least possible. It is not a question of ‘pure incommensurability’ but rather of gradations of divergence of opinion – that is, allowing for the right degree of symmetry, transcendence, etc. What remains of course is for us to withdraw useful insights, concerning our specific research topic, from this intertheoretical dialogue. At the second level, our theoretical position starts from an acceptance of the basics of the constructivist approach as explained by Latour (2003b). According to Latour, it is only useful to retain the root metaphor of constructivism, namely ‘construction’, if we insist that we retain elements of the more ‘literal’ meaning of construction, e.g. building, assembling from parts. The SCOT approach is in our view at fault on this account: due to a focus on social interactions, ‘construction’ in SCOT studies inclines more towards a ‘metaphorical’ usage of ‘creating arguments/interpretations/meaning by systematically arranging ideas/terms/concepts’. As we have shown, this approach runs into problems when dealing with the ‘harder’ materials of construction (e.g. structure). They seem to lack the ‘bricks and mortar’ in order to keep the construction together. Latour (2003b, p. 33) has very well pointed out the deficiencies of many constructivist approaches124: …Any constructivists worth their salt should be ashamed to see that everywhere things have been gypped their due : the first treats matter as master, the second as no more than wet sand in a sandbox and the third as an occasion to feel one’s own force being resisted. But with such theories of forces no one could succeed in accounting for even the simplest task: baking a cake, weaving a basket, sewing a button – not to mention erecting sky-scrapers, discovering black holes or passing new bills. And yet most debates on ‘realism’ and ‘constructivism’ never go further than the next child’s toy box. (…) Let us be serious: if the word constructivist has any sort of meaning, it is because it leads us to agencies never falling into these silly and childish roles. Yes, they act, yes they order, yes they resist, yes, they are plastic, but what proved interesting are all the intermediary positions they are able to simultaneously occupy. (…) Everywhere, building, creating, constructing, labouring means to learn how to become sensitive to the contrary requirements, to the exigencies, to the pressures of conflicting agencies where none of them is really in command…

And as he goes on to consider an alternative for ‘constructivism’ (which in the end he does not find), he states that any notion will do, as long as it conveys the meaning of (Latour 2003b, p. 39) …something which a) has not always been around, b) which is of humble origin, c) which is composed of heterogeneous parts, d) which was never fully under the control of its makers, e) which could have failed to come into existence, f) which now provides occasions as well as obligations, g) which needs for this reason to be protected and maintained if it is to continue to exist…

Thus, from constructivism we mainly retain the ‘ideolect’, a certain way of presenting our research – mainly with regard to the avoidance of strict conceptual partitions (e.g. between 123

Maturana and Varela (1992, pp. 241-242) formulate this as follows: “…The whole mechanism of generating ourselves as describers and observators tells us that the world, as the world which we bring forth in coexistence with others, will always have that mixture of regularity and mutability, that combination of solidity and shifting sand, so typical of human experience when we look at it up close. Nonetheless, we evidently cannot break away from this circle and step out of our cognitive domain. (…) Through this ongoing recursiveness, every world brought forth necessarily hides its origins…”. 124 In this quotation, the reference to ‘matter as master’ could apply to what we have called ‘structural’ approaches; the reference to ‘wet sand in a sandbox’ to the ‘interpretive flexibility’ in SCOT studies; and the reference to ‘an occasion to feel one’s own force being resisted’ could apply to some social shaping theories.


75

science/politics, nature/society, etc.). Bateson (1979, p. 61) describes this ‘ideolect’ as follows …To think straight, it is advisable to expect all qualities and attributes, adjectives to refer to al least two sets of interactions in time. (…) Language continually asserts by the syntax of subject and predicate that ‘things’ somehow ‘have’ qualities and attributes. A more precise way of speaking would insist that things are produced, are seen as separate from other ‘things’, and are made ‘real’ by their internal relations and by their behaviour in relationship with other things and with the speaker. It is necessary to be quite clear about the universal truth that whatever ‘things’ may be in their pleromatic and thingish world, they can only enter the world of communication and meaning by their names, their qualities and attributes…

And yet, in the elaboration of these basic ideas of constructivism, we will in the following chapters depart in a certain way from the usual methodological perspective of actor-network theorists. These theorists would admonish us to ‘stick to the action’ by strictly following the traces left by the ‘actants’ (human and non-human actors) engaged in activities of construction (seeking allies, performing an experiment, developing a theory, etc.). Any interpretation or explanation they propose has to obey to a strict principle of immanency – i.e. it has to be advanced by one of the ‘actants’ involved in the process, and not by the observer. Thus, actor-network theorists try to follow ‘as closely as possible’ the everyday practice and parlance – Latour (2004b) even calls himself a ‘naive realist’. The validity of scientific knowledge and the working of technological artifacts depends completely, so it is claimed, on the idiosyncratic features of the local situation in which the knowledge or the artifacts are produced or used. ‘External’ norms, rules, methods, etc. are to be understood to be no more than ‘retrospective rationalisations’. But Latour also adds that the actor-network perspective is especially interesting in situations where ‘things are changing fast’ and/or ‘boundaries are fuzzy’125 – some examples are according to Latour (2004b, p. 62) organisation studies, information studies, marketing, and science and technology studies. With this observation, he provides us with a first reason to deviate from common actor-network methodology, which is a purely pragmatic one, arising from our specific subject matter. Indeed, the switch from ‘micro-level’ studies with an emphasis on material conditions to a ‘macro-level’ study with an emphasis on theories/ideas cannot remain without methodological consequences.

125

This observation lies at the heart of the difference between the actor-network perspective and other sociological approaches. In the end, Latour is interested in those particular situations that have the potential to create entirely new ‘worlds’ by overthrowing our most ‘basic’ intuitions or concepts. He is fascinated by the example of Galilei, who through a particular staging of an experiment (the movement of metal balls on an inclining surface showing the possibility of a mathematical description of the law of gravity), not only succeeded in convincing his contemporaries of the idea a well-designed experiment could produce a truth independent from the then prevailing scholastic reasoning, but also of creating (over the centuries) the entire worldview consisting of an objective ‘world-out-there’ that exists independently from our motivations/desires/etc. (the social world), but nevertheless able to ‘speak truth’ to us through the mediation of experimental practices and scientific machinery (see Stengers (1993) for a historical account on Galilei). It is in the practice of the new ecological movements that Latour (2004a) recognises this potential to overthrow our most basic conceptual categories (see the glossary at the end of this book for an idea of what the new vocabulary of basic terms would look like). In recent publications, Latour (2004b&c) has even reverted to the same Galilean rhetorical trick (representing and re-interpreting his ideas through a fictitious dialogue with opponents voicing other theoretical points of view), perhaps hoping this will do the same for him and actornetwork theory as it did for Galilei.

76

Chapter 1

First of all, the ‘macro-level’ topic we are studying – decision making about energy systems in general – implies an attention for e.g. economic or social assessments intended to give various insights into collective choices, designed to help public decision making, to suggest institutional changes, to bring out priorities in public action, or to improve the knowledge of particular areas, topics, etc126. But these ideas about energy policy and the role of the state, institutional structures, and the energy system itself – in short, the entire ‘matrix of materiality’ (Hacking 1999) – show a great inertia (Laes et al. 2004c), and thus are far removed from the ‘normal’ field of application of actor-network theory. Secondly, following all the relevant ‘actors’ (or ‘actants’) would in this case simply be unfeasible. And lastly, our interest in this research is not so much focused on historical explanation, but rather on exploring the present situation (which of course have their roots in history) and identifying possible ‘seeds of change’ for the future. This implies that we are not only interested in ‘actants’ (theories, groups, decision-making structures, etc.) that have actually influenced the development of the energy system up till now, but also in points of view that have been repressed. The idea that analysts might begin explaining and initiating change at this macro-level with no presuppositions at all – ignoring all previous and possibly valuable insights – sounds very implausible. Hence, we argue that it might be appropriate to begin with a set of a priori (that is, relative to this research) conceptual frameworks that previous analyses have shown to be useful. However, we do not intend to merely add these concepts by way of an explanation in order to close the ‘gaps’ left in the analysis. Fundamentally, we agree with Latour (2004b, p. 74) that “…a structure is only a network on which you have very sketchy information…”. However, in view of the scope of our topic (since one cannot write about everything at once), we quite simply have to settle for some approximations127. These concepts then serve a function as ‘ideal types’ – i.e. no hypothesis or explanation, but offering guidance for the construction of hypotheses128. Furthermore, by attempting to apply the conceptual framework in question, we would not only see its relevance and the extent to which it requires adjustment, but we would also gain an idea of what kinds of reasoning fall outside the scope of the analysis. So, what kind of conceptual frameworks are we talking about? Firstly, the so-called ‘structural’ frameworks we have explored above, ranging from the ‘global’ structural transformation described by Beck as ‘reflexive modernisation’, to the more ‘local’ structural factors affecting (domestic) policy making, such as unequal bargaining power of actors, imperfect knowledge, incomplete time for deliberation, lack of inclusiveness in

126

The situation would of course be different if we were studying energy issues at a ‘micro-level’ e.g. the construction of a particular new policy measure to promote a particular technology (e.g. photovoltaic panels). 127 This is also a consequence of the inaccessibility of some ‘actants’, e.g. in our case, the political decision process that has lead to the decision to phase out nuclear power (cf. Chapter 4). 128 The notion of an ‘ideal type’ was introduced by Max Weber in his work “The Methodology of Social Sciences” (1949), in order to express the value-relatedness of the theoretical instruments of the social sciences: “…An ideal type is formed by the one-sided accentuation of one or more points of view and by the synthesis of a great many diffuse, discrete, more or less present and occasionally absent concrete individual phenomena, which are arranged according to those one-sidedly emphasized viewpoints into a unified analytical construct…” (quoted in Deblonde 2001, p. 19). An ‘ideal type’ is thus not a simple description of reality but aims to give unambiguous means of expression to such a description.


77

deliberations concerning technological choices, etc.129 Again, and we cannot stress this enough, these structural factors should not be seen as determining action in any way but rather as ‘mediators’, enabling some forms of action and restricting others; and, ultimately, as being constructs themselves, existing only as a result of their interaction with other ‘actants’ (and hence being open to modification). We believe that any attempt to develop a theory about a ‘proper’ role of the state in enabling the transition towards a more sustainable society must take, as a starting point, existing structures, and the way these existing structures are implied in either creating or solving sustainability problems. This is a further reason why we call our position pragmatic. Specifically, we look for opportunities in contemporary processes and developments and suggest how they could be goaded in ways that might bring out deeper political and structural transformations toward a system of governance that is more responsive to sustainable development. With these observations, we have already prepared the ground for a second addition in terms of a conceptual framework, i.e. Boltanski and Thévenot’s commonwealth model. As a first reason for the inclusion of this framework is that, supplementary to the ‘structural’ point of view (which tends to downplay or bracket specific dimensions of policy making), the commonwealth model emphasises the role of moral norms, the search for justified forms of action, and the significance of ‘moral entrepreneurship’ in political decision making. As we have explained, the commonwealth model shows particular attention for processes of building shared understanding in view of common action. In our view, this search for legitimate action seems especially relevant for the study of policy making processes, since no political actor can afford to not give a least good reasons for their policy choices, by trying to persuade other people rather than rely on pure coercion, even in highly distorted communicative settings where the possibility of reaching an agreement is remote. Now a predictable objection might be raised at this point that the communicative settings for deciding on energy policy are typically distorted in significant ways, so that we can expect the communicative search for legitimation to play only a minor role (compared to e.g. coercion or strategic bargaining)130. However, such an argument misconstrues the ‘ideal-typical’ role we reserve for the commonwealth model as a critical vantage point from which to observe interaction. Far from removing coercion, restraint, etc. from the equation, this counterfactual ideal enables us to observe the many ways in which the presence of power can distort communication and the search for the legitimate order. In any event, we believe society is not so removed from this ideal that it would be rendered entirely irrelevant. Again, it is not a question of ‘power’ (e.g. deriving from structural conditions) vs. ‘legitimacy’, but of all the intermediary positions in between. Any form of policy analysis (including ours of course) is inevitably caught in the tensions between these 129

Klein and Kleinman’s (2002) article gives a good overview of structural factors that can be used as sensitising concepts: the structure of relevant social groups (how did these groups come into being?, which individuals sharing common meanings were not able to unite?, how do groups enter the interaction process with?); the structure of interpretation; structural factors affecting closure (relations of power and dependence, rules of interaction); and the wider social context (accessibility to resources, concentration or dispersion of each group, etc.). 130 As we will see in our discussion of the nuclear phase-out law in Belgium (Chapter 4), this objection is not only an academic one…

78

Chapter 1

two extremes; a ‘good’ analysis therefore has to make sure that the observations do not go by in a blur of light and shade, but should rather show the unmistakable forms of ‘power’ and ‘reason’ in their subtle ramifications. The second reason for including the commonwealth model in our (meta-)theoretical framework is a more normative one. In view of our search for suitable models of governance for sustainability, the commonwealth model, with its firm roots in the tradition of political philosophy of Western societies, is thus likely to be a good candidate as a point of departure for the elaboration of this new governance structure. However, we do not exclude the possibility of the need for (significant) modifications. Again, our use of the commonwealth model should be understood in the same ‘ideal-typical’ way as we make use of ‘structures’: it is not primarily intended as a hypothesis or research question (i.e. ‘do the justifications encountered in our inquiry conform to the grammar of the six commonwealths explained in Boltanski and Thévenot’s work?’) but rather as a particular construct enabling the formulation of further hypotheses (i.e. ‘if justifications encountered in our inquiry escape the boundaries of the commonwealth model, what are the causes, and how should we then justify action in view of the higher goals of sustainability?’). In particular, the greatest challenge posed by sustainability will be to look for better ways to represent the collectivity of human and non-human actors (including future generations) (Section 3.5). Having arrived at this point, we can wind up our (meta-)theoretical investigations. The inclusion of different frameworks should allow us to become empirically sensitive to a broad repertoire of possible political action, involving as well coercive or strategic action as deliberation or persuasion. Once material conditions, structures, and the potential for legitimate action are acknowledged as possible explanatory factors for the behaviour of actors and as indicating possible ways of change there is no good reason for a priori accepting any one of these modalities as ‘more basic’ or ‘more important’, or to deduce one or more modalities from the other(s)131. Furthermore, while one can of course draw analytic distinctions among these different forms of political action, in practice, they are often enmeshed. They can only be understood in what Ruggie (1998) has called a ‘narrative explanatory protocol’. This is an interpretive account that tries to make sense of what happens, not by deductive explanation, but rather by a ‘thick description’ of events and by arranging them into a more or less coherent shape132. This is then the task we have set for ourselves in the following chapters. From a ‘thick’ description of both theoretical outlooks and the political practice regarding the difficult questions raised by sustainability in the energy field, and their roots in different ‘commonwealths’ (in Boltanski and 131

Eckersley (2004, pp. 37-38) points out very well that often an ‘unfair burden of proof’ is imposed on constructivists by their critics, by demanding a demonstration not only that moral norms matter, but also that they matter to the exclusion of power and interests, in the sense that, in order to have any explanatory power, they must be shown to be untainted by these ‘base’ material interests. 132 ‘Thick description’ is a method first proposed by the anthropologist Clifford Geertz in order to describe cultures from ‘the native’s point of view’. De Laet (1995) gives a good introduction to this notion of ‘thick description’, explaining the underlying point of view that any ‘event’ is constituted by a jumble of different meanings attached to it in different circumstances, by different actors, and in different moments in time. This rich tapestry of meanings should be clarified by the researcher to a reader who is not at all familiar with these events. Therefore, in a good anthropological account, the researcher should be able to show the multiple connections and interactions between the different threads of interpretation.


79

Thévenot’s terminology), we hope to draw a maximum number of insights on possible ways of ‘constructing’ a governance structure for sustainable energy, in view of the ‘better’ construction we envisage (Chapter 5). But before engaging is this task, we can now, armed with our new conceptual apparatus, in the following section give a new rendering of the difficulties involved in founding ‘the sustainable commonwealth’133.

3.5

In search of the ‘sustainable commonwealth’?

Having accepted the commonwealth model as a ‘searchlight’ – in the ideal typical sense outlined above – for the investigation of sustainability theory and practice, we are now able to take a new look at the theoretical and practical difficulties involved in founding ‘the sustainable commonwealth’. Armed with our new (meta-)theoretical apparatus, the question regarding the existence of a ‘sustainable commonwealth’ can only be answered in four possible ways134: 1. Sustainability theory and practice can or should be reflected within the limits of one of the six existing commonwealths; 2. Sustainability theory and practice is founded in a compromise between two or more commonwealths; 3. Sustainability theory and practice requires the foundation of an entirely new commonwealth (however, still accepting the basic grammar of the commonwealth model – cf. section 3.1.2); 4. Sustainability theory and practice is at odds with the existing commonwealths, and even with the basic grammar of the commonwealth model. Of the second possibility we will not say much right now, since most of the following chapters concern attempts to build the sustainable commonwealth with the building blocks already at hand. On the third possibility, we can be equally brief. It concerns here attempts – mostly based on a reconceptualisation of human (social, economic, etc.) undertakings as a subset of nature – to elevate ‘nature’, ‘the biosphere’, ‘the ecosystem’, etc. to a superior common principle135, which, independently of human motivations/desires/valuations, nevertheless would be able to serve as an indubitable principle for organising human action and silencing disagreement. For a number of reasons, we believe that a rejection of this possibility is justified: any attempt to posit ‘non-human’ entities as a source of value independent of human actors doing the ‘valuing’ runs into theoretical difficulties (see e.g.

133

Section 3.5 is a summary of arguments elaborated in greater detail in Lafaye and Thévenot (1993), Latour (1995) and Godard (2003). 134 Setting aside the sceptical views on sustainability as an ‘ethical wrapping’ for relations characterised by power disparities, local arrangements between a limited number of actors, etc. In any case, as we have argued, such a point of view should not be accepted a priori. The reader must also bear in mind that in this chapter we are only preparing the ground for the subsequent analytical work in the field of energy policy, so that the relevance of some theoretical propositions might only become clear in the following chapters. 135 We are referring here to the so-called ‘non-anthropocentric’ or ‘non-human-centred’ theories in the field of environmental ethics – e.g. biocentrism, ecocentrism, land ethics, deep ecology, etc. For a good overview and a critical evaluation of these various theories, see Elliott (1995), Larrère (1997) and Gimeno (1998).

80

Chapter 1

Buchdahl and Raper 1998; Gimeno 1998); the political consequences of accepting ‘ecosystems’ as a superior common principle are unacceptable, since any decision on the ‘good’ of the ecosystem would have to be taken by an enlightened elite group of scientists, who alone would dispose of the necessary means (e.g. ecosystem models) to measure ‘ecosystem health’ (see e.g. Korthals 1994; Latour 1995); elevating ‘nature’ to a superior common principle seems to be the preferred option for only a handful of radical ecophilosophers, but not for the ecological movement itself, whose practices (as opposed to their rhetoric) in political ecology reveal a completely different epistemological and moral foundation altogether (see e.g. Latour 1995, 2004a)136; and finally most practical implications (e.g. proposed policy measures) of ‘non-anthropocentric’ points of view137 can equally be defended from an ‘enlightened’ anthropocentric point of view (Gimeno 1998)138. Let us now turn our attention towards the first possibility of conceiving the sustainable commonwealth as an expression of an already existing commonwealth. Arguably the most influential attempt139 exploring this possibility aims to describe sustainable development as a management process (thus, we move within the boundaries of the industrial commonwealth) involving three different kinds of ‘capital’: human capital (including health, knowledge, culture, etc.), ecological capital (including natural resources and the capacity of the environment to absorb pollution), and the technical-economic capital 136

To summarise Latour’s (2004a, pp. 20-21) arguments (based on a number of research projects involving field work): 1) political ecology does not speak of ‘nature’, but of countless imbroglios (rules, consumers, institutions, etc.) that always presuppose human participation; 2) it claims to protect nature by sheltering it from human intervention, but this amounts to including humans in a finer, more intimate fashion (e.g. with more invasive scientific apparatus); 3) it claims to defend nature for nature’s sake, but has no idea of the ‘common good’ of a dehumanised nature – rather, it suspends our certainties concerning the sovereign good of humans and nature; 4) it claims to think in terms of a ‘system’, but finds itself entangled in scientific controversies every time it proposes to include everything (the political system, the economic system) into a higher cause; 5) it claims to seek its scientific models in hierarchies governed by ordered cybernetic loops, but finds itself confronted with a multitude of original experimental results that show ‘nature’ as sometimes changing slowly, sometimes rapidly, sometimes showing stability, sometimes fragility, etc. – all these ‘sciences’ taken together do not form a stable ‘Science of Nature’; 6) it claims to speak of the ‘Whole’, but it succeeds in upsetting public opinion or power relations only through a focus on the particular (events, biotopes, specific species, etc.). 137 With the exception perhaps of what Stenmark (2002) calls the ‘strong’ bio- or ecocentrists, who for instance advocate a substantial decrease of human populations (one should not think of simply ‘eliminating’ people; other more humane options are usually considered – e.g. through payments for periods of non-pregnancy and non-birth, tax benefits for families with fewer than two children, more funds for research in contraceptive technology, a ‘more realistic’ approach to abortion, the promotion of equal opportunities for women in all areas of life in view of the flourishing of non-human life, etc.), or a drastic increase of wild, uncultivated areas, etc. In any case, population or wildlife management policies are in practice at best marginally connected to policies for sustainable energy. 138 An ‘enlightened’ anthropocentric point of view accepts human beings as the sole source of valuation, but does not accept human beings (including their preferences/motivations/etc.) as the only objects of ethical concern (as opposed to ‘orthodox’ anthropocentrism). 139 The ‘three-capital-model’ is for instance used for the exploration of possible future scenarios by the Belgian federal planning bureau since 2002 in its biennial federal report on sustainable development (FPB 2002). This model was initially developed by the ‘International Centre for Integrative Studies’ (ICIS – University of Maastricht, webpage ), see e.g. van Asselt (2000) and Rotmans and van Asselt (2001) for applications. Furthermore, in the scientific literature dedicated to sustainable development, a vigorous debate is going on between proponents of different ‘management philosophies’, ranging from the ‘strong’ sustainability rule (setting absolute limits to the use of certain capitals) to the ‘weaker’ version (setting more lenient rules for the exchange of different forms of capital). We will not go into this debate here, but for an overview, see Dobson (1996) and Ayres et al. (2001).


81

(including material and financial assets) (FPB 2002, p. 6)140. From this point of view, nonsustainability results from consuming the interest obtained from the different capitals without investing in an equivalent amount of new capital formation. The whole question of course then hinges on this notion of ‘equivalence’ – e.g. to what extent would a generation be entitled to make up for the depletion of natural/physical capital (e.g. oil reserves) by increasing human capital (e.g. research into new technologies, investing in education, etc.)? ‘Equivalence’ plays an important role in the ongoing debate between the advocates of ‘weak’ and ‘strong’ sustainability positions. This debate involves both empirical and normative questions (Gosseries 2001, pp. 342-344). The empirical question is whether two specific capital goods can be considered as ‘perfect’ substitutes in fulfilling a certain function. The normative question is which functions are worth being fulfilled (e.g. human welfare, obligations to nature, etc.), and whether – and if so, to what extent – some functions can be abandoned to the benefit of adding/fulfilling others. Theoretical and practical answers to these questions can then be conveniently classified, running up and down the entire gamut from ‘weak sustainability’ (entertaining ‘generous’ assumptions on the substitutability of different forms of capital), over the ‘preservation of critical natural capital’ (arguing that some forms of natural capital, e.g. the ozone layer, perform functions essential to human survival and/or human wellbeing, and hence cannot be replaced by others), to ‘strong sustainability’ (arguing particularly against any irreversible damage to some forms of natural capital, e.g. biodiversity, even if they perform functions that are not essential to human survival or wellbeing) (Dobson 1996a, p. 407). So what is wrong with this picture? We believe that the whole difficulty of this approach lies in the fact that it revolves entirely around the central notion of ‘management’. From this point of view, sustainable development essentially comes down to a ‘careful management of the three capital stocks’ (FPB 2002, p. 151). Several objections can be raised, only one of which we will develop further in the present context. A first problem lies in the extrapolation of the management instruments typically reserved for industrial settings (e.g. investment decision tools, scenario exercises), where time horizons are typically limited in order to allow for some predictability, to the more complex environment of ‘steering’ an entire society on a long-term course towards sustainability. The quality of the outcome of such exercises in policy advice then becomes very much dependent on the way the policy analyst treats the inevitable condition of uncertainty141. The results obtained are typically highly sensitive to the analyst’s choice of those outcome scenarios that are worthy of attention and to divergent assumptions concerning the likelihoods of different outcomes. The second objection (related to the first one) is that the ‘three-capital’ approach upholds the traditional sharp division of roles between scientists as 140

Only the industrial and the domestic commonwealth allow for a temporality – i.e. making promises for the future – all the others establish cooperations betweens actors firmly established in the present. But since the domestic commonwealth mainly applies to the questions of the transmission of a patrimony, it seems less relevant to our purposes. 141 The treatment of uncertainty ranges from a simple consideration of risk (i.e. the probabilities of the occurrence of a negative event as well as the possible consequences are known) to the consideration of complete ignorance in decision making, and hence, the importance of maintaining diversity in investment options (for a particularly sophisticated analysis of the role of diversity in decision making, see Stirling (1998)).

82

Chapter 1

the providers of ‘facts’ and politicians or moralists as the providers of ‘values’. In this case, scientists describe the current state of the different capital stocks, the functions they fulfil, how efficiently the different capital stocks are used in fulfilling the different functions, the extent to which capital stocks can be interchanged, etc.; whilst politicians or moralists (aided by the objective ‘facts’) debate on the correct way of managing the capital stocks, depending on the functions they would like to preserve or stimulate. The maintenance of this ‘great divide’, no matter how convenient and deeply entrenched in common practice, hides the ‘political’ work done by scientists, or conversely, the selective attention for ‘hard’ facts on the political side as a convenient way to avoid ‘messy’ ethical debates142. It is not the place here to dwell any longer on this objection; the following chapters will provide ample opportunities for showing the much more subtle interpenetration of ‘Science’ and ‘Politics’ than suggested by the ‘great divide’. The third objection on the other hand is more relevant for our present purposes, since it touches upon the limits of the ‘commonwealth model’ we propose as a central interpretive instrument in this dissertation. This objection also immediately takes us to the fourth possibility outlined in the first paragraph of this section. In fact, this third objection revolves around the common denominator of ‘capital’ and the creative ways in which the concept has been understood143. If we want this concept to retain any power of analytic distinction, it should at least carry with it the intention of ‘use’ or ‘functionality’144. A least a number of observers145 have noticed that people’s attitudes towards sustainable development are not exhausted by this representation of ‘stock management’146. According to these authors, what really lies at the heart of our concern over the defence of the environment or future generations is that we are sometimes called to upon to defend them, even when we cannot give good reasons in terms of the ‘functions’ they would fulfil (even in terms of ‘intrinsic’ functions – i.e. functions which do not

142

Just to give an example taken the federal report on sustainable development (FPB 2002): the inclusion of certain scenarios (namely the ‘exploitation’ scenario outlined on pp. 158-161, which exposes future generations to greater risks than the present one) as representing a possible view on sustainability, just to assure a ‘plurality of visions’ is, whatever else one thinks about it, a political stance. 143 An example of such creative use of ‘capital’ can be found in the Belgian federal report on sustainable development (FPB 2002, p. 6), where ‘capital’ is taken to mean something ‘essential’, ‘fundamental’, ‘primordial’. While agreeing that these meanings of ‘capital’ indeed figure in the dictionary, we see the reference to these other significations in the context of this report primarily as a rhetoric trick, especially since in the remainder of the report only the management-related significations of ‘capital’ continue to figure. 144 As Marx wrote: “…Capital consists of raw materials, instruments of labour, and means of subsistence of all kinds, which are employed in producing new materials, new instruments of labour and new means of subsistence…” (quoted in Dobson (1996a, p. 409)). 145 Most notably Burms and De Dijn (1990, 1993), Holland (1994), Latour (1995, 2004a) and Dobson (1996a). This convergence of opinion is all the more remarkable in view of the very different research traditions these authors stem from (e.g. philosophy, anthropology, resource economics, political sciences, etc.). 146 One of the reactions of the adherents to the ‘three-capital model’ has been to redraw the notion of the function of a capital in ever wider circles, e.g. in order to include ‘informational’ functions (education, health, recreation, etc.) (Chiesura and de Groot 2003) in the notion of ‘critical natural capital’, or ‘heritage’ value, ‘spiritual’ value and ‘existence’ value (Ekins et al. 2003). Of course the result of these operations is that the ‘three-capital model’ is more and more able to reproduce human practice and attitudes. But this willingness to continuously adapt the model in order to replicate existing practices is of course no evidence that it is also able to reproduce the moral intuitions that lie at the base of these practices…


83

depend on human values or interests)147. This observation perhaps points at an essential shortcoming of the entire commonwealth model (not only the industrial commonwealth we discussed above), since the elaboration of all commonwealths rests, through the first axiom (the ‘principle of the human community’), rests on an a priori distinction between humans and non-humans, the latter only being deployed in order to support a certain ordering of human beings – i.e. in all commonwealths, they serve a ‘function’. But as Latour (1995) and Burms and De Dijn (1990) argue, it is very hard – if not impossible – to give a clear and strictly rational demarcation of what constitutes (the identity of) ‘the human community’ (the problem becomes even more succinct in view of the possibilities offered by modern technology)148. What these authors are not saying is that we should henceforward, when deliberating on the right course of action, extend the relevant moral community towards ‘the Whole of Nature’, or that, because our criteria of demarcation always reveal a certain particularism, ‘anything goes’. On the contrary, what they say is that we should, at least in some ethical problems, not rely on an a priori unproblematic division of humans and non-humans, quite simply because we do not know how (parts of) the non-human world is (are) implied in the constitution of the ‘human’ identity. Thus, in the words of Holland (1994, p. 178), sustainability from this point of view comes down to: …What is handed down and what is maintained does need to retain in the process something of its original form and something of its identity: there need to be continuities of form, which constitute what may be called ‘units of significance’ for us, as well as continuities of matter…

Perhaps a bit paradoxically, the undeniable presence of the arbitrary or the particular in the thing we value most – our human identity – can be a source for the establishment a new kind of universal order. It is precisely through this experience of particularity, of strangeness in ourselves that we can become sensitive to the experience of strangeness in general. In the words of Burms and De Dijn (1990, p. 91, our translation): …The things to which one is attached most, can be considered from an objective point of view as futile and hence vulnerable. Realising that the things we value most are in fact vulnerable and arbitrary, we become aware of the inevitable arbitrariness of every curtailment of our moral concerns. The corresponding attitude is one of a tension between two elements: the acceptation of particularistic restrictions as inevitable and, at the same time, the realisation of the arbitrary and non-rational character of these restrictions. This means that the deepest form of loyalty is experienced as an attachment that, from a strictly rational point of view,

147

A good example of (the dangers of) a miscast defense of nature in terms of intrinsic functions is the socalled ‘stability-diversity’ controversy explored in Gimeno (1998, pp. 36-38). Gimeno explains how (political) ecologists were very quick to accept the hypothesis that a greater diversity in ecosystems would lead to more stability. Hence, according to this hypothesis, biodiversity would serve a function of ‘ecosystem support’. This hypothesis remained very popular, even in the face of contrary empirical evidence and mathematical modelling which suggested that diversity could actually constitute a form of instability. Defending biodiversity on a functional account thus seems to be an a posteriori rationalisation of a concern that seems to be harder to put into words, but is nevertheless very real (as witnessed by the sometimes dogged defence of biodiversity)… 148 Burms and De Dijn (1990, pp. 89-90) give an example taken from the discussion about animal rights. A rational justification for the different treatment we reserve for humans and animals can for instance rest on morally relevant differences between humans and animals, e.g. types or levels of intelligent behaviour, imagination, sensitivity, etc. But no matter which of these relevant differences is chosen, there will always be human beings who do not dispose of these characteristics, e.g. because they are severely handicapped. Nevertheless, we would never subject these persons to the same treatment as animals, e.g. in medical tests. In the end, the only justification we can give for this different treatment comes down to a vague sense of ‘belonging to the human species’, which, from a strictly rational and universal point of view, is a particularistic and arbitrary reasoning.

84

Chapter 1 has to appear as a coincidence. But it is precisely because one realises the arbitrary extension of one’s loyalty, that one can perhaps become sensitive to a particular kind of respect for the beings that appear near the border of this loyalty: the stranger, the guest, the handicapped person, the recently deceased,… Through the acceptance of the strangeness or arbitrariness in ourselves one becomes responsive in principle to the acceptance of strangeness in general…

To give a more concrete example in the case of future generations of what Burms and De Dijn (1990, p. 88) call a ‘badly understood particularism’ (as opposed to the ‘particularismleading-to-universalism’ so vigorously defended in the above quote), we might refer to a case where an argument of particularity has been used to deny any moral obligations towards (far-off) future generations149. The argument goes as follows: in order for moral relations to exist between people, they have to be part of a moral community, implying a certain consensus on fundamental ethical values. But there is no guarantee whatsoever that (far-off) future generations will share these fundamental values. Taking decisions now in the name of future generations thus resembles a lottery game, since we can never be sure that future generations will agree on our judgment. Hence, we should not be too concerned with future generations and reserve our moral concerns to the present (or nearby) one(s). The argument is bad because it raises the ‘strangeness’ of future generations to the point that we no longer recognise them as human; this is something different altogether than saying that we have no strict criteria to demarcate them as human, no more than we have for ourselves, and that this is precisely the reason why we can form a ‘moral community’ with them… Reformulating all of the above in the idiom of the commonwealth model, Latour (1995) sees the emergence of a new commonwealth, defined by a superior common principle based on an extension of Kant’s categorical imperative to non-humans: namely, in all forms of action, to not only treat them just as means to an end, but as ends in themselves. In this new commonwealth, the ‘state of smallness’ would be to know exactly and irrevocably what a thing ‘does’ (what its function is); while the ‘state of grandeur’ accrues to leaving open the question of the ordering of means and ends. The contours of this new commonwealth imply a suspension of the certainties of the other commonwealths; it induces a moment of ‘perplexity’. This is not comparable to an attitude of scepticism, but rather it designates the starting point for a collective process of building associations between the ‘actants’ involved in a situation. The ‘objects’ involved are subsequently ordered (e.g. assigned to the fulfilment of certain ‘functions’), but only after a due process (cf. Chapter 5). All in all, from this section it becomes apparent that sustainable development cannot serve as a new generic superior common principle in itself, and neither can the established commonwealths (in particular, the industrial commonwealth), with their particular instrumentations, serve as a durable base for the guaranteed achievement of the sustainable society. Thus, for lack of a better solution, we will have to resign to the fact that 149

The argument was originally developed in a 1972 article by Martin Golding (“Obligations to future generations”, Monist, No. 56), but is resumed by Visser ‘t Hooft (1991, pp. 38-39).


85

sustainability will be achieved as the result of compromises between commonwealths, including the not-yet-established ‘extended Kantian’ commonwealth. This then implies a double movement. The first movement entails a search for well-known and firmly established concepts and objects that might serve us as a base for the construction of the sustainable commonwealth. This implies an investigation of existing theories and practices in the context of sustainable energy, and judging their potential application – a task we have set for ourselves in chapters 2, 3 and 4. Next these concepts and objects have to be harnessed in view of a ‘better’ construction (in terms to be specified) – a task that will be taken up in chapters 5, 6 and 7.

4 Summary and conclusions Drawing upon the vocabulary of constructivism, we might characterise sustainability as attempt to rebuild parts of our (social, political, scientific, etc.) fabric. This attempt certainly raises many difficult questions. Within the context of this chapter, we have mainly dealt with two of such questions. Our first question concerned the point of departure of this re-edification effort. A literature survey taught us that appeals to the principle of sustainability could be broadly classified in three classes. The first one concerned the vernacular appeals to sustainability voiced by the inhabitants of the present edifice (sustainability as a ‘manifest image’ – Section 2.1.1), expressing their concern over particular aspects of their everyday experience, often without having any clear positive conception of desired changes. A second form of appeal did have a general idea of what the new edifice should look like, and of the necessary structural alterations this entails, without however disposing of the detailed construction plans (sustainability as a ‘vision’ – Section 2.1.2). The third possibility implied an appeal to the use of standardised, measurable construction elements, without necessarily having a vision of what the new edifice will look like in its entirety (sustainability as a ‘policy target’ or ‘goal’ – Section 2.1.3). The second question we had to answer concerned our role as observer/analyst, or perhaps even architect. Should we, in our analysis, accept the fact that people, in their constructive effort (e.g. when denouncing situations or when searching for a legitimate common ground in order to coordinate action), always act with a limited number of idealised blueprints (i.e. established principles and orders of justification) in the back of their minds – even if they cannot realise them fully (Boltanski and Thévenot’s ‘commonwealth model’ – Section 3.1)? Or should we accept the idea that the re-edification efforts are largely determined by structural forces – e.g. that some parts of the present edifice (e.g. ‘risky’ technologies) are rapidly deteriorating to the point that they cannot be upheld by any constructive effort whatsoever (Beck’s theory of ‘risk society’ – Section 3.2)? Or should we rather just describe the constructive (or destructive) work done by the people involved (including the interaction with building materials, the plans they employ, etc.), accepting it at face value, without pronouncing any opinion ourselves (the ‘(social) constructivist account of technology’ – Section 3.3)?

86

Chapter 1

Our provisional answers to these questions have been that, on the first question, we leave a maximum degree of freedom, at least as a point of departure, without however allowing everything – one cannot simply repaint the existing edifice in a different (most likely green) colour and claim that it has become sustainable. This means that we do not propose our own vision or (technical) criteria of sustainability at the very outset of our research, but rather accept a list of characteristics which we deem necessary in order to speak of sustainability at all (Section 2.3). This list will be used by us as a guideline for questioning existing theories and practices, in order to examine how they respond to the ‘minimal’ requirements – e.g. by justifying certain choices, arguing the greater importance of one requirement over others, etc. On the second question, our response has been to first clear away the apparently insurmountable theoretical barriers between the different perspectives (Section 3.4). This in its turn has lead to an acceptation of the central tenets of constructivism, understood in an inseparably empirical, pragmatic and moral sense. This implies that in the following chapters we will first investigate how existing theories and practices in the field of energy policy respond to the sustainability challenge. In turn, we will deal with theoretical models of governance (Chapter 2), the (influential) neo-classical interpretation of sustainability in terms of external costs (Chapter 3), and the actual policymaking practice in the context of the decision taken to phase out nuclear energy in Belgium (Chapter 4). But our story does not end there. We are not only interested in an empirical description of existing ‘building blocks’, but also in the question how these building blocks could be arranged in view of a better, generally satisfying, ‘sustainable’ construction. The question then becomes one of finding amongst the existing theories and practices of decision making a procedural structure able to produce intellectually and morally forceful argumentations in view of a better decision support in the field of energy policy (Chapter 5). The last chapters of this dissertation will then be devoted to a limited test (a ‘pilot exercise’) of this new procedure (Chapters 6 and 7). For a befitting ending to this (lengthy) (meta-)theoretical exploration, we once again turn to Latour (2003b, p. 27) and his unique view on constructivism: …Negotiations toward a viable and peaceful common world are possible among constructivists, but radically impossible if fundamentalists are expected to show up at the diplomatic table, and religion is not the only domain for bigotry : nature can trigger zealots as well, so can markets, so can ‘deconstruction’. Between war and peace stands a realistic definition of what a construction is; this, at least, is my argument…

It is precisely this search for a ‘realistic’ definition of construction that lies at the heart of the argument in the pages to come.

CHAPTER 2 IN SEARCH OF SUSTAINABLE ENERGY POLICY: FOUR THEORIES OF GOVERNANCE In chapter 1 we have developed and defended the position that constructivism – at least when it is stripped of its materialistic and essentialist pretensions – allows us to understand how ‘things’ are produced – i.e. how they are seen as separate from other ‘things’ and are made ‘real’ by their internal relations and by their behaviour in relationship with other ‘things’. And, following Latour, in the previous sentence ‘things’ should not be understood in opposition to ‘subjects’: the notion of ‘construction’, normally only applied to artefacts, has to be extended to scientific knowledge, values, concepts, etc. In this chapter, these insights will be applied to policy making strategies with an emphasis on the problems typically encountered when devising a strategy for sustainable (energy) governance. Based on a survey of existing theories and historical evidence, we reconstruct four idealtypical governance schemes. We show how each of these ideal-typical schemes relies on a ‘network’ of relevant knowledge, norms, actors, practices, institutions and technical artefacts for reducing the (often intractable) uncertainties facing any (political) decisionmaking process. In each case we also show how the particular reduction of uncertainty – though an absolute precondition for action – is not unproblematic when applied in the context of ‘radical uncertainty’ often exposed by demands for more sustainable policies. The focus on (the reduction of) uncertainty as an entry point into the debate leads us first to consider different types of uncertainty facing policy makers (Section 2.1). Next, we consider a typology of different policy problems and solution strategies (in terms of the scientific support, norms, practices, etc. mobilised in order to ‘solve’ the problem) which can be considered as ‘dominant cooperative schemes’ of governance in democratic societies (Section 2.2). In order to arrive at solutions, each of these strategies has to work through a number of steps, each having its own functional exigencies in terms of the required justifications (Section 2.3). These governance schemes are presented in section 3 in terms of their strengths and drawbacks. Expert-based governance (Section 3.1) operates according to a (deep-seated) logic that political authority and decision making can be separated from the scientific authority provided by experts, based on disciplinary competence. Governance by aggregation (Section 3.2) seeks a resolution of policy problems through negotiations between different ‘interests’. Governance by pacification (Section 3.3) represents a widespread political strategy to tackle (politically organised) value positions of a different kind, and has shown its value in the past, also in Belgian energy policy. A last scheme – deliberative governance (Section 3.4) – is strong in its insistence on giving a ‘voice’ to the voiceless and promoting the values of ‘unconstrained dialogue’ and ‘social learning’.

88

Chapter 2

Overall, our investigation is thus guided by the same logic as set out in chapter 1: as each scheme proves to be vulnerable from a particular perspective, we will try to reduce weaknesses by systematically bringing in other perspectives on governance. The more practical level of investigation (at least compared to chapter 1) allows us to bring in more concrete examples taken from the history of (nuclear) energy policy in Belgium. Hence, the present chapter functions as a bridge between the (meta-)theoretical explorations of chapter 1, and the more down-to-earth (combined scientific-political) practices empirically reconstructed in chapters 3 and 4.

1 Introduction In the following sections we will take a closer look at the possible contribution of a number of governance strategies150 to the construction of the ‘sustainable commonwealth’ (cf. Chapter 1). Because of our subject matter, our focus will be on the processes involved in the assessment and shaping of technology in society, and related issues such as the (unavoidable) management of risk and uncertainty; this focus thus constitutes a first demarcation of our research. As mentioned before, a second demarcation consists in the fact that actions and decisions of individual engineers, scientists and/or the producers and users of technology will not be discussed here. Neither will we discuss organisation-level decisions concerning ‘relevant’ companies for the technology in question (e.g. designers of new nuclear reactor types, electricity companies operating nuclear power plants, etc.)151. Rather, we will discuss here the societal and/or governmental mechanisms (mechanisms of ‘governance’) for somehow ‘steering’ energy technology, focussing on actions taken at the

150The

term ‘governance’ rather than ‘government’ will be used here, since this notion is not limited to the specific formal functions of government and the political system; but it also includes the informal organisation and regulation of collective affairs that is often taken for granted in society (although the focus of our analysis will still be on the actions of the political system within this enlarged governance structure). ‘Governance’ therefore recognises the strong influence both formal and informal networks have on forging shared beliefs, allocating rights and obligations among parties, legitimating initiatives taken by policy and promulgating collective interests (‘the common good’) (Burgess et al. 1999). 151 This also implies that, in the context of this dissertation, we will not discuss a fifth possible model of governance, commonly described as ‘self-regulation’ (Pellizzoni 2003, p. 336-337). The European white paper on governance (EC 2001b) and many other policy documents and academic papers insist on the importance of self-regulation by social partners. Self-regulation can take many forms: quality assurance between suppliers and customers, third-party certification of environmental performance, publication of guidelines or standards of good practice by sectoral business associations, publication of codes of conduct and ethical codes for various fields and professions, etc. This is not meant to say that these instruments cannot play an important role in directing society towards more sustainability – although different critical objections can be raised, e.g. concerning the non-legally binding character of sectoral standards, the danger of merely ‘greenwashing’ the public image through the adherence to these instruments, etc. However, ‘local’ (in functional and spatial terms) solutions have to be fitted in with more ‘global’ solutions, or else they run the risk of producing different kinds of ‘externalities’. Therefore, we will devote most of our attention to this ‘global’ level of ‘official’ decision making institutionalised in the political system, although our suggestions for an improved decision-making architecture do include the two levels of ‘official’ decision making and the actors carrying out the desired policy changes (cf. Chapter 5).

In search of sustainable energy policy: Four theories of governance

89

level of the ‘political system’152. With the notion of the ‘political system’ we intend to capture a whole range of forms of political organisation, ranging from administrative bodies, ministry officials, political parties, parliament, parliamentary discussion groups and committees, scientific advisory groups, etc. We immediately admit that in practice much of the assessment of different (energy) technologies, and its surrounding climate of relevant knowledge and sufficient levels of certainty, exist in the private corporate sphere where the various agents of public policy have limited access. Also, as shown by Beck (cf. Chapter 1), ‘civil society’ (e.g. through actions of NGO’s) or even individual consumers nowadays at least have increased opportunities to engage in ‘informal’ forms of technology assessment and the exertion of ‘sub-political’ influence on technological choices. Nevertheless, it seems to us that against this backdrop of broader societal evolutions the role of policy influence on the development, uptake and social acceptance of new technologies has not declined to the point that it becomes irrelevant to study it and/or suggest improvements (Meadowcroft 1997). The starting point of our analysis will be the problem of ‘decision making under uncertainty’ (set in the context of technology policy) which, at least according to some authors (e.g. Beck 1992; Stirling 1999a; Wynne 2001), has become one of the defining issues of modern times153. This increasing salience of the issue of uncertainty is for instance mirrored in the rise of the precautionary principle in national and international law and jurisprudence154. Our primary interest in decision making under uncertainty should not be seen as an implicit value judgment concerning the other dimensions of sustainability. It merely provides us with a convenient entry point into the discussion, as we will show that the treatment of uncertainty in the different theories of policy making cannot be separated from an articulation of the broader sustainability concerns we set out in chapter 1 as a ‘thin conception of sustainability’ (Chapter 1 – Section 2.3): a focus on values beyond economic growth (and especially, dealing with the so-called ‘incommensurable’ values, i.e. value perspectives that are very hard to reconcile), inter- and intragenerational equity, prospective analysis, an integrated view on development problems, special attention for

152

We will therefore also not treat the ecological critique of the administrative state as a source rather than a potential solution for environmental problems. The eco-anarchist Carter (1999) for instance argues that states have, over the centuries, sought ways of extending their revenue base, enlarging their sphere of influence within civil society and expanding and rationalising their administrative apparatus. According to Carter, these are the ‘state (or the state/military complex) imperatives’: an autonomous source of power independent of society or capitalism. The result has been the development of a hierarchical system of depersonalised and specialised bureaucratic power, the production of an apathetic public and the reliance on ‘non-convivial technologies’. Contra Carter, we will argue that state power (or administrative power) can be harnessed to bring about more sustainability. This is a matter of principle (environmental benefits are public goods that ought to be managed by democratically organised public power) as well as of pragmatics (rather than designing society de novo – as some bioregionalists would seem willing to do – it seems more useful to us, in view of the urgency of many environmental problems (e.g. global warming), to build on existing state structures and try to make them more ecologically accountable). 153 Beck’s theory of ‘risk society’ is discussed at some length in chapter 1. 154 For an overview and analysis of the impact of the precautionary principle on the legal system, see de Sadeleer (1999) and Lierman (2004, pp. 45-129).

90

Chapter 2

environmental problems, global responsibility and participation in decision making155. Moreover, using uncertainty as an entry point for studying technology policy seems justified for a number of reasons. The first is an empirical one: irrespective of the actual ‘conscious’ attention given to uncertainty in decision-making processes, the key point is that intractable problems always have to become (and have become) somewhat tractable in practice156. Understanding the how and why of such patterns of creating and maintaining tractability furthermore enables us to reflect on the quality of the processes and outcomes, and if necessary, to shift or at least modulate them (i.e. a moral reason). Admittedly, our third reason is somewhat more pragmatic: in chapter 5, we will show how our proposal for a design of more adequate arrangements, rules, and perhaps even institutions, has a better chance of success when it is firmly located in the growing international (at least, European) acceptance of the precautionary principle as a guideline for policy making. The overall argument is thus in line with what we have called our ‘inseparably moral, empirical and pragmatic’ constructivist perspective. The present chapter should therefore be understood as an effort to make visible certain (more or less historically applied) constructive patterns of dealing with (energy) policy problems, to consider their practical application in the context of the (often difficult) questions raised by the demand for more sustainability, and, based on an assessment of their respective strengths and weaknesses, to derive some practical insights for the formulation of new perspectives on sustainable energy policy.

2 Technology policy problems and policy-making strategies

2.1

A typology of uncertainty

As Wynne (1992, 2001) asserts, decision making under uncertainty has become ‘the defining issue of modern, scientifically-informed times’. One should of course always take care not to fall back too rapidly on generalisations, but Wynne’s analysis (for that matter closely resonating with Beck’s theory of risk society) nevertheless seems to have touched upon a particularly sensitive chord of modern times, particularly when attention is focussed on the domain of technology policy, where complex environmental and health risk have to be traded off against the presumed benefits of new technological developments.

155

Given our emphasis on actions at the level of the political system, participation can now be circumscribed more accurately as follows: “…Political participation in the context of sustainable development encompasses every political interaction between the state and civil society or between public actors aimed at solving societal problems. This includes the process by which governments and civil society have a dialogue, from partnerships to solve problems, exchange information and also the interactions between the state and civil society during the development, the implementation and the evaluation of politics, programs and projects aimed at a more sustainable future…” (Bruyninckx 2002, p.298). 156 This is especially salient since we will also use historic examples in our overview. Wynne (2001) gives a convenient ‘crude late-20th century chronology of the multiple faces of uncertainty’, arguing that the treatment of uncertainty has evolved from ‘deterministic decision and scientific advice’ (1950-1960s) to an ‘institutional acknowledgement of reductionist reflexes’ and a ‘precautionary embrace of unanticipated consequences as an unavoidable predicament of policy making’ (2000s). However, this last assertion is followed by two big question marks in Wynne’s paper…


91

Uncertainty arguments concerning environmental and/or health risks (implicitly or explicitly) therefore occupy a central place in political sciences and/or ethical analysis of (technology) policy choices, having been cited both as a basis for adopting ‘an ethic of caution’ (e.g. in limiting resource use), as for discounting dire predictions of environmental decay on the ground that these cannot be proved definitively (e.g. in the debate following publication of the famous “Limits to Growth” report by the Club of Rome – see e.g. de Vries 1996). However, in our effort to understand how uncertainty has been managed in practice, we face a first difficult issue – that is, the issue of finding an encompassing viewpoint on uncertainty that allows us to make visible the multifarious ways the concept has been defined as a necessary first step towards practical management. It seems that here again, as was the case for finding a definition of ‘sustainable development’ in chapter 1, we are confronted with the same fundamental difficulties. Slovic (1999, pp. 691-692) has nicely phrased these for the related concept of ‘risk’: …Just as there is no universal set of rules for games, there is no universal set of characteristics for describing risk…

Indeed, ‘risk’ and/or ‘uncertainty’ seem to be very fundamental concepts, lying very close to the everyday experience represented in ‘manifest images’ (cf. Chapter 1 – Section 2.1.1), and hence being prone to multiple attachments in subsequent rationalisations157. On the positive side however, the similarities between the difficulties involved in defining ‘uncertainty’ and/or ‘risk’ and ‘sustainability’ allow us to transpose our (meta-) theoretical reflections entirely to our present concern of finding an adequate typology of uncertainty. Indeed, browsing through uncertainty-related literature, one again encounters the by now familiar positions of a realist representation of uncertainty as an independent condition of ‘objective-reality-out-there’ as well as various forms of (social) constructivism. The former would for instance maintain that scientific uncertainty, on whatever scale it is thought to exist, provides the objective boundaries to possible legitimate policy divergence158. A related belief is that the more scientific certainty and precision – including a clear delineation of the remaining scientific uncertainties and, possibly, the means to reduce them – could be achieved in the knowledge relating to a policy decision, the less problematic would be the ensuing policy decision, both to make it

157

For a good overview and a critical discussion (in the constructivist vein) of ‘definers’ and ‘definitions’ of concepts such as ‘safety’, ‘risk’ and ‘incertitude’ see Bombaerts (2004, pp. 33-45). 158 van Asselt (2000) might be cited as an example of this ‘realist’ position. According to van Asselt, uncertainty has both an ‘ontological’ (variability) and an ‘epistemological’ (lack of knowledge) origin. Different sources of variability are e.g. ‘inherent random processes in nature’, ‘human behaviour’, ‘technological surprises’, etc. Epistemological uncertainty covers the entire gamut of ‘inexactness’ (measurement errors), ‘lack of observations or measurements’, ‘practical immeasurability’, ‘conflicting evidence’, ‘reducible ignorance’, ‘indeterminacy’ and ‘irreducible ignorance’ (van Asselt 2000, pp. 86-87). The point is, as Bombaerts (2004, p. 43) convincingly argues, that all of these categories implicitly start from the idea of a comparison between ‘theory’ and ‘reality’, even (in the case of ‘irreducible ignorance’) involving implicit judgments on the future capacities of science !

92

Chapter 2

and to justify it159. At the other extreme, the latter (social constructivist) position claims that uncertainties are negotiated, for instance during a process of policy formulation, and hence represent the interests or power positions of the parties involved in the negotiation160. As Wynne (2001) points out, these questions are not merely about ‘discussing the sex of the angels’, nor are they ‘innocuous’ in a political sense. For in a ‘realist’ view, political or social responsibility is limited to the management of ‘given’ uncertainties (unequivocally defined by ‘science’), while in the social constructivist view responsibility is entirely shifted on the shoulders of the social actors involved in negotiating uncertainty. As was the case for our discussion of ‘sustainable development’ in chapter 1, we argue here also that it is not necessary (and furthermore counterproductive) to make an a priori decision in favour of one of these positions. In our constructivist view, uncertainty should be seen as deriving from broadly conceived processes of human commitment – including material interventions in nature, our attempts to theorise about these interventions through a creation of credible knowledge, policy-making processes, and everything in between. Nevertheless, despite the obvious difficulties, it seems useful to have some further conceptualisation of the blanket term ‘uncertainty’ at hand that covers (as far as possible) the entire breadth of uncertainty as encountered in the literature covering ‘decision making under uncertainty’161. Our intended use of this typology is not in a prescriptive sense (e.g. pinning down uncertainty on an ‘objective scale’), but rather in a diagnostic sense, in order to arrive at some clarity in the confusion that is often (even unintentionally) caused by the way different actors refer to ‘uncertainty’ in science or technology-related controversies. We are certainly not claiming that this is a definitive classification, but rather that it is good enough for our purposes. For instance, many more distinctions could be made in our overarching categories of ‘uncertainty’ and ‘ignorance’; we however are more interested in the borderline between ‘risk’ and the other categories as the principal instigation for a transition from ‘traditional’ risk regulation to more ‘precautionary’ based measures (this will hopefully become more clear in chapter 5 when we discuss the invocation of the precautionary principle as a crucial justification for our proposed decision-making framework). Also, the categories should not be seen as mutually exclusive; in practice, they can be overlaid one on the other. For instance, a lot of authors in the risk field refer to the so-called ‘new’ risks (such as BSE, GMO’s, climate change, etc.) as manifestations of a new and important form of ‘radical uncertainty’ (e.g. Pellizzoni 2003, p. 328).

159

We will return to this belief and critically assess it under section 3.1 of the present chapter. Let us limit ourselves here to one telling counter-example: contrary to the realist assumption, it is frequently found in the empirical study of several risk and technology fields that the more scientific understanding is given to an issue, and the more ‘knowledge’ is being produced, the result is not more, but less certainty, along with intense expert disagreement (see e.g. the chapter on ‘controversy’ in Jasanoff et al. 2002). 160 Smithson (1989) seems to be one of the first analysts explicitly talking about the social construction of uncertainty. A clear problem with the ‘social construction of uncertainty’ is the way in which social construction is imagined to be purely cognitive-strategic, leaving out the possibility of a surprising ‘practical’ discovery of new uncertainties (e.g. in technological innovations). 161 In Table 3, a more precise specification of ‘uncertainty’ is given. In the remainder of the text, we will clearly indicate whether we are talking about uncertainty in the broad encompassing sense, or in the more limited sense employed in Table 3.


Decision context

Explanation

Examples

Risk

Known consequences / Quantifiable probabilities; uncertainties may have statistical (e.g. stochastic) nature

Defining clear thresholds (level of protection); preventive policy measures; ALARA principle in radiation protection

Uncertainty

Known consequences / Unknown or uncertain cause-effect relations, therefore unknown probabilities

Measurement errors (e.g. modelling imprecision, extrapolations, etc.); lack of knowledge (but knowledge is principally possible); indeterminacy (e.g. nondeterministic outcomes)

Ignorance

Unknown scope of consequences; however, degree and/or nature of ‘seriousness’ can be estimated in qualitative terms (e.g. ubiquity, persistency, carcinogenicity, mutagenicity, bio-accumulation, etc.)

Scientific controversies; contested knowledge (epistemic uncertainty) – e.g. GMO’s, climate change, etc.

Ambiguity

Variability of interpretations based on identical observations or data assessments

Uncertainty about socio-political positions (most likely also conflict)

Hypothetical effect/ imaginary risk

Arguments on the basis of a fully conjectural knowledge, no scientific indication of possible occurrence of the claimed effects

93

Table 3. The different dimensions of uncertainty (Source: based on Smithson 1989; Funtowicz and Ravetz 1993; Stirling 1999a; Wynne 2001; Boudourides 2003; Renn et al. 2003; and von Schomberg 2006).

Radical uncertainty is then understood to be …different from the simple ‘disagreements’ of routine political debate. The latter can be resolved by appealing to the ‘facts’ – that is, by using shareable kinds of rational argument referred to scientific research, witnesses, past experience, and so on. The former cannot. In this case, the parties in dispute tend to emphasize different facts, or give them different interpretations, so that each party seeks to confute the empirical evidence adduced by the others. There is no consensus either on the relevant knowledge or on the principles at stake. Facts and values overlap…

Referring to our Table 3, radical uncertainty thus appears to be a conflation of ‘uncertainty’, ‘ignorance’ and ‘ambiguity’, and therefore does not escape from our attention. All in all then, our attempt at categorisation represents a compromise between a broad enough coverage, analytic simplicity and practical use, whilst taking care not to prejudge the ‘ontological status’ (i.e. objective or socially constructed) of risks and uncertainties.

2.2

Structuring policy problems

In policy sciences (not limited to the context of technology policy or risk management), different authors have tried to explicitly trace out the links between the definition of policy

94

Chapter 2

problems, knowledge production and methods or practices of political decision making (see e.g. Hisschemöller 1993; Hisschemöller and Hoppe 1996; van de Graaf and Hoppe 1996). Authors from this school thus claim that different approaches to decision making and knowledge production are justified in light of different conceptions of the problem at hand. The fact that policy problems are political constructs is generally recognised in policy sciences; however, much less attention is given to how these problems are constructed or defined. In other words, much attention is usually given to the strategies deployed to solve problematic situations, rather than to the logically preceding phase of problem structuring. Furthermore, these authors argue that the ‘real’ power lies with the one who is able to impose his/her definition of the problem at hand, because together with the problem definition, the available scope for solution finding is ‘smuggled’ furtively into the decisionmaking process. Hence, policy making is not limited to answering the question on how to solve a particular problem; more importantly, it is also a question of who decides on the ‘right’ way of defining the problem, who gets access to information, which groups have the opportunity to influence the policy decision, etc. It is important to add here that we do not want to suggest that the different ways of structuring problems and/or conceiving solutions are necessarily and always the result of a deliberate strategy. Even setting aside the scientific difficulties of liberally assigning intentions to political (or other) actors, such ‘politics of intentions’ would never be a sufficient explanation of the particular strength and perseverance of policy strategies. This crucial insight is for instance embodied in Buchanan’s (1996) notion of ‘dominant cooperative schemes’. As Buchanan maintains, the dominant cooperative schemes (a notion he even extends to entire societies) have, strictly speaking, never been chosen. Instead, according to Buchanan, they have emerged from the cumulative (and largely) unanticipated effects of many interactions among many generations of individuals162. Such schemes are thus never entirely ‘rational’ (no single actor has the ability to foresee the full consequences of the application of a particular scheme) or entirely ‘imposed’ by powerful interests (in democratic societies, policy measures are always supported by a mixture of power and legitimacy). According to us, this view comes very close to our elaboration of ‘structures’ or ‘institutions’ in chapter 1. Our aim in the following paragraphs will then be to explore these structures with regard to problem definition and problem solving in the context of technology policy. In view of the still rather abstract (however less so than the previous one) character of this chapter, we can only provide summary ‘sketches’ of these structures (cf. Chapter 1), whilst however taking care of not being overly simplistic, or, even worse, erecting straw men as convenient foils for our theories. Elam and Bertilsson’s

162

Starting from yet another angle (that is, a sociological perspective on technological controversies), Rip and Talma (1998) and Rip (1999) inquire into the nature and functioning of arrangements to create some tractability in difficult (technology) policy problems, for example in regulating for occupational health and safety, and how to improve such arrangements. Following this train of thought, Rip and Talma (1998) and Rip (1999) identify a design challenge in helping to create arrangements and processes which achieve tractability. They see ‘design’ both as an intentional activity with a product that must be implemented, and as a process of de facto design in which new practices, procedures, norms, and institutions emerge. Thus, a distinction is made between an ‘intentional’ design strategy, and a ‘pattern’ design, the goals and approaches implicit in the actions and interactions as they occur, and which can be, but need not be, made explicit and reflected upon.


95

(2003, p. 234) admonition should serve the reader as a statutory reminder of also our unavoidably moral commitments: …behind every authoritative account of major changes of science and society relations stands a more or less explicit vision of how the future ‘knowledge-based society’ should be organised. The work of accounting for change is never innocent of a desire to make a difference to change…

This being said however, we see no reason why moral commitments would automatically degenerate into unfair representations of other positions (of course, the ultimate judgment on this lies with the reader). The ‘sketches’ will afterwards be fleshed out in more detail in our exploration of a particular (well-developed) example of a science/policy arrangement in the context of energy policy (the theory of external costs – cf. Chapter 3) and our description of the policy development cycle in the context of the Belgian decision to gradually phase out nuclear power and the resulting societal debate (Chapter 4). Hisschemöller (1993), Hisschemöller and Hoppe (1996) and van de Graaf and Hoppe (1996) essentially (with some minor variations between them) identify four types of discourse on policy problems mapped out in two dimensions. One dimension refers to (the lack of) certainty concerning the kinds of knowledge a problem may require. The other refers to the (lack of) consensus on the common values at stake which need to be politically defended (i.e. ‘the common good’, ‘basic rights’, etc.)163. Referring to the different dimensions of ‘uncertainty’ as set out in Table 3, a problem is termed structured when there is a rather high degree of consensus and certainty concerning the problem at hand (i.e. the problem is principally amenable to scientific insights). A problem is referred to as moderately structured (ends) when there is a broadly shared consensus on the common values at stake but, outside of the boundaries defined by this consensus, ambiguity on the interests of different social groups (or citizens). A problem is called moderately structured (means) when there is a consensus on what kind of knowledge is relevant, but ongoing dissent with regard to the (balancing of the) common values at stake. A problem is called unstructured when there is neither value consensus nor certainty. Table 4 gives a schematic representation of the different types of governance and the degree of ‘unstructuredness’ they permit in order to function effectively. All of these strategies are compatible with the basic idea of the democratic constitutional state, but promote (fundamentally) different conceptions of democratic method or practice to address the issue of ‘decision making under uncertainty’164.

163

Of course, a whole spectrum of problems can be imagined in the margin between ‘structured’ and ‘unstructured’. It would be more precise to speak of ‘relatively’ structured vs. ‘relatively’ unstructured problems, with ‘completely’ structured and unstructured problems as (ideal typical) extremes. Furthermore, ‘facts’ and ‘values’ can only be identified a posteriori, after the political construction work has been done. 164 Hisschemöller (1993) for instance supports his classificatory scheme with arguments found in different political theories. More specifically, he distinguishes the technical approach to policy making (based on certain elite theories), the market approach (based on economically rational conceptions of political participation), the distributive justice approach (based on theories which stress the role of the state in protecting vulnerable interests in society) and the public participation approach (based on egalitarian or anarchist political theories).

96

Chapter 2

Dissent on principles for

Conditions of

Conditions of

Risk / Uncertainty

Ignorance / Ambiguity

Structured problem

Moderately structured problem (ends) (only ambiguity is admitted)

Expert-based governance

Governance by aggregation

Moderately structured problem (means)

Unstructured problem (ignorance/ambiguity are both admitted)

the ‘common good’ small

Dissent on principles for the ‘common good’ large

(no ambiguity since socio-political positions are well-known)

Governance by pacification

Deliberative governance

Table 4. A typology of policy problems and approaches (Source: based on Hisschemöller 1993; Hisschemöller and Hoppe 1996; van de Graaf and Hoppe 1996)

Broadly speaking, structured problems require solutions according to standard procedures involving quantifying technical methods. Unstructured problems require methods for clarifying or finding the problem itself. These methods focus on social and political interaction, rather than on quantifying techniques in order to reveal the assumptions of policy actors holding divergent views. Some measure of problem structuring is needed in these cases, implying awareness-raising of a problem by means of confronting, evaluating and integrating as much (potentially) contradictory information as possible. Before proceeding with the intended examination of the different policy problem solving strategies when confronted with conditions revealing variable degrees of uncertainty, more clarification seems to be in place here. Table 4 in fact makes two different claims. The first is an empirical one. The four types of problems correspond to empirical observations of problem structuring (e.g. in the context of technology policy, through different kinds of technology assessment) and policy approaches in a lot of policy fields, which makes it a suitable candidate for our purpose of finding ‘dominant cooperative schemes’. The second claim is a more normative one. Put simply, it is claimed that particular combinations of problem definition/policy strategy are more robust then others165. Hisschemöller (1993) exemplifies this claim by his concept of the ‘Type III-error’166. A Type III-error occurs when the ‘wrong’ problem is solved by employing a method that does not apply to the problem at hand. For instance, when the expert-based approach is applied to a problem which implies a political choice regarding the values at stake, political differences of opinion need to be actively suppressed or blurred (of course depending on the political

165

Stirling (1999a, p. 27) gives the following definition of ‘robustness’: “The capacity to sustain performance under external perturbation by maintaining an established internal structure.” 166 A Type I-error occurs when both the problem and method are well-defined, but a mistake is made in applying the method (e.g. an error of measurement). A Type II-error occurs when the problem is stated correctly, but a faulty method is applied in solving the problem.


97

salience of the issue). After all, a structured problem implies a large consensus on the technical nature of the problem at hand. The crux is then to demonstrate or identify the mechanisms by which certain interests or points of view are marginalised. Here again, we have to be careful not to fall into the trap of pre-defined ontologies. Therefore, the scheme should not be used as a technical manual for an objective approach of the policy problem at hand (i.e. ‘we encounter policy problem x, so we should apply strategy y’). One should always keep in mind that policy problems are constructs. Whether or not there is a consensus about the values at stake or about relevant knowledge cannot be judged outside of the ‘network’ of actants involved in the policy-making initiative (depending for instance on whoever is implied in the decision-making process). The classificatory scheme should therefore rather be regarded as a methodological aid for empirical observation and as a guideline for a substantiated problem choice. Furthermore, its main value does not lie in providing new and directly ‘usable’ knowledge for policy making. Rather, it points to the ‘white spots’ and the uncertainties in the information on a policy problem because possible alternative problem approaches, and the corresponding policy strategies, are made visible. All of these strategies are characterised by a certain approach to knowledge production and utilisation, by patterns of interaction and inclusion/exclusion mechanisms. The aim of the scheme is thus to make the one-sidedness of certain approaches more visible, without however giving a decisive answer on the ‘correct’ interpretation. A justified problem choice always remains a (partly) normative choice, be it from the part of a policy analyst, an observer, a policy maker or other stakeholder groups.

2.3

The policy development cycle

Problem structuring and the choice of an appropriate policy strategy thus require an active and constitutive form of labour in order to reduce the inevitable uncertainties surrounding the policy choices. Moreover, in order to be legitimate, the policy maker in principle has to be able to justify his/her choices to a larger constituency (which can be either physically present in the decision-making process or can be represented through some form of ‘representative thinking’ from the part of the policy maker). In this process of justification, a policy maker can, again in principle, be called upon to justify his/her choices on four levels of political judgment. These levels were already explained in chapter 1; but for purposes of clarity, we will recall (and somewhat expand) these here: First order judgments: • Technical verification: concerned with empirically checking whether certain given policy goals are indeed attained by the policy measures which were introduced in order to solve or alleviate a problematic situation. This is a matter of ‘factual’ observation, including a possible search for reasons why a certain policy goal is not attained (‘do we get what was promised?’); • Situational justification: concerned with justifying or criticising the stated policy goals, given a certain layout of the field of application of the proposed policy measures. This is a matter of searching for the right ‘values’ – i.e. the ones that are

98

Chapter 2

‘applicable’ to the situation at hand (‘what can we propose as legitimate goals for policy, given the circumstances?’); Second order judgments: • Institutional support: concerned with justifying or criticising the stated policy goals, keeping in mind the contribution of these goals to the maintenance of the established order. This is a matter of checking whether the newly proposed situational justifications fit in with the hierarchy of already (implicitly or explicitly) ‘institutionalised’ justifications (‘what can we propose as legitimate policy goals, given a certain way of life, (political) culture, dominant ideology, social order, etc.?’); • Reasonable choice: concerned with justifying or criticising policy goals or political institutions in light of a certain value system (possibly independent of prevailing societal values). This is a matter of making ‘ultimate’ judgments – i.e. on the wished for identity of the political community, on the final hierarchy of facts and values to be taken into account concerning the policy decision, etc. (‘which values do we want to embody in our political community, and which institutions are needed to realise this political identity?’). Of course, while we began this section by stating that an individual policy maker can in principle be held accountable on all four levels of justification, policy-making practice is, much as many other social activities, specialised and differentiated to a large extent. Taking this into account, it would be more correct to replace the individual policy maker by ‘the political system’ (in the sense explained above), viewed then as one large ‘politicaljudgment-producing’ mechanism. This mechanism then ensures that – at different moments, on different political levels, in different places – political judgments are formulated, identifying and transforming ‘policy problems’ into ‘desired end states’, and assigning means and responsibilities in order to move towards the ‘desired end state’. In policy sciences, this process is often referred to as the ‘policy development cycle’. In this policy development cycle, different separate constituent processes are identified. We will discuss these briefly here, as they provide a convenient way for structuring our further analyses167: • Agenda building: the process of agenda building ensures that ‘matters of concern’ (more or less sharply experienced problematic situations) are transformed into a certain common understanding about the urgency and more precise definition of problems to appear on the political agenda. Because the political agenda is always at risk of getting overburdened, problems need to be selected and structured. As explained, the structuring of problems always implies a transformation of the ‘matters of concern’, whereby certain policy solutions are already implicitly implied in or excluded from the problem definition. Ideally then, agenda building requires a judgment on all four levels of political reasoning (cf. supra). The process of agenda building maps out the ‘demand

167

A more complete description can be found in Leroy and Nelissen (2000).


•

•

•

•

99

side’ of the political system and includes formal as well as informal mechanisms. Problems can for instance appear on the political agenda because of certain unexpected developments (‘surprises’, e.g. a reactor incident or accident), media attention, deliberate agenda-building by an individual politician or a political party, demands made by the administrative bodies, etc. Referring to the context of sustainable development, we will argue (in chapter 5) that the process of agenda building ideally should show a great sensitivity to the ‘manifest images’ (cf. Chapter 1 – Section 2.1.1.), allowing itself to be ‘surprised’ by the multiplicity of ‘voices’ making demands for sustainability; Policy development: the process of policy development ensures that the items on the ‘political agenda’ (e.g. a declaration of policy, political programmes, etc.) receive an ‘adequate’ solution through a process of generating, selecting and comparing different possible policy options. This again is a process that takes place on many different levels; e.g. media pressure, lobbying by stakeholder groups, demands made by political parties, etc. all will play a role. In this phase, judgments will mostly concern situational justifications (the situation being defined as it is in the process of agenda building) and institutional support. Referring to the context of sustainable development, policy development entails the formulation of a strategic ‘vision’ (Chapter 1 – Section 2.1.2.) – somehow, the ‘cacophony’ of voices demanding attention has to be directed towards a form of ‘polyphonic’ agreement; Judgment: the process of judgment leads to a political ratification (e.g. in policy notes, texts of law, decisions on the government budget, etc.) of one of the policy options generated during the previous phase. Political judgments are arrived at in a very delicate balancing operation between ‘reason’ (i.e. at one extreme, the ‘neutral’ choice for the ‘best’ option) and ‘power’ (i.e. at the other extreme, the option preferred by the most powerful actor). Political judgment sets down the goals and the instruments of a certain policy plan, including the mandates of the different organisations involved in carrying out the policy plan. Referring to the context of sustainable development, political judgment entails the formulation of an operational (specific, measurable, well-timed) plan (Chapter 1 – Section 2.1.3.); Execution of policy: the process of the execution of policy leads to a mobilisation of means (e.g. people, organisations, budgets, etc.) in a structured way in order to carry out the political decisions. In practice, this can lead to a divergence between the intended effects of the policy measures in question and the ‘real’ effects; Collective dynamic: the ‘collective dynamic’ (an envelope term) leads to the transformation of the (intended and not-so-intended) results of previous policy measures and ‘autonomous developments’ (e.g. economic growth, growing individualisation, etc.) into matters of concern directed at the ‘political system’. These may then be taken up in another policy development cycle.

100

Chapter 2 Agenda building Matters of

Selection and

concern

structuring of policy problems

Collective dynamic Autonomous developments

Results of previous policy

Policy development Policy analysis

Initiation policy

Definition,

instruments and

comparison and

organisation

selection of options

Judgment

Execution of policy Political decisions

Figure 1. The policy development cycle (Source: van de Graaf and Hoppe 1996, pp. 79-99; Leroy and Nelissen 2000)

Figure 1 gives a schematic representation of the policy development cycle which we will use throughout our work, in particular as a convenient tool to structure our policy analysis of the Belgian parliamentary decision to gradually phase out nuclear power (Chapter 4) and for ordering our suggestions for an adequate decision making aid in the context of sustainable energy policy (Chapter 5). In order to avoid misunderstandings, two remarks seem in place here (see also van de Graaf and Hoppe 1996, pp. 93-98). The policy development cycle explained here is clearly a model – i.e. a reconstruction of policy development as it occurs in the ‘real world’. For instance, ‘real-life’ policy making will not go through the different phases of the cycle in a neatly ordered way, moving from ‘problem definition’ to ‘policy implementation’ after each development phase has been duly considered and closed. In practice, different phases will take place more or less simultaneously, and there will be a lot of overlap and feedback between the different processes. However, this observation does not take away from the value of the policy development cycle as an analytic instrument, since the different caesurae, e.g. between policy formation and policy implementation, have to be recognisable from the policy result (policy making under democratic conditions at least requires that a policy makers gives some explanation why certain measures are undertaken). Also, the above discussion might give the wrong impression that in order to arrive at justified policy decisions each and every step of the


101

policy development cycle has to be completed anew. In fact, the greatest part of ‘everyday’ policy making consists of tinkering with existing policy measures (responding to signals from policy evaluations or monitoring) or hasty responses to acute problems. Van de Graaf and Hoppe (1996, p. 96) make a distinction between a ‘practical’ and a ‘reflective’ form of judgment, the former being largely conditioned by technical verification of existing policy and an assessment of the immediately available policy resources or, at the most, situational justifications. While this does not exclude the more ‘reflective’ forms of judgments which clearly are needed in the context of policy for sustainable development, there clearly exists a tension between the exigencies of everyday political practice and the ‘higher’ forms of judgment (e.g. at the level of ‘reasonable choice’). If we want to make our suggestions for an improved decision-making practice somewhat credible, this tension has to be taken into account.

3 Four governance schemes for sustainable development The time has come now to show how the different ‘dominant cooperative schemes’ identified by political sciences can be put to use in the context of governance for sustainable development. In doing so, we can of course build on a lot of previous work done in this field. However, bringing together the different existing approaches in one systematic overview seems necessary to us for at least two reasons. The first is that such a systematic account seems to be lacking in literature168. A large number of approaches rely heavily on the model of expert-based governance, and thus (as will be shown in section 3.1) tend to give a rather simplistic account of the problems of governance in complex policy fields. Other accounts do present more than one possible theory of governance, but they seem to do so in order to demonstrate the ‘obvious’ superiority of one preferred mode of governance over the other, rather than looking for possible synergies169. While not downplaying the merits of these previous analyses, our intention is not so much to show the advantages of one scheme over another (in fact, we will show that each of them taken separately shows deficiencies) but rather to show the possibilities for a constructive interplay between the different governance strategies. A second reason is then that we intend to use this systematic account as a starting point for our own model of governance for sustainable development, as set out in chapter 5.

168

With the exception perhaps of Pellizzoni (2003), who discusses the challenges posed by the new conditions of ‘radical uncertainty’ to the production of knowledge, governance structures and public deliberation. His subject coverage is however broader than ours, ranging from the international level (European governance) to the decentralised level (self-regulation). Nevertheless, we have found inspiration in Pellizzoni’s analysis of the interplay between knowledge production and governance structures. 169 We can of course not do justice to the particularities of each of these approaches. However, broadly speaking, ‘expert-based governance’ is commonly pitted against the advantages of (some form of) ‘deliberative governance’, e.g. in Burgess et al. (1999) and Vandenabeele (1999). Deblonde (2001) promotes a form of ‘governance by deliberation’ (close to but not quite the same as the model of ‘deliberative governance’ we will discuss – but this is of course quite irrelevant for the general tendency we wish to demonstrate) whilst stressing the problems encountered through the application of ‘governance by aggregation’.

102

Chapter 2

Our discussion of the four theories of governance in the next sections will be structured along the following lines. Firstly, the theoretical underpinnings, in terms of arguments taken from political philosophy (thus aimed at the level of ‘reasonable choice’), will be discussed170. We will show how each of these theoretical outlooks supports a particular way of structuring policy problems and arriving at solutions through a reliance on a ‘network’ of relevant knowledge, norms, actors, practices, institutions and technical artefacts. Where possible, more concrete examples (taken directly from our historical account of (nuclear) energy policy in Belgium) will be given. Each strategy will then be assessed for its possible merits and drawbacks when applied in the context of sustainable development.

3.1

3.1.1

Expert-based governance

Theoretical background

Knowledge, and more precisely scientific-technical knowledge, is a crucial resource in expert-based governance. This is not meant to imply that in the other theories experts play no role whatsoever, but simply that in the expert-based governance discourse this role is much more accentuated: policy measures are typically legitimated with a reference to (scientific) expertise. Typical examples of expert-based governance include bureaucratic or technocratic governance modes171. In the latter case, often encountered in policyrelevant situations were scientific/technical knowledge is more ‘uncertain’ (but still – at least in the discourse on the policy problem – considered to be reducible by future scientific research), the scientist/expert is given an important role as an arbitrator, providing the necessary directions for collective action. In terms of Boltanski and Thévenot’s commonwealth model, the expert-based governance discourse is firmly situated in the industrial commonwealth, were scientists/experts appear as subjects imbued with ‘grandeur’. This ‘grandeur’ is derived from the fact that scientists/experts are conceived as independent researchers, who look for facts and are free from any value commitment, apart from those proper to the scientific community. As explained above, it can be argued more 170

Thus, in this chapter, we will follow the ‘reflective order’, working our way down from general philosophical arguments to more practical reflections. Our intention is however not to discuss each of the philosophical theories we invoke in great detail. We merely want to show how each governance strategy can be supported by certain arguments taken from a range of (particularly well-known) philosophical examples. This implies for instance that one cannot conclude from the following discussion that the particular political theory in question wholly supports the governance strategy in question. On the other hand, in our reconstruction of the policy development cycle in the case of the Belgian decision to phase out nuclear power (Chapter 4), the focus will lie on the ‘practical order’ of argumentation. 171 Hisschemöller (1993, p. 164) gives the following defining characteristics of bureaucratic governance: different actors and groups can take part in the preparation of a policy decision, but only one political actor (appearing as a ‘monolithical block’, e.g. a political leader, ‘the government’, etc.) has the power to make binding decisions; insofar as there exist conflicts over the policy measures to be implemented these are not characterised to the outside world as ‘being of a political nature’; there is a clear hierarchy and division of responsibilities between the institutions or groups involved in policy making, with a clear demarcation of competences; and a dominant role is accorded to ‘specialists’ ensuring an efficient use of policy mechanisms in order to arrive at uncontested and broadly accepted policy goals.


103

convincingly that this ‘grandeur’ is not so much associated with ‘real’ intentional or behavioural features of the expert, but rather with the ‘structures’ embodying the values according to which the experts are presumed to act172. According to Merton (1973), in modern knowledge discourses these values include the principles of impartial evaluation of claims (through the system of peer review), knowledge-sharing, systematic doubt and personal disinterestedness – in short, the wellknown canon of ‘organised scepticism’. In this logic – i.e. the grammar of the ‘industrial commonwealth’ – knowledge controversies can only arise because either the legitimacy of the expert discourse or its veracity is questioned (Pellizzoni 2003, p. 329). Legitimacy refers to a normative dimension as it concerns a value judgment on the ability to speak. Experts can for instance be criticised because they lack the necessary competence or skill, as officially attested (e.g. through the role played in educational or professional institutions, through professional experience in the field, etc.). Therefore, experts will usually take great care to stay within the limits of their mandate, as any transgression possibly exposes them to criticisms by colleagues. Veracity on the contrary refers to a more cognitive dimension. Criticising a knowledge claim based on its veracity means raising a problem either regarding the pertinence of the knowledge claim (i.e. not producing the right kind of knowledge as required by the situation at hand) or, of course, regarding the correctness of the claim as an adequate representation of reality. This ‘internal’ regulating mechanism for the production of relevant and correct scientific-technical knowledge is however exposed to greater strains when confronted with other forms of reasoning in policy problems. In fact, translating ‘scientific’ knowledge into ‘policy-relevant’ knowledge is an inherently unstable and underdetermined process, and thus requires a substantial amount of ‘boundary work’173, as shown by numerous studies in the field of the sociology of science174. This inherent instability hinges on the essentially open-ended character of the scientific enterprise. As both Karl Popper and Thomas Kuhn have shown in their field of the philosophy of science, there is no absolute assurance that findings, and/or new arguments, will not undermine present achievements. There are degrees of solidity, of course. But the nature of scientific observation and experiment, and the precarious shift from specific findings to more general knowledge claims, always leave openings for doubt and further checks. Closure of the quest is a practical matter, not a logical step. A first problem is then that in the field of environmental policy, knowledge is often of a controversial nature and thus often involves a debate on the epistemic level (i.e. a debate on

172

Examples of such ‘structures’ are 1) scientific communities, ad hoc or ‘independent’ experts, consultancies, scientific advisory committees; 2) stakeholders with ‘in-house’ experts; 3) functionaries of various government agencies or ministries (EC 2001b; Radaelli 2002). 173 ‘Boundary work’ entails the creation of ‘objects’ (models, scientific methodologies, assessment criteria, etc.) that are accountable to both ‘social worlds’ of science and policy (see e.g. Guston 2001). For an example in the field of nuclear expertise, see Bovy et al. (2003). 174 For an overview of the field, see e.g. the reference works by Vinck (1995) and Jasanoff et al. (2002).

104

Chapter 2

the ‘right’ theoretical perspective)175. Policy makers then have to make ‘hard’ decisions based on ‘soft(er)’ facts. Or it might be that a certain experiment works under the circumstances specified in a laboratory – but who is to ensure that the circumstances in ‘the-world-out-there’ are exactly the same176? These possible difficulties notwithstanding, there is something more to be said about the problem of translating science for policy purposes than is suggested by this ‘realistic’ account of the uncertainties involved. This second set of problems confronting ‘science for public policy’ has been adequately labelled by Wynne (2001) as the ‘janus-face’ of science. Wynne means to indicate by this the way in which science is regularly presented as offering definite predictive knowledge (e.g. concerning the consequences of new or existing technologies, in the form of institutionalised risk assessment), yet at the same time (e.g. when the hidden uncertainties underlying such assessments and predictions become apparent in the form of ‘unintended side-effects’) science is exonerated from blame because after all, following Popper and Kuhn, scientific knowledge is forever provisional177. Rip (1999) has noticed something similar in his sociological account of the ‘variable tolerance of uncertainty in scientific results and expertise in general’. Metaphorically speaking, this phenomenon is identified as the ‘trough of uncertainty’ (Rip 1999, p. 100): …Uncertainty can be, and generally is, tolerated within a scientific specialty, especially by the ‘core set’ working at the research front. When knowledge is to be used in professional activities (in another scientific specialty, in professional practice, in preparing policies and decisions), however, actors want it to be as solid as possible. (Except in situations when there is an interest in finding an opening for alternatives, and uncertainties are welcomed). When scientific findings are disseminated to broader audiences, the link with action is absent, or indirect, and there is more tolerance (or perhaps just indifference) for uncertainty. A graph with tolerance to uncertainty on the vertical axis, and closeness to knowledge production on the horizontal axis, has the shape of an animal feed trough – the trough of uncertainty.

It is important here to notice that, again, our intention is not to lay blame on either ‘the experts’ or ‘the politicians’ for this situation – it seems to be part of a large and broadly shared dominant scheme of interaction, which undoubtedly has shown to be ‘productive’ (e.g. able to arrive at decisions in an efficient way) in the past. The productiveness of the scheme, often operating under time pressures or resource limitations, hinges for a large part on the ability of institutions to represent and translate issues they have to deal with into

175

An example from the nuclear field is the debate on the health effects of the exposure to low doses of ionising radiation. Radiation protection is based on epidemiological evidence, involving an extrapolation from the observed stochastical health effects at the higher dose region (based on the so-called linear no-threshold assumption). Because exact experimental epidemiological results are impossible for the low dose region (this would require the exposure of an immense research population to low doses of radiation), the debate on the health effects continues, informed by disciplines such as molecular biology or genetics. 176 This is the argument used by Beck and others that the environment and society as a whole is becoming ‘one big laboratory’ (cf. Chapter 1). 177 This is also why Latour (2004a) accuses the realist (in the epistemological sense) understanding of science of engaging in ‘(political) epistemology’ – i.e. engaging in politics (in the large sense of ‘constructing the common world’) without however admitting to do so (hence, the brackets around the ‘political’).


105

forms which reflect control and orderly management178. Wynne (2001) speaks of ‘the universal tendency’ of organisations and institutions to reduce explicit agendas to match existing forms of intellectual and practical control, and he goes on to discuss several typical manifestations: ...whilst the (maybe very few) specialist knowledge-actors at the very production-centre may be aware informally of some of the contingencies involved in testing, validating, reproducing results, etc., this awareness disappears for actors in neighbouring specialities or user-groups. The latter operate with a simplified account of that knowledge in which some of its key uncertainties and contingencies have been deleted, not by anyone’s deliberate manipulation but just by the routine economies of communication. Further away, socially-speaking, others may again re-inflate scepticism, though for different reasons from those shaping the author’s own circumspection (…) In those typical situations where policy knowledge requires synthesis of several scientific specialist knowledges, the synthesis may embody an aggregate of the understated (and translated) uncertainties that tend to characterise the understandings and representations of adjacent (user or research) fields (…) Studies of scientific communication which have little to do with policy fields note a general tendency for scientists to translate the local contingencies of actual scientific research into ‘sanitised’ representations, which results in the understatement and translation of explicit uncertainty so that only limited and tractable forms of uncertainty are given explicit recognition…

The above argument thus enables us to characterise (in an ideal typical way) the ‘dominant cooperative scheme’ of expert-based governance as follows. The ‘traditional’ policy discourse of expertise provides specialised knowledge for well-delineated policy issues, and moreover, it does so under the banner of the ‘common good’ (or principe supérieur commun, which, in the case of the ‘industrial commonwealth’ can only be the ‘efficient performance of the (governance, or societal, or ecological etc.) system’). The scheme essentially assumes that a technical definition of the policy issue at hand is possible, so that it can be settled by relying on established specialised forms of knowledge. The forms of specialised knowledge required are decided upon by referring to the exigencies of legitimacy and veracity explained above. As a side effect, the appeals to scientific knowledge usually play down or, at least, set apart other forms of ‘non-scientific’ knowledge (e.g. particular, private, unverifiable, intertwined with personal interests, etc.) by playing on the fact/value and relevance/marginality oppositions – e.g. stakeholders in the decision can voice their ‘values’ or ‘interests’ after the experts have decided on the ‘facts’, politicians decide after a careful weighing of the ‘values’ and ‘interests’. Lay knowledge is disabled because citizens usually lack the ability to speak pertinently and

178 Fischer (1990) signalises a surreptitious narrowing down of governance strategies to technocratic ones. In his view, this is caused by the tendency to limit public participation with an eye on the efficiency of decision making : “…for technocrats the solution is to replace the ‘irrational’ decision processes of democratic politics (group competition, bargaining and compromise, in particular) with ‘rational’ empirical/analytical methodologies of scientific decision making…” (Fischer 1990, p. 22). For Fischer, this tendency is not ‘imposed’ upon society from the outside, but rather emerges from within.

106

Chapter 2

appropriately179. This inability is usually not considered to be a permanent fixture (else, people would be excluded a priori from the commonwealth), but rather a temporary ‘deficit’ which is amenable to reason (given enough time and education efforts)180. In the case of sustainable development, taking responsibility is thus reduced to an alignment of individual or group behaviour with the existing scientific knowledge (translated into institutionally-anchored norms, rules, laws, etc.) The expert-based governance discourse thus not only provides an interpretative framework for the role of scientific knowledge in policy processes, it also produces ‘identities’ (Wynne 1996, p. 55): …Furthermore, the most fundamental dimension of risk expressed in such social interactions is that of the risk of social identity which is felt to be involved in this kind of dependency, upon expert institutions which disseminate and impose such models of the human and the social, whilst pretending to deal with objective facts…

Expert knowledge on the other hand is able to provide the necessary repertoire of disciplines, models, theories, approaches, etc. as required by an unequivocal issue definition. Moreover, as observed by Pellizzoni (2003, p. 330), a cooperative scheme focused on specialised knowledge implies the tendency to increasingly narrow down the definition of ‘relevant’ abilities – i.e. increasing the specialisation of usable expertise. This (admittedly somewhat broad) characterisation of expert-based governance enables us to investigate in a second step its merits and drawbacks in providing guidance for governance in the field of sustainable development. Before doing so however, we believe it might be useful to descend from the abstract theoretical level of reflection of the

179

On this issue, it is relevant to mention Wynne’s (2001) observation of a ‘profound dislocation’ between understandings of uncertainty within scientific and policy institutions and those which informed lay public concerns (about the management of GMO’s). Wynne reports how the institutional and expert understanding was based on the ‘best available science’, recognising only known and observable uncertainties within the boundaries of existing risk assessment. These were considered to be fully tractable to further research were needed (e.g. through farm-scale evaluations of the effects of GM crops on biodiversity). Regarding the issue of public opposition to GMO’s, according to Wynne this was usually thought to result from the ‘false public expectations’ of an absolute proof of safety prior to issuing a permit for a particular GM crop. Research into the issue however revealed that people did not expect certainty – in fact, they took uncertainty for granted. Thus, their concern was not so much that science could not predict ‘all’ possible consequences, but rather that this state of affairs was effectively suppressed by the existing institutions in denying their responsibilities in handling the consequences beyond the boundaries of established risk assessment. This concern for the more long-term unpredictable consequences would then explain widespread public demands for labelling of biotechnological applications. Furthermore, in light of this public recognition of ‘ignorance’, public questioning often turns to the interests behind the ‘uncertain’ innovation (‘Given the ignorance, are the driving purposes at least valid?’; ‘What kind of contingency plan, if any, could be established?’, etc.). See also the results of the public panel (the ‘Publiforum’) on genetically-modified food organised by the Flemish institute for technology assessment (viWTA 2003). 180 This argument can be considered ‘mainstream’ in political theories of the early 20th century, for instance in Schumpeter’s view that the ‘average citizen’ is incapable of a rational judgment on complex issues that go beyond the experiences of daily life. In his book “Capitalism, Socialism and Democracy” (1942), Schumpeter sees the democratic method as essentially analogous to the ‘normal’ competition and marketing in commercial markets: politicians accede to power positions as a result of a competition (conceived as) for the favour (and votes) of citizens. Once elected, the politicians have to assume responsibility for the decisions; citizens should be kept as far away as possible from actual policy-making processes. Citizen participation would, in Schumpeter’s view, undermine the quality of the decisions taken and the stability of the democratic state in general. The general argument is summarised in Hisschemöller (1993, pp. 92-93).


107

foregoing paragraphs to discuss a practical illustration taken from our historical account of the nuclear power debate in Belgium (Laes et al. 2004c, pp. 48-71)181. 3.1.2

A historical example: nuclear safety regulation

Our illustration takes us back to the mid-’70s, when, as a result of a siting conflict, a wider societal debate on the sense or non-sense of the nuclear option for Belgium erupted182. The fundamental decision to build nuclear power plants in Belgium had been taken earlier, after consultation in a closed committee (the ‘Boereboom Commission’) involving both government representatives and representatives of the electricity holdings183. As a result of the work of the ‘Boereboom Commission’, the ‘ministerial committee for economic and social coordination’ agreed (in December 1966) in principle on the suitability of Doel, Tihange and Zeebrugge to serve as candidate locations for nuclear power plant construction. However, while the scope of this decision was limited in principle (concerning only possible locations for nuclear power plants, whilst not prejudging the results of the procedures required by law in order to receive permits to actually build and operate nuclear power plants), historical analyses clearly reveal that the then government in fact sanctioned a significant nuclear investment programme in the decades to come, as witnessed for instance by ministerial declarations, press conferences, and policy notes (Laes et al. 2004c, p. 51). Furthermore, the role of official organisms (e.g. state administrations, provincial or communal authorities) in judging the applications for construction and operation permits was limited, since nuclear know-how was almost exclusively owned by the electricity holdings (through their engineering branches) and the officially recognised private organisms responsible for nuclear safety control and radiation protection, with considerable overlap between the roles of the ‘controller’ and the ‘controlled’ (e.g. representatives of the electricity utilities had a seat in the board of management of the control organisms, the control organism was appointed and remunerated by the electricity utilities, etc.)184. Early decisions concerning the nuclear power programme in Belgium thus represent a schoolbook example of technocratic decision making (cf. Section 3.1.2).

181

This report, containing a historical overview of the nuclear power controversy in Belgium starting from the early post-war period till the recent debate on the phase-out law (2003), should be read very much as a companion to the present volume. 182 In January 1974, Belgian electricity utilities applied for a construction permit for a nuclear power plant in the harbour zone of the town of Zeebrugge (situated on the Belgian coast). At that time, three nuclear reactors were already under construction and would become operational shortly after: Doel 1 (with a capacity of 392.5 MWe) later in 1974, and Doel 2 (392.5 MWe) and Tihange 1 (962 MWe) in 1975. Contrary to Doel and Tihange (where there was no significant opposition), in Zeebrugge a local action group called ‘REM-U-235’ took the lead in the opposition against this proposed nuclear siting. 183 Called after the president of the commission, the chief of cabinet of the then minister of public works. The ‘Boereboom Commission’ counted thirteen members, four of which represented the electricity holdings. The commission met seven times in the period October 1965 - June 1966 (Reynebeau 2000, p. 125). 184 At that time, three organisms existed: ‘Controlatom’, ‘Corapro’ and ‘Association Vinçotte Nucléaire’ (AVN). Only the latter had sufficient resources and expertise in order to control the Belgian nuclear power plants. Nowadays, AVN is still the designated private control organism for the Belgian nuclear power plants.

108

Chapter 2

Since the Belgian nuclear power plants are of an engineering design developed by the American multinational company Westinghouse185, the safety design and control of the Belgian power plants were heavily influenced by the American regulations in vigour, as set down by the ‘US Nuclear Regulatory Commission’ (USNRC)186. Nuclear safety analysis received a new impetus in the mid-’70s due to the publication of the USNRC (1975) “Reactor Safety Study”, a landmark application of probabilistic safety assessment (PSA) through computer modelling. The “Reactor Safety Study” is more commonly known under its document number (WASH-1400) and its connection with the author, the ‘Massachusetts Institute of Technology’ (MIT) professor Norman Rasmussen (hence, it is also commonly called the ‘Rasmussen-report’). WASH-1400 gives an account of the first application of a probabilistic method to the analysis of the safety of nuclear power plants187. It relies on a 185

ACEC, an at the time important Belgian company active in electrical engineering (under the control of the Société Générale, an influential Belgian financial holding), had obtained in 1957 a licence from Westinghouse for the construction of nuclear power plants of the ‘pressurised water reactor’ (PWR) type, in exchange for a participation of Westinghouse in ACEC. In 1970, Westinghouse assumed control of the entire nuclear engineering division of ACEC. This meant that, when the nuclear power plants were actually built, Belgian companies could only contribute in the field of ‘classical’ construction (e.g. the concrete containment buildings) and electro technical engineering and components (e.g. turbines). 186 One noticeable exception (in view of the debate on nuclear power in Belgium): while USNRC criteria following the Three Mile Island (TMI) accident (1979) called for the organisation of emergency planning (and possible evacuation of people) in a 10-mile radius surrounding a nuclear power plant, this requirement was relaxed in the Belgian context to a ten-kilometre radius, since the application of the ‘ten-mile’ criterion to the Doel power plant would have been impossible in view of the location of Antwerp (a densely populated city), downwind from dominant wind directions from Doel. Belgian power plants were however equipped with more extensive safety measures than required by USNRC guidelines (e.g. double containment buildings, an extra separate bunkered control room for emergency cases where the normal controls have been destroyed, etc.). After hearings in the Belgian senate (which took place in the period 1986-1991, spanning two legislatures) following the accident at Chernobyl, it was confirmed in the recommendations that emergency planning should not be limited to the 10-kilometre zone (Gedr. St. van de Senaat, Doc. 113-26 (B.Z. 1991)). Furthermore, the senate also recommended that nuclear power plants should not be located within 30 kilometres from major population centres (a condition which is not met by neither the Doel (near Antwerpen) nor the Tihange (near Luik and Namen) sites). Following the Chernobyl accident, a zone with a 30-kilometre radius around the destroyed reactor was permanently evacuated. Admittedly, the reactor at Chernobyl was of a completely different type than the PWR-reactors used in Belgium; and safety practices and culture in the former Soviet Union were blatantly inadequate (SCK•CEN 2001). But nevertheless it is evident that the debate about emergency planning and siting requirements has a large relevance for a densely populated country such as Belgium. 187 Before the publication of WASH-1400, nuclear safety analysis and regulations were based solely on a deterministic philosophy – i.e. ensuring that the public would be protected from the consequences of any possible accident (in practice, a set of postulated accidents, including a ‘maximum credible accident’) through an appropriate combination of containment and isolation (relying on a combination of expert opinion, engineering rules of thumb, and operational and experimental experience). Effectively, the assumption was that the containment building would hold under any accidental circumstances. The fundamental shift in safety philosophy was provoked by a 1967 study by the ‘US Atomic Energy Commission’ (USAEC), predecessor of the USNRC, which showed that under certain circumstances (the so-called ‘loss-of-coolant-accident’ or LOCA), the reactor containment could be breached (the so-called ‘China syndrome’, whereby the molten reactor core would melt through the reactor pressure vessel and the concrete floor of the containment building into the ground). The ‘China Syndrome’ debate introduced the direct correlation between core melt and a loss of containment integrity. Emphasis shifted from the design of containment buildings to the design of ‘Emergency Core Cooling Systems’ (ECCS) for the prevention of core melt; and the LOCA received primary attention as the most probable (although still very small) source of an accident which might lead to core melt. In 1971, small-scale experiments at the Idaho reactor testing station showed that an ECCS would possibly fail to inject enough cooling water into the reactor core (the high steam pressure that was created in the vessel by the loss of coolant blocked the flow of water from the ECCS). The USAEC reacted by publishing strict ‘interim acceptance criteria’ for new licensing applications, pending more scientific research.


109

detailed methodology called ‘fault tree analysis’ in order to give an estimate of the probability of different accident sequences that might lead to a melting of the fuel during nuclear power plant operation. In a fault tree analysis, one constructs a logical model of the sequence of events involved in a nuclear reactor accident. An accident in this analysis requires first the failure of one component in the engineering system, and then the successive failure of components in the safety systems designed to operate in the case of an anomaly occurring188. Hence, it becomes possible, based on ‘reliable’ information (e.g. based on engineering experience, modelling of components or expert guesses) on the probability of failure in the physical systems responsible for a ‘normal’ functioning of the reactor, to calculate the probability of an accident sequence. A ‘good’ analysis thus comes down to deriving an accurate model of the dependent probabilities among the different engineering systems and using a (theoretically simple, but computationally complex) application of probability theory to assess the probability for an entire sequence of events (Thompson 1986, p. 62). More important for our purposes here is that the conclusions of the ‘Rasmussen-report’ (which put very briefly considered the risks of a reliance on nuclear power acceptable, especially compared to other risks society was willing to take)189 were

It wanted to settle the uncertainties about safety without arousing a public debate that could place hurdles in the way of the quickly expanding market demands for nuclear power plants. For a number of action groups (e.g. the ‘Union of Concerned Scientists’) and individual experts, this approach was unsatisfactory, and the debate finally ended in a controversy on a national scale (public hearings in the period 1972-1973). The hearings damaged the USAEC’s credibility and led to its division into the aforementioned ‘Nuclear Regulatory Commission’ (USNRC) and an ‘Energy and Research Development Agency’ (ERDA) in 1974. In that way, the two original functions of the USAEC (promotion and regulation) were effectively split, meeting the demands of the critics. Afterwards, it became clear that the insufficient core cooling resulted from a faulty estimation of the high thermal inertia of the small-scale reactor experiment. Thus, while later large-scale testing programs did confirm sufficient core cooling, the strict acceptance criteria were maintained and confirmed in 1974 as ‘final acceptance criteria’, which are still in force. For a good overview of the history of nuclear safety regulation, see the USNRC-website . 188 PSA studies can be aimed at three levels: Level 1: consists of an analysis of plant design and operation that is focused on accident sequences • that could lead to core melt, their basic causes and their associated frequencies; • Level 2: Consists of an analysis of the physical processes of the accident and the response of the plant and the containment system, to assess the likelihood, the time, the mode, and the quantity of radioactive releases to the environment (i.e. the conditional probability of containment failure and/or release, and accident source terms); • Level 3: Consists of an analysis of the transport of radionuclides through the environment to assess the public health and economic consequences of various accidents, ultimately arriving at the estimates of the risks of severe nuclear reactor accidents. (Source: “Probabilistic Safety Assessment and Risk-informed Decision Making (PSARID)”, Eurocourse (5-9 March, 2001), Garching). 189 The final report, released in October 1975, concluded that in comparison to other risks, including fires, explosions, toxic chemicals, dam failures, airplane crashes, earthquakes, tornadoes, and hurricanes, those from nuclear power were very small, based on a ‘traditional’ risk estimation (i.e. a multiplication of the ‘probability of an adverse event to occur (P)’ by the ‘magnitude of the adverse event (M) (e.g. number of deaths)’ gives an estimation of the ‘expected outcome of the adverse event’) (Commissie van Beraad 1976, p. 40). WASH-1400 found that transients, small break LOCA’s, and human error could be important contributors to risk. Four years later (in 1979), the Three Mile Island (TMI) accident confirmed that conclusion: the TMI accident sequence (for a concise description, see e.g. Gillon 1986, 1990) was taken up in the possible scenarios set forth in the ‘Rasmussen-report’.

110

Chapter 2

generally presented as one of the key documents in favour of nuclear power, also in Belgium (e.g. by the working group on nuclear safety of the ‘Commissie van Beraad’ (or ‘Consultation Commission’) (1976)) 190. Hence, opponents of the nuclear option and critical experts were obliged to open up this ‘black box’ in order to undermine the generally favourable conclusions the proponents were willing to advance. The debate hinged on uncertainty arguments, which, following the publication of the ‘Rasmussen-report’, were quickly articulated and almost coalesced into a ‘standard critique’, commonly known (in one guise or the other) to most commentators on the nuclear power issue191. Our analysis of the nuclear debate in Belgium also reveals how critiques formulated by US scientists, official agencies and public interest groups (e.g. the ‘Union of Concerned Scientists’, the ‘US Environmental Protection Agency’ (USEPA), the ‘US Advisory Committee on Reactor Safeguards’ (USACRS), etc.) were quickly imported into the Belgian context and put to use here. The following paragraphs set out the lines of argumentation used in the debate, each of which builds upon the one preceding it (Laes et al. 2004c, pp. 63-68; Thompson 1984, 1986). The first and basic argument is a methodological critique. Essentially, the argument boils down to the following statement: fault trees make empirical claims, but these claims can never be proven to be completely right. Since the methodology comes down to an investigation of possible accident sequences, one should be able to prove conclusively that ‘all’ possible accident sequences have been taken into account – and this is of course, as a consequence of the general fallibility of all scientific enterprise, quite impossible. Of course, even the most convinced proponents of the nuclear option would therefore never claim the infallibility of fault-tree analyses, so the argument actually comes down to a qualitative judgment concerning the remaining uncertainties. This insight is well explained by Parry (1996) in his discussion of PSA methodology. Parry makes a distinction between aleatory uncertainty and epistemic uncertainty. He explains that the aleatory uncertainty is addressed when the events or phenomena being modelled are characterised as occurring in

However, the human and organisational factors which clearly influenced the probability of the accident sequence (i.e. the safety personnel committed some errors caused by an incorrect, poorly organised and misinterpreted flow of information) was not sufficiently addressed by Rasmussen. In 1994, the French nuclear safety expert Tanguy demonstrated that the probability of an accident caused by human error was underestimated by a factor of 5-25 in the ‘Rasmussen-report’ (IBC Conference, London 1994), so that an accident of the TMI-type was not at all improbable in view of the accumulated number of operational reactor.years at that time. Since then, important lessons have been learnt and implemented concerning the training of reactor operators (e.g. through the use of simulators), safety culture, quality control, the exchange of safety-relevant information between utilities, etc. 190 This expert commission (also colloquially known as the ‘Commission of Wise Men’) was asked at the height of the nuclear controversy in Belgium (1975-1976) to evaluate energy options for the future, without however questioning earlier engagements taken by the government and the electricity utilities. These utilities had by then already placed orders for the most important engineering components of four new nuclear power plants (Doel 3+4, Tihange 2+3), which would become operational in the mid-’80s. This observation lends credence to the conclusion that the main political motive for asking the advice of a ‘Commission of Wise Men’ was to soften up increasing social tensions and gain time while the commission was working. The conclusions of the commission were to serve as the base for a parliamentary debate on energy policy, which however only took place in 1982-1983. 191 Wynne (2001) speaks of a ‘well-rehearsed’ critique of the ‘Rasmussen-report’. For an overview of such comments in the United States in the period 1975-1980, see Thompson (1986, p. 61).


111

a ‘random’ or ‘stochastic’ manner, and probabilistic models are adopted to describe their occurrences (in our terminology, ‘aleatory uncertainty’ therefore falls under the ‘risk’ category of Table 3). Epistemic uncertainty is the uncertainty associated with the analyst’s confidence in the predictions of the PSA model itself; thus, it reflects the analyst’s assessment of how well the PSA model represents the actual system being modelled (in our terminology, ‘epistemic uncertainty’ therefore falls under the ‘uncertainty’ category of Table 3). This distinction between and assessment of ‘aleatory’ and ‘epistemic’ uncertainty lies at the core of many criticisms of the ‘Rasmussen-report’. In the United States, the controversy which erupted after the publication of WASH-1400 actually led to an official USNRC assignment for a critical review of Rasmussen’s conclusions by the ‘Risk Assessment Review Group’ (RARG 1978) led by prof. Lewis (hence, also called the ‘Lewis-report’). The RARG came to a balanced conclusion: the group expressed its appreciation for the innovative character of the research carried out by Rasmussen and recognised the valuable contribution of fault-tree analysis to nuclear safety investigations, but also expressed its doubts about certain statistical analyses in the ‘Rasmussen-report’. They found that although investigators had been able to secure reliable figures on the probability for failures of many components and for human errors in nuclear reactor plants, there was a rather uneven character to the quality of the frequency data, with some frequency estimations being little more than surmise (RARG 1978, p. 11). Furthermore, the ‘Lewis-report’ found WASH-1400 to be lacking in ‘scrutibility’, thus making an essential component of the scientific process – peer review – virtually impossible. These criticisms thus did not question the legitimacy of the scientific method as such, but it did call into question the WASH-1400 derived figures on the risk of nuclear power, and it did challenge the practices of the investigators who conducted the study. As a result of these criticisms, the USNRC in 1979 publicly renounced the conclusions of the ‘Rasmussenreport’ as they were published in the executive summary. ‘Epistemic’ uncertainties continued to play a role in the debate, since most notably the contribution of the human and organisational factor to nuclear safety continue to resist an unproblematic integration in probabilistic calculations. The question then comes down to the confidence one can ascribe to the scenario-based calculations192.

192

Following the Chernobyl accident, the Dutch health council (Gezondheidsraad 1989) submitted an opinion that the results of PSA-studies concerning the frequencies of core-melt accidents with severe consequences for mankind and the environment were inevitably surrounded by large uncertainty intervals (the health council mentions a factor of 10–100). In the latest biennial FISA-meeting (10-13 Nov. 2003, Luxembourg), presenting the conclusions of the projects funded in the area of operational safety of existing nuclear installations through the EC ‘Framework 5’ programme, it was again confirmed that the main challenges for the future consist in providing adequate regulatory supervision for human and organisational factors, in increasingly competitive environments. The former have been integrated up to a certain extent in PSA-studies (based mostly on the performance of reactor operator behaviour during simulation exercises), while the latter remain very resistant to quantification. This is a significant observation, since according to Heyes (1995, p. 1031), the ‘Institute of Nuclear Power Operators’ estimated that 65% of nuclear system’s failures involve human error and 51% are caused by human error.

112

Chapter 2

These criticisms already foreshadow a second kind of uncertainty argument, aimed at the science-policy interface193. In effect, the core of this second critique concerns the intended use of the ‘Rasmussen-report’ (or, more broadly, PSA-studies) in the formation of public policy. The policy-relevant question then becomes: ‘In view of the questions raised by the need to simultaneously address issues of ‘aleatory’ and ‘epistemic’ uncertainty, which aspects of the policy development cycle can be ‘safely’ addressed by PSA-studies?’; ‘Can or should they be used to evaluate different options for electricity generation in the problem structuring phase through the use of comparative risk assessment (e.g. safety of nuclear vs. coal power plants), or to pronounce an opinion in light of the risks that are ‘generally accepted’ by society (e.g. floods, fires, etc.)?’; ‘Can or should they be used to inform policy implementation (e.g. nuclear safety regulation)?’; or even ‘Should we delay policy decisions and wait for more scientific findings?’, etc. Opponents or critics of the nuclear option have for instance brought forward the argument that the possible consequences of a large-scale accident in a nuclear power plant are qualitatively so different (e.g. due to possible long-term genetic impact on entire populations, spread of radioactive substances over a large surface area, etc.) from other (industrial) risks that it is hardly justifiable to use comparative risk assessment in order to pronounce a general opinion on the desirability of the nuclear option. While these critics admit that PSA-studies can prove their general value in optimising the design, operation, exploitation and safety analysis of nuclear power plants (e.g. by comparing between different reactor types), they reject its use as a solid basis for deriving conclusions concerning national energy policy decisions. A clear distinction was thus introduced into the (Belgian) debate between the ‘design issue’ (i.e. concerning safety measures in existing nuclear power plants or engineering designs for new reactor types) and the ‘development issue’ (i.e. concerning the further deployment of nuclear power plants), which demand, at least according to the critics, different kinds of evidence and different decision rules (Laes et al. 2004c, p. 111). Furthermore, in the Belgian debate, a clear link was established by the critics between the ‘development issue’ and the ‘siting issue’ (i.e. concerning the question were nuclear power plants should be built); this link proved to be very salient for a densely populated country such as Belgium (cf. footnote 186). The debate informed by the second kind of uncertainty arguments thus addresses the general question which form of policy commitment might be justifiably based upon the general conclusions of PSA-studies. The picture is then completed by a third kind of uncertainty argument, which was used to raise further questions about the general desirability of the nuclear option. This third type of uncertainty argument concerns an argumentation based on ‘fundamental principles of democracy’ (Leroy 1979; Laes et al. 2004c, pp. 66-68). In this argumentation, the organisation of expertise at the science-policy interface is not judged by the standards of

193

This distinction between doubts raised at the level of an intra-scientific process of criticism and doubts about the interanimation of science and the policy process is of course the result of a reconstruction ‘after the facts’. In fact, several of the critical sources on the ‘Rasmussen-report’ were completed only several years after the report ceased to function as a policy document. Nevertheless, it seems to us that the example is very instructive to the general philosophical problem of finding the ‘right way’ of dealing with uncertainty in public policy processes.


113

the ‘industrial commonwealth’ (i.e. the first type of uncertainty argument) but by those of the ‘civic commonwealth’. The point here is not that science should be replaced by voting, but simply that since policy decisions in the field of nuclear issues depend heavily on the credibility of expert advice and scientific studies, the processes through which these results are acquired should respond to the same fundamental democratic principles as any other public policy process – i.e. transparency, accountability, independence from commercial interests, etc194. Thompson (1986, p. 64) reports how the methodological flaws and irregularities of the ‘Rasmussen-report’ were construed by some opponents as evidence of the general level of competence and honesty of the entire nuclear industry, for instance by accusing scientists and policy makers alike of an intentional ‘uncritical’ use of the report in order to influence public opinion. In the Belgian context, criticisms of this kind were aimed at the rules governing the functioning of the ‘Commissie van Beraad’ (cf. footnote 190), strengthened by the general background of ‘less-than-transparent’ technocratic decision making regarding nuclear power (or the electricity sector in general)195. We give but one interesting (that is, philosophically speaking) example of an argument developed at that time by the philosopher of science Etienne Vermeersch196. Based on insights taken from the philosophy of science (Thomas Kuhn’s theory of scientific paradigms) and social psychology (Leon Festinger’s theory of cognitive dissonance), Vermeersch (1976) argues that scientists, e.g. due to public stands or the adherence to particular methods, will generally experience serious difficulties when challenged to reconsider earlier opinions. Following this line of reasoning, Vermeersch develops some guidelines for the composition and the functioning of an expert commission such as the ‘Commissie van Beraad’; and in the particular case of this commission, he concludes that (Vermeersch 1976, p. 475, our translation) 197 …in this case, one can speak of a collection of opinions, which certainly are often solid, but which are also characterised in many places by conscious or unconscious omissions, limited views and rigid frames of

194

On the transparency of typical PSA-studies Heyes (1995, p. 1033) makes the following observation: “…It is worth noting in this context, how the complexity of PSA effectively obscures the regulatory enforcement process from the view of outsiders. To simply review someone else’s PSA can be expected to cost up to half a million of dollars, performing one can require $5 million as well, of course, as requiring the cooperation of the reactor operator. This makes it next to impossible for any figure posted by the regulatory agency ever to be externally verified, and reduces the extent to which real accountability can be said to exist between the regulatory body and the public…”. 195 These arguments are explored in greater detail in Laes et al. 2004c (pp. 67-68), and range from criticisms on the ‘information monopoly’ of the electricity utilities and the associated engineering divisions, the lack of independent control on the investment plans proposed by the electricity sector, the influence of lobbying on all political levels (from the national to the communal level), the market monopoly position enjoyed by the electricity utilities (united in the ‘Beheerscomité der Elektriciteitsondernemingen’ (BCEO) – i.e. the ‘Management committee for electricity producers’) which ensured a guaranteed margin of profit on investments through the ‘cost-plus’ tariff system (therefore, the investment risk was greatly reduced), the lack of effective and independent safety control in the nuclear sector, etc. 196 The example is particularly interesting to us because rather than being aimed at the individual members of the ‘Commissie van Beraad’ (which, after all, smacks strongly of an ad hominem attack, possibly unrelated to the ‘scientific quality’ of the work done by these individuals), it targets the ‘scientific enterprise’ as such… 197 These include the confrontation of experts from different ‘schools’ with different commitments, the disclosure of possible conflicts of interests, the disclosure of public stands, the open discussion of the scientific stature of a certain assertion (e.g. derived from a generally-accepted theory, a hypothesis, speculation, informed opinion, etc.), etc.

114

Chapter 2

thinking. These opinions are worthy of respect, but should in any case be confronted with other opinions, which, if possible, are derived from other commitments…

The point here is then that, although this third type of argument indeed risks being construed as an attack on the personal integrity of experts (and therefore also as an explicit questioning of the validity of the scientific results as a deliberate misrepresentation), it is at bottom a legitimate (from the point of view of the ‘civic commonwealth’) application of an uncertainty argument. Since policy decisions in the field of nuclear power are so heavily dependent on expert opinion, it is only legitimate to ensure that the processes governing the production of expert knowledge are worthy of the standards of confidence generally required in democratic societies, all the more so since science at the frontier of public policy shows a general tendency to downplay uncertainties based on more ‘informal’ expert assessments (cf. Rip’s ‘trough of uncertainty’ argument). Wynne (1982) therefore introduced ‘trust’ as an additional factor in risk judgments. Moreover, he argues that this factor is not just a matter of ‘perception’, but that it may actually influence the ‘objective’ level of the risk; and further, that the ‘scientific’ estimation of the risk makes assumptions about these factors too (e.g. through the selection of ‘trustworthy’ experts). When trust in the process erodes, it is perfectly reasonable for policy makers or the general public to become more conservative or risk-averse. This observation is furthermore important to us because, making a jump through time, it reveals an essential characteristic of present-day questions regarding environmental policy or sustainable development in general. As Beck (1992) has convincingly argued, experts have progressively replaced our own ‘eyes, ears and noses’ as concerns risks to our health or environment, since the detection and diagnosis of these problems requires ever more intricate forms of expert intervention (e.g. measurement devices, mathematical models, calculation tools, etc.). An efficient and effective implementation of (health and environmental) policy measures therefore depends for a large part on the trust invested in the experts who have the power to structure the relevant problem features. However, it is precisely this trust in expert systems which seems to be fading in contemporary industrialised societies. Durant (1999) speaks of a ‘paradox’ of science: on the one hand, science and science-based technologies have been “…among the most powerful energising ideas of the 20th century…”, and countries which have enjoyed little science-based development seem to want more; but on the other hand, opinion polls show that the highly industrialised countries seem to be plagued by doubt and even disillusionment. This situation has been diagnosed variably as: 1. the result of a more sceptic (but not necessarily ‘anti-scientific’) attitude caused by the proof of expert failures in the past (e.g. Chernobyl, asbestos, etc.) which makes people ‘think twice’ about blindly relying on expert information (Beck 1992); 2. the result of a complex attitude of ‘virtual trust’ (Wynne 1996) – i.e. people have no choice but to rely on expert information, but they do so in a resigned way, leading to general feelings of anxiety; 3. the result of the ‘communications revolution’ (e.g. internet communications), which makes people generally more informed about all kinds of issues (Durant 1999); or


115

4. the result of lay people stressing other contextual factors (e.g. value conflicts, clashes of commercial and political interests, etc. – cf. footnote 179) besides knowledge claims in public policy issues concerning environmental or health risks. Whatever the ‘true reason’ of the distrust in expert systems (and the ‘true reason’ might very well include a mixture of all of the above), the relevant observation for our purposes seems to be one of a stalemate situation, whereby the expert-based cooperative governance scheme as a proposed solution to steer society towards a more sustainable future seems to be robbed of its very motivational basis. This observation is not meant to imply that governance can do without science, but rather that any strategy which keeps the motivational basis of its decisions well-hidden from public scrutiny is likely to run afoul of complex social reality198. 3.1.3

The contribution of expert-based governance to the sustainability debate

It is perhaps a bit ironic that the same advocacy of technocratic decision making which allowed a ‘swift and efficient’ deployment of nuclear power plants in Belgium is nowadays often advanced in the sustainability debate in circles of ‘ecological thinkers’ (we submit this term here by lack of a better one), and often with the same sense of urgency199. We discuss one influential example here, namely the so-called ‘environmental space’ (ES) approach200. The ES-approach can be characterised as an attempt to estimate the limits imposed upon our use of ‘the environment’ (since it has only limited regenerative abilities, a limited resource base and a limited carrying capacity) by the requirement that this use should be sustainable in time (i.e. for the next generations) (Opschoor and van der Ploeg 1990) – although it is also asserted by some that these limits should not be understood in an ‘objective’ sense, but rather as ‘risk thresholds’ telling us something about possible hazards for the environment and human activities if this or that particular threshold is transgressed (Van Assche 2000, p. 266). These limits (however conceived) thus demarcate a ‘vault’ in which all social activities have to take place. Biesot (1998) – in a summary volume of a decade of research – gives us a good example of an application of the ES-approach in the field of energy production and consumption. His methodology consists of the following steps, which should be repeated iteratively: firstly, one should make a demand-side analysis (an investigation of the different energy

198

Latour (2004a) distinguishes between ‘Science’ with a capital ‘S’ and ‘the sciences’ (in plural, with a small ‘s’). ‘Science’ refers to the monolithical entity speaking undisputable ‘truth’ to power, whilst ‘the sciences’ refer to an ongoing practice of constructing ‘realities’ (in plural), thereby also rendering apparent the ‘scientific apparatus’ (methods, measurement tools, etc.) involved in constructing realities. 199 Once again, it is not our intention here to make a judgment concerning e.g. the ‘irresponsible’ way science has been ‘misused’ to promote nuclear power, or to show how ecologists ‘abuse’ science for political purposes. Rather, we want to show how in both cases, very similar uncertainty arguments can be made and derive from these observations a more systematically justified approach to science for policy making. 200 Different indicators and approaches fall under this broad characterisation, e.g. the ‘Factor 4’ approach, ‘ecological footprints’, ‘environmental rucksack’, ‘MIPS’ (‘Material Input Per Service unit’), etc. – for an overview and a critical assessment, see Chapter 4 on “Environmental indicators of sustainable development” in Moffatt et al. (2001).

116

Chapter 2

services required by individuals in a particular region); secondly, one should investigate along the entire chain of production the material and energy resources mobilised in order to provide that particular energy service (‘from the cradle to the grave’, i.e. from resource extraction to the final waste streams); thirdly, one should look for opportunities to reuse or recycle materials along the different production chains (the so-called ‘quality cascades’ – e.g. used wood can be recycled as paper / used paper can be recycled as cardboard / etc.); lastly, the resulting environmental pressures of all the production chains are compared to the available ES. This last step is often combined with a ‘backcasting approach’ – i.e. starting from a ‘sustainable’ future situation (commonly the horizon 2050 is used) ensuring an equal share of the available ‘environmental space’ for everyone, different possible scenarios for attaining this future are investigated. Not surprisingly, these scenarios include some far-reaching breaks with the ‘business-as-usual’ situation, e.g. in the form of accelerated efficiency gains, the implementation of sufficiency strategies and the accelerated penetration of new technologies. Van Assche (2000, p. 267) gives an example of such normative criteria for a global sustainable energy future: firstly, each citizen of the world should be able to enjoy an equal level of energy services (by the year 2050); secondly, these energy services should be provided for by renewable energy sources (somewhere between 2050 and 2100, depending on the magnitude of fossil fuel reserves); and lastly, in the meantime (as long as the sustainable situation is not attained), fossil fuels can be burned to the extent that they do not present unacceptable risks to the climate system. Other goals are possible, but they should always be almost ‘self-evident’: e.g. survival of the human species or the ecosystem, simple equality, fulfilling (universally accepted) human needs, preventing social disasters (e.g. when resources run out), etc. It is then up to the ‘political system’ to implement the necessary changes, e.g. through direct (norms and sanctions), indirect (subsidies, taxes, etc.) or one-sided communicative regulation (education, technology transfer, etc.). We will not engage here in a detailed critique of the ES-approach, but already from the brief introduction given above, it will be clear to the reader that the approach is firmly situated within the logic of the ‘industrial commonwealth’: the (social, economic) system has to be ‘steered’ within boundaries through a rationalisation of resource efficiency and human (including social and political) behaviour, based on exigencies of either ‘system survival’ as such, or principles of simple equality from other commonwealths (e.g. equal access for all to the same level of energy services) – thereby simplifying, reducing or simply not addressing the likely tensions arising from the fact that complex democratic societies depend on a careful balancing of different commonwealth exigencies. It will come as no surprise that the ES-approach has attracted severe criticisms, e.g. concerning the treatment of equity, value judgments in the aggregation of different material streams, the adequate spatio-temporal scale to use, the supposed or desired link with policy making, etc.201. We will limit ourselves here to some examples, which will be systematically 201

For a more in-depth treatment, we refer the reader to Korthals (1994), van den Bergh and Verbruggen (1999) and Moffatt et al. (2001). Moreover, some adherents of the ES-approach also openly discuss the normative presumptions, the methodological weaknesses involved and the necessary synergies with other methods – see e.g. De Jonge et al. (2000).


117

explored in the logic of the ‘commonwealth model’, also in order to show the symmetries with the case of nuclear safety regulation explored above.

•

•

•

•

202

Essentially, criticisms in both cases come down to: Doubting the solidity of the mobilised ‘entities’ within the ‘industrial commonwealth’: the ‘entities’ mobilised by the experts (e.g. models, accounting methods, etc.) reveal more uncertainty in their functioning than the experts are willing to accept or admit – e.g. Korthals’s (1994) argument that governments no longer can fulfil the ‘central’ systemic steering role required by the ES-approach; Doubting the solidity of the mobilised ‘entities’ when translated to other commonwealths: the ‘entities’ are shown to be ‘hybrids’ which need to function also in other action logics, e.g. the argument for the provision of ‘contingency plans’ in case expert reasoning proves to be wrong when applied in the ‘real world’, especially considering the far-reaching social changes advocated by the ES-approach; or the argument that real indeterminacy, inasmuch as it results from open-endedness in the sense that the outcomes depend on how the intermediate actors will behave, can only be addressed by involving these actors themselves (Wynne 1992); The ‘subjects’ imbued with grandeur in the ‘industrial commonwealth’(i.e. the experts) are ‘de-masked’ as having interests in other commonwealths: e.g. by questioning the ‘political’ role of experts202; Other commonwealths are mobilised in order to dispute central value questions: e.g. by questioning the priority of the ‘simple’ strictly egalitarian approach assumed in the ESmethodology203, or Korthals’s (1994, pp. 39-41) objections to the increasing ‘pastoral power’ and government interventions in daily life.

Such political commitment is, in the case of ‘green’ scientists, however often covered with the cloak of charity. Van den Bergh and Verbruggen (1999, p. 70) conclude their (otherwise sharp) critique of the ‘ecological footprint’ method (developed by Wackernagel and Rees (1996)) with the bland words: “…We are in sympathy with Wackernagel and Rees in their concern about the impacts of humans on natural systems, as well as with their effort to construct an indicator using an explicit accounting framework and a detailed database. Nevertheless…”. Latour (2004a) mentions in a footnote to his book (footnote 15, p. 255) that he is “…well aware that there is no lack of good reasons that would make it possible to explain why, in the heat of the new battle, ecological thinkers have not devoted all their strength to discussing the political nature of nature. Like Sartre before them, they did not want to dishearten the proletariat by beginning to doubt the Science that seemed to them to serve as the indispensable lever for public emotion. This ‘strategic naturalism’ allowed them to turn these famous ineluctable laws of nature against their enemies. Their tactics may have been good ones at war, and it is somewhat unfair to criticise them for this expedient use of nature, but it still remains bad political philosophy…”. 203 Cowell (1995) mentions at least nine different interpretations of ‘equity’, each of course having different consequences for the appropriate measurement of equity unbalances: • The 100% approach: a completely equitable distribution over persons – each person receiving an equal amount of the good in question (i.e. le principe de commune humanité in Boltanski and Thévenot’s idiom); • The social minimum: ensuring that no-one receives less of a particular good than required by a minimal standard (i.e. le principe de commune dignité in Boltanski and Thévenot’s idiom); • Lifetime equity: measures equity and inequity in the enjoyment of a particular good based on an entire lifetime; • The mobility approach: aimed at removing structural barriers for the attainment of higher levels of equity (i.e. grandeur can never be the result of a permanent fixture in Boltanski and Thévenot’s idiom);

118

Chapter 2

Our point of course is not to reject the use of such normative analyses or to claim that they are ‘wrong’ or ‘methodologically flawed’, but rather that these methodological approaches and their underlying normative questions should be openly debated, possibly confronted with other approaches, and justified within a larger circle of stakeholders – the ‘good common world’ (Latour 2004a) should be a shared construction effort, not to be closed in advance by (quasi-)scientific ‘facts’. 3.1.4

Conclusion

As set out in the introduction to this chapter, our interest in the investigation of existing emergent governance practice lies not merely in providing criticism, but chiefly in reflecting on the quality of these practices and outcomes as a first step towards ‘better’ practice. So, what can be learnt from our account of expert-based governance? A reasonable conclusion to put forward is that expert-based governance has proved to be an excellent cooperative scheme for providing swift and efficient solutions – hence no doubt its popularity in the sense that it truly has become a ‘dominant cooperative scheme’ in highly industrialised societies. This success results from an inherent aptitude for providing ‘closure’ to a debate, whether on an epistemic level (e.g. reducing the more intractable forms of uncertainty to the general category of risk), on the scope of the problems addressed (e.g. implicitly structuring the political agenda in terms of ‘measurable phenomena’), or on a social level (e.g. inclusion/exclusion mechanisms regarding those who are ‘able to speak’ because of their expertise). But there are costs and risks associated with achieving tractability (as witnessed in the nuclear safety case), because sometimes problems are reduced rather than solved them, and what has been ‘reduced away’ may return at an inopportune moment204. One can try to separate ‘facts’ from ‘values’ and ‘interests’ in advance, but this cannot be a general default solution. From a normative point of view, the choice between interest in maximising the technical efficiency of problem-solving, and the interest in maximising actors’ rights to have a say (based on fundamental understandings of democratic principles such as transparency, openness, the right to information, etc.) is ultimately a political one, which should always be justifiable

•

The inclusive approach: aimed at avoiding excessive levels of inequity in the enjoyment of a particular good (i.e. based on le principe du bien commun in Boltanski and Thévenot’s idiom); • The relative approach: aimed at improving the conditions of the most disadvantaged group (e.g. increasing the income of the lowest decile of a population); • The ‘upper limit’ approach: aimed at lowering the enjoyment of a particular good of the most advantaged groups in society (i.e. questioning whether these groups – based on le principe d’investissement in Boltanski and Thévenot’s idiom – have they really done enough to earn their state of ‘grandeur’?); • The avoidance of privilege: aimed at combating unfair privileges in education, political rights, employment opportunities, etc. insofar as they are the result of an unfair distribution of income (i.e. removing the effects of a transport de misère in Boltanski and Thévenot’s idiom); • The ‘benchmarking’ approach: aimed at an equitable distribution of a particular good based on a comparison with a ‘similar’ nation. 204 Of course, this moment not necessarily falls within the term of office of the politician responsible for the ‘reducing’. As a testimony to the widespread recognition of this political strategy, an abbreviation is now commonly used to denounce it: ‘NIMTOO’ or ‘Not In My Turn Of Office’…


119

(and challengeable) under the banner of ‘the common good’205. Furthermore, we maintain that this solution is also deficient from an analytic point of view: thinking the expert-based governance scheme through to its utmost consequences, the ‘best’ solution would be to draw the borders of the ‘fact-finding’ process ever more tightly before it enters the domain of ‘meaning’ or ‘valuing’; but in doing so, politics remains essentially blind to the inevitable value-based choices going on in the sphere of science. In fact, there is now growing evidence that in the field of technology policy, where decisions often have to be made under conditions of radical uncertainty, we are at present experiencing a ‘critical juncture’ whereby increasingly the weaknesses of the dominant cooperative expert-based scheme are exposed (e.g. its failure to achieve public assent to science-led policies) – hence, an increasing interest in new approaches to knowledge, public deliberation and governance (see e.g. Pellizzoni 2003)206. The precautionary principle, which explicitly accommodates recognition of scientific uncertainty and the mechanisms for translating this uncertainty into policy decisions as a problem, has emerged as a principle for decision making at the international, European and national level. In a sense, it can be said that the ‘archetypal’ expert, associated with notions of social and rational control, is increasingly put under pressure to become a more ‘flexible’ actor regarding their approach to knowledge production and utilisation. However, there is still significant evidence that this recognition is far from widespread; experts often remain unwilling to accept that scientific analyses are replete with normative judgments that can (or should be) opened up to wider scrutiny (see e.g. Fischer 1999) 207. Thus, while it is clear that ‘sound’ scientific analysis forms a necessary contribution to decision making in the field of sustainable development (and in chapter 5 we will even

205

Precisely on this point, Jänicke (1990, p. 27) observes what he has called a “…steady deterioration of the control ratio between politics and the machinery of government…”. That is, Jänicke criticises the growing disjuncture between those who make the decisions, those who are politically responsible for them, and those who are affected by them. The influence of the democratically elected legislature over the state administration must pass through what he calls “…the needle eye of ministerial responsibility…”. Indeed, one of the core points of the general green critique of the administrative state is that it makes something of a mockery of the liberal democratic ideal of public accountability (Eckersley 2004, p. 88). 206 One can also point out a raft of new discursive designs that have already emerged as partial antidotes to the technocratic dimensions of governance structures, such as third-party litigation rights, environmental and technology impact assessment, citizens’ juries, consensus conferences, and public environmental enquiries. Each of these initiatives might, to some extent, be understood as attempts to confront both public and private power with its consequences, to ‘widen the range of voices’, to prevent or expose problem displacement, etc. (Eckersley 2004, p. 92). 207 The tight connection between science (and especially the social sciences) and political practice was already analysed by some ‘forerunners’ at the beginning of the 20th century. Otto Neurath (a logical empiricist, member of the famous ‘Vienna Circle’ – a group of mainly natural scientists with a philosophical interest in the methodology of science) for instance has devoted a large part of his oeuvre to the historically contingent and non-neutral character of the sciences, which according to him explains its political relevance. Neurath proved to be very sceptical about the particular conception of ‘science speaking truth to power’: “…The assumption that scholars enjoy a kind of social extra-territoriality is above all a product of a period which was inclined to accord an exceptional position to scholars, as substitute priests, and was ready to use scientific assessments as a basis for taking political measures; but this was done not because politicians wished to be scientific but because they knew that scholars are ultimately politicians…” (quoted in Deblonde 2001, p. 95). This excerpt was written in 1931, but it seems to us that it still carries great relevance for present conditions…

120

Chapter 2

argue for a strengthening of the role of independent scientific research), on its own it cannot be sufficient, both in a normative and in a pragmatic sense. Scientific knowledge, when deployed as public authority, is not just a private matter for scientific bodies first autonomously to resolve, define, or otherwise interpret on behalf of the public policy domain, before it is rendered visible to the latter. Achieving ‘integration of the economy, society and the environment’ (which is often presented as the key question for sustainability) is an inherently political process (in the broad sense of ‘constructing a common world’) that is not just guided by scientific judgments about environmental limits208. Perhaps the most important point then to emerge from these theoretical reflections and the discussion of some examples concerning the production of scientific knowledge (involving the reduction of uncertainty) and its relation with governance is that far from abandoning ‘sound’ science, the notion of ‘soundness’ should indeed be radicalised. Scientific soundness roughly means adopting approaches that are transparent, systematic, peer-reviewed, accountable, independent, capable of learning – but mostly in a context of immediate colleagues or ‘peers’. ‘Sound science’ as a basis for a societal project such as sustainable development is then confronted with the challenge of upholding this ‘soundness’ in front of a larger (and often critical) constituency. Science has to become an integral (and not a separate) part of a science-policy arrangement. This means that the question of ‘soundness’ (i.e. ‘how sound can you be, and should you be?’) cannot be kept out of the ‘political’ discussion, but becomes as much a practical organisational challenge (through processes of quality assurance) as an epistemological issue. In particular, the level of soundness required will depend on the prior framing of the policy issue at stake (problem structuring), which means that these wider assumptions and commitments should be rendered explicit and be democratically debated as far as possible. What shape such an arrangement could take will be set out in chapter 5.

3.2

3.2.1

Governance by aggregation


The preceding section taught us that the ultimate weakness of the expert-based governance discourse lies in its inability to openly deal with ‘values’. Such ‘values’ are either smuggled furtively by the expert into scientific analysis, and consequently into the policy process, or are considered to be self-evident. At best (and mostly grudgingly), the dependence of scientific advice on value commitments is admitted, and it is then left over to deliberations in the ‘political sphere’ to decide upon the most salient ones – thereby leaving the scientific analysis itself untouched by ‘politics’. Even if not looked upon as

208

For example, approaches that envisage integration as being achieved in the manner of a ‘Russian doll’ model, in which ‘the environment’ successively encapsulates ‘society’ and ‘the economy’ (e.g. the ESapproach), will always be contested because they conceal the processes contributing to the production and recognition of environmental limits and to their solution (cf. Chapter 0).


121

utterly irrational or fully determined by the interplay of powerful interests, ‘hard’ science (that is, based on a self-understanding of scientific enterprise situated within the logic of the ‘industrial commonwealth’), although it would like to see its ‘informal’ influence on the policy sphere maximised, nevertheless is fundamentally (on the basis of its own selfunderstanding) unable to provide guidance on the process of value selection taking place in the ‘political sphere’. It is precisely this gap that the model of governance by aggregation, which we will discuss in the following sections, aims to fill in. The fundamental idea behind the different possible manifestations of ‘governance by aggregation’ is that policy should be guided by an aggregation of ‘preferences’ or ‘values’ expressed by citizens. Broadly, with aggregation we mean to indicate a procedure that renders policy problems tractable by either • defining socially-relevant ‘intervals’ that can be added or substracted to make valid choices (a prime example would be the organisation of a direct vote); or • defining socially-relevant ‘ratios’, so that it is possible to combine quantities and values of diverse entities (a prime example would be the market pricing mechanism). However, many different manifestations and variations to these strategies exist, which are mainly set apart by their intended scope or depth of penetration in the policy development cycle209. A very far-reaching interpretation sees ‘governance by aggregation’ as the fundament upon which almost the entire edifice of the liberal democratic states should be built. The basic idea is then that under modern conditions, notions like ‘the common good’ and/or ‘the general will’ have to be relinquished and that the pluralism of values and interests had to be acknowledged as coextensive with the notion of ‘the people’ (Mouffe 1999). Moreover, given that in this view self-interest is what motivates individuals to act, these theories declare that it is interests and preferences that should constitute the lines over which political conflict should take place and provide the matter over which bargaining or

209

This variation of scope is supported by different versions of (philosophical) liberalism. Nevertheless, all of these versions share some fundamental insights. Philosophical liberalism departs from a normative ideal of free and autonomous individuals. In order to realise this ideal, liberalism posits a number of basic principles. One is that the state should be neutral with regard to private matters, based on the belief that the state should not judge between the different conceptions of ‘the good life’ (e.g. religious choices) held by its citizens, or at least that the state should guarantee equal liberties for each individual citizen necessary to fulfil these visions of ‘the good life’. Liberalism does not deny that encompassing views on ‘the good’ are still possible (it remains ‘agnostic’ on this issue), but it takes the position that these should not be at the core of political culture. Therefore, the public forum should be organised according to principles which are unbiased towards particular visions of ‘the good’, and should be able to produce a broad consensus on the necessary conditions for the fulfilment of individual versions of ‘the good life’ (whatever they are). This principle of neutrality is sometimes also referred to as a principle of ‘exclusion’, because it makes clear that some arguments should be barred from entry into the political debate as a foundation for public policy (e.g. measures that can only be justified from one particular ideological or religious point of view). In some theories this implies a strict and ‘empty’ neutrality; others promote a more ‘active’ form of neutrality, but in any case, even when the state apparatus promotes particular broadly-shared ideals (e.g. education), it should remain neutral with regard to the concrete interpretation of these ideals. Another basic principle of liberalism is the ‘principle of damage’ and the separation of the public and the private sphere. The ‘principle of damage’ posits that every citizen is in principle free to act as he or she wants, as long as this action does not entail damages for other citizens. Only in the latter case can the state ‘invade’ into the private sphere in order to limit liberties or take corrective measures (Mortier and Raes 1997).

122

Chapter 2

voting (two prominent forms of aggregation) could take place. For a number of reasons, we will take issue here with a more limited version of ‘governance by aggregation’, that is one that sets apart a separate domain for aggregation procedures. One reason is that criticisms on this more limited interpretation of ‘governance by aggregation’ will a fortiori hold true for a more extended version. Moreover, it can be shown on a theoretical level that the self-understanding of constitutional liberal democratic states cannot be grounded solely upon motivational assumptions of self-interested behaviour (Habermas 1996, pp. 290-295). More limited versions of ‘governance by aggregation’ are supported by attempts to provide a solid basis of allegiance to liberal democracy by reconciling the idea of popular sovereignty with the defence of liberal institutions. Their aim is not to relinquish liberalism but to recover its moral dimension and establish a close link between liberal values and democracy. Thus, these theories affirm that, despite the ‘fact of pluralism’, it is nevertheless possible to reach a consensus that would go deeper than a mere ‘agreement on procedures of aggregation’ – a consensus that could qualify as ‘moral’. One of the most authoritative voices of this version of philosophical liberalism is undoubtedly John Rawls210. Rawls has always opposed the utilitarian idea of society as a goal-oriented machine, aimed at maximising people’s preferences; instead, he has built his theories around the idea of society as a structured collection of deontological rules that constitute societal identity and reality211. Rawls is concerned with seeking and making explicit rules of justice as the ‘first virtue of social institutions, as truth is of systems of thought’ (Rawls 1971, p. 3). However, these deontological rules of justice apply only to the ‘basic structure’ of society (i.e. the political constitution and the principal economic and social arrangement) – the structure that regulates the distribution of ‘primary social goods’ (i.e. basic liberties, freedom of mobility and professional choice, the advantages and power connected to public offices and responsible positions; income and property; and the social fundaments of self respect). Thus, Rawls actually seeks a compromise between the utilitarian and the deontological framework: only in the case of the ‘primary social goods’ comparison between individuals is accepted (because these goods are so ‘basic’ that everyone needs them, regardless of a particular ‘design for life’) and deontological rules of a just distribution are formulated212. The ‘primary social goods’ are an expression of a ‘thin 210

Our choice for Rawls’s theory is not only guided by its profound impact on political philosophy, but also because Rawls’s ideas have been used as a platform for tackling typical sustainability problems, most notably the problem of intergenerational equity (see e.g. Gosseries 2001; Damveld 2003; Naert 2004). The following paragraph is mainly inspired on the chapter on Rawls in Mortier and Raes (1997, pp. 149-165). We have also greatly benefited from the thoughtful exposition of Rawls’s ideas in Naert (2004). 211 Again, many variants of utilitarianism exist, but all share the following characteristics: utilitarians assume a neutral position with regard to the substantive choices made by people, and are only interested in the consequences of these choices in terms of the augmentation or diminution of total or average (feelings of) wellbeing. In other words, utilitarians pronounce no judgment on the intrinsic value of these choices, but rather use an extrinsic, instrumental standard of judgment. 212 These are: “…Firstly: each person is to have an equal right to the most extensive total system of equal basic liberties compatible with a similar system of liberties for all. Secondly: social and economic inequalities are to be arranged so that they are both: a) to the greatest benefit of the least advantaged, consistent with the just savings principle; and b) attached to positions and offices open to all under conditions of fair equality of opportunity…” (Rawls 1971, p. 266).


123

theory of the good’ – i.e. a kind of ‘minimal morality’. Thus, ‘the right’ in Rawls’s theory gets a limited substantial interpretation so that every individual would be able to realise his or her vision of ‘the good’. In his work “Political Liberalism” (Rawls 1993) – for a large part an ‘update’ and a reformulation of his most important arguments as set out in the seminal “A Theory of Justice” (Rawls 1971) – Rawls is mainly concerned with situating his theory firmly within the political practice of present-day democratic constitutional states. His theory of justice is hence presented as historically and contextually situated, whereas in “A Theory of Justice” the theory was rather presented sub specie aeternitate. Rawls is still of the opinion that, despite fundamental differences of value perspectives (the ‘fact of pluralism’), policy making in pluralistic societies needs to be grounded in a shared conception of justice. Contrary to “A Theory of Justice” however, these basic principles of justice are discovered in a search for an ‘overlapping consensus’ on how one should best handle the different value perspectives. This ‘overlapping consensus’ is to be found in the ‘background political culture’ of constitutional democratic societies – i.e. fundamental ideas or principles, as found in documents such as the American Declaration of Independence, jurisdiction, or ideas about society as a ‘fair system of cooperation’. Rawls calls his method ‘political constructivism’. What justifies the particular conception of justice (the basic principles themselves are not substantially changed in comparison to “A Theory of Justice”) is thus not its being true to an order antecedent to and given to people, but its congruence with the deeper understanding of citizens of constitutional democracies, and the realisation that given a certain historical development and given the traditions embedded in political life, this is the most reasonable doctrine to follow (Mortier and Raes 1997, p. 164). Rawls is referring here to liberal distinctions between ‘public’ and ‘private’, ‘justice’ and ‘the good’ – distinctions that emerged in a history of power struggles and compromises and that are now solidly inscribed in the liberal democratic state. Free, equal, rational and reasonable citizens can generally agree with these principles, although they will still justify them based on precepts taken from their own ‘reasonable comprehensive doctrines’. Rawls thus sees this ‘overlapping consensus’ as the basis, and not as the result, of public debates, for questioning this consensus (e.g. by inquiring into the reasons for supporting it) might lead to unproductive and even insurmountable or destructive value conflicts between the different ‘comprehensive doctrines’213. The substantive content of public reason is therefore limited to the rather abstract level of constitutional essentials and questions of basic justice – i.e. fundamental rights and freedoms, and the justification of socio-economic disparities (Rawls 1993, p. 214). Questions of environmental (i.e. the distribution of environmental and health risks) and intergenerational justice (which were 213

Rawls (1993) does concede that, as long as society is not ‘well-ordered’, arguments derived from a particular worldview can be brought into a public debate. He gives the historical example of arguments used against slavery, which was considered by the abolitionist movement to be ‘against the law of god’. With hindsight, this ‘non-political’ argument (i.e. stemming from a particular religious point of view) proved to be necessary to abolish slavery. However, Rawls is of course only able to make this judgment post factum – who is to say whether an argument derived from an at that time ‘private’ worldview will prove to be ‘political’ in the future?

124

Chapter 2

addressed to some extent by Rawls in his ‘principle of just savings’ – cf. footnote 237) were however considered by Rawls to be ‘problems of extension’ of the theory of justice – problems which were indeed taken up by other authors, but as we will show in section 3.2.2, with limited success. Outside the realm of public reason, the clash of private, non-generalisable preferences prevails. Without recourse to generally accepted ‘reasonable’ principles for a substantial conflict resolution, questions involving a need for collective coordination and action can only be settled in terms of procedures enabling an aggregation of preferences, leading to a modus vivendi214. For instance, some liberals might argue that concerns for sustainability represent no significantly different challenge to liberal democratic states than those already posed by ‘the fact of pluralism’ – that is, the problem is simply one of competing human preferences. Therefore, the argument goes, if, after engaging in lawful means of persuasion and utilising all available liberal civil and political rights (e.g. holding public meetings, campaigning, organising demonstrations, bringing legal actions, standing for political office as a ‘green’ candidate, voting, etc.) an effective majority for sustainability policies cannot be mustered at crucial decision moments (e.g. general elections, parliamentary votes, etc.) then sustainability advocates must simply learn to live with the outcomes. The point is then simply that the liberal democratic state cannot, and ought not, put forward any particular interpretation of sustainability – just as it has to refrain from imposing any other version of ‘the good life’. Besides the fact that a simple reliance on the majority vote of ‘the people’ is unlikely to bring about (through the election of representatives) even the ‘thin conception of sustainability’ we set out in chapter 1 (simply because other issues will most likely compete for the public attention in times of elections), there are more principal arguments possible against this reliance on the ‘political market’215. Another straightforward possibility of aggregation is strategic negotiation. All groups with a sufficiently large interest in the issue at stake216 (and thus, presumed to be sufficiently organised and sufficiently powerful) are then invited to the negotiation table; the result of 214

Perhaps it is also the right moment here to point out some of the crucial insights of the Nobel prize-winning economist Kenneth Arrow. Arrow (1951) argues that one cannot rationally aggregate individual preference structures into a single joint preference structure. This ‘Impossibility Theorem’, still generally acknowledged to be the outstanding problem in the philosophy of economics, is that, given certain assumptions (‘reasonably acceptable’ to society at large), there can be no ideally rational aggregation device. These assumptions are 1) unrestricted scope of the device (or universality); 2) the Pareto principle; 3) non-imposition or citizen sovereignty; 4) non-dictatorship and 5) independence of irrelevant alternatives. Arrow’s theorem says that if the responsible decision-making body has at least two members and at least three options to decide among, then it is impossible to design a social choice function that satisfies all these conditions at once. 215 We will not go into a detailed critique of the conception of politicians as ‘entrepreneurs’ in a ‘political market’ aiming to gain votes (see Hisschemöller 1993, pp. 95-96), but the general argument developed under section 3.2.2 can be transposed with some minor modifications. 216 Negotiation is typically encountered in problems of distribution. In that case, policy goals (‘the common good’) are broadly accepted, or, in any case, non-negotiable; the issue is to realise these goals within the limits of acceptable costs for society. A necessary condition for policy by negotiation to work properly is that costs and benefits of policy alternatives have to be reasonably well-known. For example, in the context of sustainability, the problem of distributing greenhouse gas reduction efforts for a given overall national reduction target (for Belgium, - 7.5 % compared to 1990 levels in the period 2008-2012) over different sectors (industry, transport, households, services) and regions (the Flanders, Brussels, and Walloon region) is a typical negotiation problem.


125

the negotiation will be a ‘local bargain’ (in Boltanski and Thévenot’s terminology) between the represented interests. Power plays a major role of course, but those who come to the negotiation table have good reasons for accepting some democratic rules. They ensure that confrontation among each position may take place in commonly acceptable conditions, allowing the expression of interests while reducing the possibility of reciprocal oppression and humiliation. This aggregation strategy obviously faces difficulties because of its inability to address the more diffuse interests of the citizens not sufficiently organised to put their demands on the political table – one then has to assume that the parties doing the negotiation are sufficiently representative of the interests in society at large. Market-based governance aims to accommodate this shortcoming by taking into account the preferences of all those who stand to gain or lose from a specific governance initiative. To do so, preferences of those concerned should of course be made comparable on an equal basis. The market-based governance model therefore assumes that these preferences are revealed in actual or hypothetical market situations. Because there seems to exist a large (combined scientific/political) interest for the application of the market-based governance model in the domain of sustainable energy choices (the EC-funded ‘ExternE’ project), the next chapter (Chapter 3) will be entirely devoted to a detailed methodological reconstruction of this particular application of the model. In the next section, we will limit ourselves to a discussion on a more general level. 3.2.2

The contribution of governance by aggregation to the sustainability debate

In our discussion of market-based governance (as one possible form of governance by aggregation), we do not want to focus on all the respective pros and cons raised against this method when applied in the context of sustainability217. As we will show, there are substantial difficulties which cannot (and generally are not) denied by even the most neoclassical economist or full-blooded utilitarian. Conflicts often seem to arise rather from a profound (perhaps even intentional) misunderstanding of the respective positions taken by neo-classical economists and other perspectives (often loosely referred to as ‘ecological economists’) than from a willingness to learn from other insights. Rather, we will therefore ask the question whether market-based governance is, based on the weight of the arguments, so seriously deficient that it ought to be discarded altogether. If not, our task will be finding an adequate use for the model in our overall search for governance for sustainability. From the point of view of political theory the principal idea of market-based governance rests on the assumption that a model which may be appropriate (or in any case widely used) in the case of private goods is also appropriate for public ones. Broadly speaking, this normative political theory can be said to be grounded in four principal hypotheses (Jacobs 1997, p. 212): 217

There is a very substantial literature on the subject, so it would be impossible to deal with every argument in detail in a few paragraphs. Good overviews (and personal assessments of the merits of the different arguments) can be found in Schrader-Frechette (1985), Deblonde (2001) and Mofatt et al. (2001); and in the context of energy policy, Stirling (1997).

126

Chapter 2

1. In social behaviour, human beings can be represented as separate, autonomous individuals seeking to satisfy preferences of variable importance to them. Moreover, the means needed to satisfy these preferences (the ‘commodities’ in a market situation) are considered to be limited (i.e. individuals act under the condition of ‘relative scarcity’), and these scarce means are capable of alternative application; 2. These preferences are exogenously determined (they are ‘givens’) and ethically unchallengeable; 3. The role of collective choice institutions is to discover these preferences and aggregate them to give the overall preference of society; 4. The optimal public decision is the one which maximises the total preference satisfaction (benefit over cost) of all individuals. When the first and second conditions are fulfilled, human behaviour will necessarily take the form of ‘rational economic choice’ (Robbins 1984). As stated above (and explained further), in the context of policy recommendations for sustainable development usually not all of these hypotheses are accepted as a whole. Most notably the fourth hypothesis is nowadays not advocated; a common feature of market-based governance recommendations for sustainability is usually the maximisation of social welfare subject to some intertemporal sustainability rule and possibly additional normative constraints on the exploitation of natural resources (and thus, the argument that neo-classical approaches to sustainability are yet another expression of a ‘cold-hearted’ utilitarian attitude does not hold). Nevertheless, the other hypotheses have also provoked an intense debate. For the sake of convenience, we will discuss these criticisms (although they are intimately related) under two separate headings. Broadly speaking, the first critique (the ‘critique of inadequate representation’), mainly directed at the first and second hypothesis, claims that the market-based governance approach is not capable of adequately representing (and thus, defending) people’s attitudes or expectancies towards sustainable development, and thus should be discarded. The second critique (the ‘institutional critique’), mainly directed at the third and fourth hypothesis, claims that an application of the market-based governance approach to sustainability issues implies an (ethically) inappropriate transposition of decision-making models valid in ‘markets’ to social institutions in general, which should be avoided at all cost. 3.2.2.1 The critique of inadequate representation To start of the discussion of this critique on the right foot, we believe it to be absolutely necessary to first burn down a straw man erected by a certain ecological critique on the market-based approach (and liberalism in general – at least if liberalism is understood as fundamentally entailing an embrace of the first and second hypothesis concerning human behaviour mentioned above). A concise version of this critique can for instance be found in Eckersley (2004, p. 104): …The fundamental problem with the ideal of liberal autonomy is that it rests on an incoherent and undesirable ontology – that of social and biological detachment. Given that ontology precedes ethics (i.e. underlying assumptions about being and reality constrain the field of ethical possibilities), it is necessary to question these basic liberal conceptions of the self before it is possible to rethink what autonomy might mean in a new ecological age…


127

She then goes on to point out that by upholding this image of the ‘self-contained rational maximiser’ denies the “…non-instrumental dependency on ecosystems and the biological world in general…” (p. 104) and the “…fundamentally social character of individual conceptions of the good…” (p. 105). In fact, Eckersley is revisiting (with the addition of some ecological accents) a rather old debate between liberal thinkers and the so-called communitarians, who maintain that community is a structural precondition for human selfhood and moral agency218. Taylor (1996, pp. 15-52) – himself a communitarian – has suggested a way out of the debate by positing a difference between two levels: one ontological, the other practical. On each of these levels, ‘liberal’ or ‘communitarian’ points of view can be accepted, but – and this is the crux of the distinction – ‘ontological communitarianism’ can go hand in hand with ‘practical liberalism’ (i.e. notwithstanding an acceptance of community influence on individual definitions of ‘the good life’, individual rights can be defended over ‘the good of the community’ as a matter of pragmatics). Taylor for instance goes on to declare himself a communitarian ontologist and a liberal pragmatist. This distinction, though useful, still cannot resolve all difficulties219. We suggest dropping the ‘ontological’ level altogether and instead replacing it by our notion of a ‘dominant cooperative scheme’. Accepting a constructivist point of view implies that any reference to an ‘ontological’ level should be exposed for what it really is: an attempt to settle once and for all an ineradicable political tension (Mouffe 1999)220. Moreover, accepting ‘ontologies’ (which are of course non-discussable) puts undue constraints on the political debate, whereas a ‘dominant cooperative scheme’ remains principally open to challenge221. Restating the tension in these terms allows to bring out the political nature of conflict (as will be discussed in section 3.2.2.2), instead of being presented as ‘ontological’ requirements. While this clarification seems necessary to us on the general level of liberal tenets as such, this is a fortiori the case for the more restricted market-based governance approaches. In fact, advocates of these approaches generally make much more modest pragmatic claims

218

A good summary of the debate can be found in Kukathas and Pettit (1990, pp. 74-118). Naert (2004, p. 31) points out how an attempt to characterise Rawls’s theory of justice (reformulated as a form of political liberalism) on these levels leads to some ambiguities. On the ‘ontological’ level, on the one hand Rawls seems to reduce the basic structures of society to a simple aid for the satisfaction of privately-held preferences, based on an agreement between rational ‘mutually disinterested’ individuals, but on the other hand he characterises this agreement as the result of an artificial construction invented by citizens situated in a political community (with a shared conception of free and equal relations). On the ‘practical’ level, on the one hand a defence of equal individual rights and freedoms is formulated, but on the other hand these basic rights constitute a shared ‘good’ of political culture. 220 Yes, people are ‘dependent on the social environment for their selfhood and value-commitments’; yes, people are ‘biological beings’; yes, people act as ‘rational egoists’; but it depends on the context, and more importantly, we have no transcendent a-historical (rational, metaphysic, ontological, etc.) standard by which to measure and compare people’s behaviour. 221 ‘Dominant cooperative schemes’ might however become so dominant that they are not recognised as such – at that point, they indeed start to resemble ‘ontologies’. 219

128

Chapter 2

than ‘metaphysical’ or even strictly ‘empirical’ ones222. The battle is therefore often waged on the wrong battlefield. Far from being metaphysical, the first two hypotheses should best be regarded as ‘pure methodological fictions’. Hodgson (1997, p. 49) mentions how right from the start (e.g. already by John Stuart Mill in the middle of the 19th century), the assumption of utility-maximising ‘economic man’ was recognised as unrealistic but nevertheless as an allegedly acceptable abstraction to enable economic science. Subsequently, Hodgson continues, the ‘fiction’ of humans as utility-maximising machines is defended, not on the basis of strict empirical validity, but because it represents a useful foundation for theories that would make valid and meaningful predictions, or that it was simply necessary for the theoretical system in which it is employed. Furthermore, with the objective of mathematical formalisation in mind, it is much easier to assume agents with exogenous preference functions rather than considering more complex theoretical economic systems where individual preferences are endogenously determined223. Hodgson (1997, p. 54) adds a further more culturally inspired motive to this drive towards ‘hardheaded reasoning and ubiquitous quantification’: in his view, neo-classical economists strived for an emulation of the successes of ‘hard’ sciences such as physics. Thus, neo-classical economic theory tries to capture human behaviour in a logically consistent axiomatic system. In doing so, it is not simply presenting a ‘mirror’ of human behaviour, but instead requires a particular way of ‘seeing’ from the part of the economistobserver. Wynne (1997, p. 146) has well described the conceptual operation this entails when the theory is applied to policy relevant questions. Firstly, a particular policy problem has to be defined as a problem of meeting several (broadly shared) competing values demanding public attention. Formal methods and protocols for eliciting public values then take a fixed monetary value (such as a willingness to pay for a certain environmental asset, e.g. clean air) and assume it to represent a stable, intrinsic feature of the target group’s individual preferences for the public values. This monetary value is subsequently ‘transported’ as an (de-contextualised) ‘object’ fit for expert calculations and manipulations – but all the while maintaining a commitment of value representation to the target group. Through this operation, the ‘subjective’ (the expressed preferences) is made ‘objective’ –

222

There are exceptions however. Robbins (1984) for instance claims that the basic hypotheses of analytic economics are known by ‘immediate acquaintance’ (i.e. direct psychic evidence). These basic premises of economics are according to Robins derived from simple introspection, and therefore bear much resemblance to Descartes’ idées claires et distinctes. 223 Moreover, as one moves away from the assumptions of the ‘perfect market’ (i.e. under a series of conditions which guarantee ‘perfect’ competition), the economist’s view of ‘rationality’ becomes increasingly complex. For instance, in a situation with imperfect or asymmetric information available to participants in the market, expectations require a much higher capacity of calculation because they lead to infinite regressions – e.g. in terms of ‘expectations of the expectations of the opposite party’, or of the ‘optimisation of the optimisation cost’ (Thévenot 2002). Without hypotheses of equilibrium, perfect competition and market regulation, the very concept of ‘economic rationality’ is threatened, because the perception of others becomes an element of rationality.


129

i.e. the economist is not expressing a value judgment, but ‘merely’ measuring people’s preferences (Turner et al. 1994, p. 132) 224: …We tended not to ask where the preferences came from, or whether they were ‘good or bad’ (subject to the law, that is). This is because cost-benefit approaches try to be ‘democratic’ by using individuals’ preferences rather than some expert’s view. Otherwise the way is open for the ‘tyranny of the expert’ whereby expert values are imposed on others…

People are thus allowed to participate; they are offered means of representation, but only through the language of ‘interests’ or ‘preferences’. In fact, from the above quotation it becomes apparent that many economists see this as a ‘democratic’ aspect to their methodology. Thus, at first sight, governance would stand to gain from the theoretical strength of the neo-classical economic framework, as a result of the logical and systematic consistency of its axiomatic system. The ideal picture is that all assumptions are made explicit and that the results obtained follow logically from the application of the research protocol225. But surely, mathematical formalism is not enough to support practical recommendations in the policy sphere. Whether this theory is also practically relevant thus necessarily depends on other aspects than pure theoretical validity. One is of course the central empirical question: whether the presence of other possible causes – besides the one psychic motive of maximisation of satisfaction of one’s own competing desires – which are deemed relevant for human behaviour can be observed226. On this issue, there is growing evidence stemming from research in social sciences that economic behaviour represents only a limited aspect of the human experience. Fiske and Tetlock (1997, p. 284) for instance maintain that individuals indeed continuously make (and are forced to make) trade-offs; but – and herein lies the big difference with the economist’s assumptions of ‘rational man’ – very few of these choices are made reflectively, and fewer still publicly or explicitly227. This is because the ‘burden of conscious rationality’ is carried over for a large part to situational exigencies. This observation resonates largely with the fundamental tenets of Boltanski and Thévenot’s commonwealth model. In the words of Thévenot (2002, p. 193), 224 Boltanski and Thévenot (1991, p. 71) therefore see the central concept of ‘utility’ in neo-classical economics as a compromise between the language of the ‘market commonwealth’ (revealing essentially subjective interests of participants in a pure market situation) and the ‘industrial commonwealth’ (translating and rationalising these subjective interests into stable objective measures of preference satisfaction). Consider the following example they give: in the ‘market commonwealth’, a rose ‘acquires’ its beauty only because the participants in a market situation direct a longing gaze towards it. The rose ‘is’ (or rather, becomes) beautiful because it is an object of desire. Translated into the language of the ‘industrial commonwealth’, man has a preference for beauty (a functional model of man is built), and the rose helps in fulfilling this preference. It can do so because it ‘is’ (objectively) beautiful (it has a ‘utility’), to an extent revealed by the market price. 225 There are of course also within economics methodological disputes. We shall leave these aside for the moment. 226 To continue the analogy of representation, the question is here whether people are represented ‘faithfully’. But unlike the case of ‘normal’ political representation (where people ultimately have the power to express their dissatisfaction with the representative by simply not voting for him/her), economically represented people become entirely dependent on expert interpretation. This raises the question of the ‘soundness’ of interpretation, a matter that we have discussed under section 3.1. 227 In practice, according to Fiske and Tetlock (1997, p. 282), even money is often segregated into different types, linked to different relationships and uses, without being integrated into a common, psychologically convertible currency.

130

Chapter 2

in a complex society, each individual experiences ‘shifting involvements’, depending on his/her almost subconscious ability to attune his/her conduct to various settings: …The same person can, successively, and even within a short period of time, engage in a market transaction which requires detachment from the objects and persons he is dealing with (market coordination), then rely on domestic loyalty and stick to ingrained local customs (trust coordination), before finally planning investments on the basis of technical tools designed for accurate forecasting (industrial coordination)…

These findings in fact indicate that all central hypotheses of economics only hold under the very circumscribed ‘pure market’ situation. In short, critics thus claim that neo-classical economics may be a good ‘methodological fiction’ for dealing with market interactions, but not for the broader issues involved in public choice. Remember that in Boltanski and Thévenot’s model the commonwealths function as ‘regulative ideas’, meaning that we do not dispose of an ‘independent’ measure that would tell us to what extent this ‘pure market’ conditions are realised. There is no shortcut to argumentation and dispute. But the commonwealth model does aid in making visible certain processes of social interaction, thus rendering a qualitative judgement and a principle-based resolution of the dispute easier (but not self-evident). With the aid of the commonwealth model, we can thus easily summarise the main points of critique. Most critics see a fundamental problem in the fact that governance most of the time has to deal with questions of the ‘common good’ (dealt with in the ‘civic commonwealth’) which cannot simply be translated into competing preferences – that is, questions of the ‘common good’ are seen as having value to society as a whole over and above the value they give to individuals228. People therefore try to value the aspects of the governance question from a wider perspective, taking into account other people’s interests, their own ethical values and their view on what is good for society as a whole. By forcing these wider perspectives into the mould of preference satisfaction a bias is introduced229. Secondly, objections are raised against the assumed structure of preferences, sovereign and independent of social relations. Values, it is claimed, are in practice much more ambiguous and indeterminate than suggested by the economic model. This critique extends to the assumption of economic rationality. The argument goes that in view of the ‘common good’ character of many governance questions, forming attitudes towards them is 228

The ‘common good’ can be understood in two senses. A standard economic sense implies that a common good is collectively consumed and indivisible (e.g. a landscape, clean air, etc.). In a second sense, common goods are a matter of ethical concern, to be discussed in terms of ‘right’ or ‘wrong’ rather than ‘costs’ and ‘benefits’. The point is that what the economist (e.g. when designing a valuation questionnaire) sees as the domain of private preferences is not necessarily shared by the respondents. 229 Jacobs (1997, pp. 216-217) discusses the problems inherent to ‘contingent valuation’ (CV) methods, whereby people are asked their ‘willingness to pay’ for a certain (environmental) good, or their ‘willingness to accept’ compensation for damages to this good. Firstly, there is evidence that people do express ‘citizen’ motivations in these exercises, but this in fact raises problems for CV, since the results can then no longer be interpreted as measures of consumer surplus in the way that the theory supporting the CV exercise requires (in fact, if people do value the benefit of the good in question to other people this would imply a double counting). Secondly, there is a significant amount of ‘protest’ bids and refusals to participate in these CV exercises. Thirdly, more qualitative research (questioning the respondents afterwards) revealed that a significant amount of people refused to make trade-offs, or that (environmental) goods should not cost money in the way specified by the questionnaires (mostly private contributions to a trust fund).


131

therefore of a different kind than forming attitudes (preferences) towards private goods. It involves reasoning about other people’s interests and values (as well as one’s own – this aspect should of course not be excluded entirely) and the weight which should be given to them; about the application of and conflict between ethical principles in particular circumstances; and about the nature of society one wishes to create or sustain. In sum, this rationality implies almost by definition an engagement with other people; what is needed are public ‘debate forums’ and not simply a ‘market’. ‘Value’ is therefore not something one should accept as a given, but rather something one reasons towards. This view again seems to be corroborated by research in the social sciences. Button and Mattson (1999, p. 621), who did extensive comparative research on several practical efforts by different organisations around the US to organise such public forums on diverse issues, make the following observation: …In observing how deliberation can inform specific policy questions, we witnessed first hand one of the fundamental errors of the many models of democracy and collective action (often informed by rational choice and economic theory) that assume pre-existing and ordered preferences among citizens. It is clear to us that many of the participants in these sessions came to their opinions about the issues being discussed as a result of a process of talking and listening to others, not from their state of opinion previous to deliberation. The most common refrain in all of our post-event interviews was: “I learned a lot from being there”…

Thirdly, it is claimed that people in making value or preference comparisons, also express and constitute their respective social identities. This observation for instance explains the feelings of indignation often expressed by respondents in contingent valuation exercises when asked about the economic value of a certain environmental good. The economic model thus replaces every possible form of identification with that of ‘economic man’. But do all of these objections constitute sufficient reasons for an outright rejection of market-based approaches to governance? We argue that this is not the case. Furthermore, we argue that even a neo-classical economist might very well agree with the above criticisms without seeing a reason to abandon the neo-classical framework altogether as an aid for policy making (see e.g. Craig et al. 1993). Why is this so? The point is that while critics have been right in pointing out the many limitations to methods of valuing people’s preferences, these criticisms, though necessary, are not sufficient for rejecting the entire framework. The argument in defence of retaining the neo-classical framework is then that, even if the formalisation of people’s preferences does not fully capture the subtleties of actual behaviour – and thus gives no ‘true’ results – it at least forces values and criteria out into the open, thus making governance more accountable. Furthermore, while people in their daily lives are generally not commonly confronted with difficult trade-offs, requiring a balancing between ethical concerns and ‘mere’ preferences (situated within the different commonwealths), such trade-offs do form the essence of political struggles. Thus, proponents of economic methods might very well concede that their framework is far from perfect, but even so, they challenge critics to come up with something better. The assumption is of course that governance would benefit from any kind of ‘rational’ decision theory, instead of relying on the ‘wisdom’ of normal democratic practice and processes or

132

Chapter 2

the ‘intuition’ of political leaders. Such ‘political intuitionism’ is for instance advocated by the ecological economist Holland (1997, pp. 129-130): …We already suspected we were losing too much of the natural world, and causing too much damage, when figures started to come in about lifeless rivers and streams, increased desertification, loss of biodiversity and so forth. Indeed, it was not the measurable loss which concerned us so much as the immeasurable loss. Similarly, we suspect that the current generation of humans (some of them, at least) is appropriating an unfair amount of earth’s resources both in relation to future generations of humans and in relation to current and future generations of non-humans. These beliefs are simply ‘abroad’ in the community (…) Perhaps our suspicions are wrong? But we are unlikely to rely on economic measuring to put us right. (…) As a consequence, any discrepancy between our suspicions and the result of economic studies is as likely to feed our scepticism about the methods used in those studies as it is to make us question these suspicions…

Research in political sciences clearly shows the dangers of relying on ‘broadly shared intuitions’ to steer governance initiatives. Tetlock et al. (1996) have for instance identified a set of individual and institutional coping strategies used by politicians for dealing with difficult ‘taboo trade-offs’ (trade-offs involving values from the different commonwealths), possibly exposing them to the dangers of moral outrage (and loss of political support). These strategies include obfuscation (secrecy, keeping a low public profile, rhetorical obfuscation), decision avoidance (buck-passing, procrastination) and demagoguery. Another frequently used strategy is, according to these authors, burying the trade-offs in bureaucratic, regulatory or judicial enclaves where the light of public scrutiny rarely extends (see also Fiske and Tetlock 1997, pp. 288-292). Thus, the ‘business-as-usual’ scenario seems to be one where elected politicians resort to demagoguery or simply refuse to address difficult trade-offs, even if this would bring some benefit to society at large230. Seen from this point of view, a mere reliance on ‘intuition’ or ‘subjectivity’ (as sometimes advocated by opponents of neo-classical methods) seems unlikely to bring about any fundamental changes towards more sustainability. A further danger of such ethical and political intuitionism is of course that it becomes very difficult to hold policy makers accountable for their decisions if the criteria are not openly discussed. Our point is not meant to deny the fact that intuition or emotions do play a role in political debate – or even that they should not play any role – but that reliance on emotions or intuition alone is unlikely to be sufficient to guide governance for sustainability. The fact remains that policy measures for sustainability will have to turn, as any other policy matter, on the central role of a deliberated trade-off. An agreement on the ‘correct’ policy, which of necessity will have to be capable of being generalised across society as a whole, will quite simply have to address the fact that there are limits (e.g. to government spending, to levels of taxation, etc.) and that the possibilities for genuine ‘win-win’ situations are usually limited. Furthermore, in a complex and highly differentiated society, such conclusions will only be defensible when they are applied consistently (i.e. there has to be some kind of a collective agreement that ‘these costs are outweighed by these benefits’ in any similar situation) and coherently (i.e. there has to be some kind of logic to the collective agreement

230

This is partly because the organised interest groups who would stand to lose from a particular policy initiative usually have a louder ‘voice’ in the policy process than the diffuse ‘broader community’ (including perhaps the environment and future generations) who would stand to gain.


133

so that anyone who follows the line of argumentation arrives at the same conclusions concerning the ‘costs and the benefits’) 231. Market-based approaches to sustainability, for all their manifest deficiencies, have nevertheless at least been trying to establish exactly this kind of basis for policy measures. As Foster (1997, p. 235) puts it …Of course values are plural, in the sense that lots of things matter, and they matter in lots of different ways even to the same people, never mind as between different people. Of course there is no single implicit ordering out there which some sudden improvement in our ethical instrumentation would permit us to at last register. And equally clearly it is not the case that in order to make practical judgements among these values for ourselves, we must buy into some single master-value which we can use as a measuring rod (…) But while the idea of value pluralism is all very well when it serves to remind us of these important meta-ethical points, our eagerness to emphasise the pluralism must not lead us to overlook the fact that what we are taken to be plural are still values (…) Indeed, if a decision has to be made, as in this fallen world it so often does, between two mutually incompatible real options, then the reasons for one must be reasons against the other. Values, as it were, can be left to disport themselves in happy plurality when off-duty, but the need for a decision calls at least the relevant ones sharply into line…

Now, in fairness to the critics of neo-classical approaches, we should admit that they generally do propose an alternative, which we can broadly characterise as a ‘deliberative’ approach. We will deal with this approach in more detail in section 3.4. For now, it suffices to say that such approaches advocate the direct involvement of people or communities who are affected by proposed policy measures. In the case of sustainable development, this comprises that the elaboration of the concept is itself subject to a continuous process which is promoted by stakeholder participation. An obvious advantage over neo-classical methods is that such public forums would enable not only reasoning about one’s preferences, but also about other people’s preferences and values and the weight which should be given to them; about the application of and conflict between ethical principles in particular circumstances; and about the nature of the society one wishes to create or sustain. As to representativeness – the issue we are dealing with here – deliberative institutions typically involve much smaller numbers of people than economic surveys. But the point of deliberative institutions is to dig deeper, to explore opinions, and not just to reveal preferences. It is therefore argued by proponents of this approach that even though smaller numbers are involved, people are actually better represented. In any case, there seems to be a trade-off here between the qualitative and quantitative dimensions of ‘representativeness’ (Jacobs 1997, p. 223). In section 3.4, we will argue that neoclassical economic insights about sustainability can be integrated in a deliberative approach (the reason is quite simply that ideal conditions of communication are inter alia realised through the provision of information and expert knowledge). For now, we will explore the claim that governance for sustainability can rely solely on the kinds of institutional improvements suggested by economic analysis, and hence has no need for more ‘deliberative’ practices and institutions. In case of an affirmative answer (quod non), we could stop our search for the adequate governance response right here.

231

We leave open the possibility for a broad definition of ‘costs’ and ‘benefits’.

134

Chapter 2

3.2.2.2 The institutional critique At first sight, market-based approaches to governance offer better opportunities than the expert-based governance scheme for providing the stable foundations needed for the common construction effort called sustainable development. This is because, as Latour (2004a, p. 132) has noted, market-based governance introduces ‘nature’ right inside the political sphere, whereas expert-based governance had to rely on an external nature to bear upon the political deliberations. Economics (as explained above) claims to be able to treat both ‘facts’ and ‘values’ on an ‘objective’ basis. Latour (2004a, p. 134) – with his usual poignancy – denounces this ‘exploitation of the ambiguity between facts and values’ in economics as follows232: …If you say that this discipline is scientific and must therefore describe in detail the complicated attachments of things and people, it will reply that it does not have time to be descriptive, because it has to move very quickly to the normative judgment that is indispensable to its vocation. If you acquiesce, albeit in some astonishment at this casual tone, you will be surprised to see that, in order to produce the optimum, economics does not burden itself with any consultation, and its work of negotiation is limited to the calculus alone. (…) If you become indignant at this cavalier attitude, economics will signal you to be quiet: “Shh ! I’m calculating…” and will claim not to need either to consult or to negotiate, because it is a Science and because, if it defines what must be, it does so in the name of its laws cast in bronze, as indisputable as those of nature. If you point out politely that it is difficult to be counted as a science before devoting a great deal of time to the requirements of description, before plunging into controversies, before deploying instruments that are as fragile as they are costly, it will reply that it prescribes what must be done; and if you object once again, losing all patience, that economics does not respect values because it has jumped over all the requirements of prescription, it will retort scornfully that it only describes facts, without concerning itself with values ! By allowing the discipline of economics to unfold, one thus keeps the collective, by the cleverest of schemes, from having to produce any description in the name of prescription, and from having to hold any public debate in the name of simple description…

232

This ambiguity concerning the role of economics in governance is intimately linked to the (self-perceived) role of the economic profession. Deblonde (2001, pp. 4-7), based on Nelson (1987), discerns three different interpretations of the economists’ role in governance which prevailed in American history. The first one is that of the ‘Progressive Neutral Expert’. The political theory underlying this first role distinguished between two government functions, namely those that involve basic policy questions and those that are administrative or instrumental in nature – decisions of the former type belonging to the realm of ‘politics’, decisions of the latter to the separate realm of ‘administration’. ‘Administration’ then is supposed to be the domain reserved for experts in economics, who are expected to provide governments with efficient means (and not with political ends). The ‘Progressive Neutral Expert’ thus stresses extensive gathering of data and measurement of social phenomena. The second role, the ‘Entrepreneur for Efficiency’, is a prolongation of that of the ‘Progressive Neutral Expert’, but adapted to a different political theory. Government, in this view, is no longer presumed to be driven by the ‘public interest’, but rather by continual competition and bargaining among privileged interest groups. Political leaders and interest groups are thus believed to be no longer respectful to the progressive boundaries between the properly political and the properly expert. This political theory thus gave rise to an interpretation of the economists’ role as active advocates of efficiency. Only as active advocates of efficiency could the economist continue to view himself as a spokesman for the diffuse and weakly represented interests of the ‘general public’, thus acting as a counterweight to organised interests. This new role as ‘Entrepreneur for Efficiency’ thus demanded command of both scientific/technical knowledge and political and tactical skills. The third and last role identified by Deblonde is that of the ‘Ideological Combatant’. This third role is supported by a political theory that sees politics not only as a terrain for interest group struggles, but also of ideology. Concomitantly with this third political theory, the role of the economist is that of a proponent of his/her particular framework of thinking. This role generally entails a propagation of markets and efficiencyrelated values, rather than simply an analysis of the economic aspects of particular policy details.


135

Thus, according to one possible interpretation, economics is able to perform all the functions required of collective decision making (and thus, it provides a possible shortcut to the construction of the common world): it is able to speak at once for the ‘human world’ (called consumers and producers) and the ‘non-human world’ (called commodities); it upholds values of personal autonomy and freedom (in the market); and through its calculations, it is able to produce a hierarchy of values (the optimum allocation of resources). As a result of these calculations, neo-classical economists might concede the bias of actual market situations, but they would attribute the difference between ideal models and ‘vulgar’ realities to ‘market imperfections’ (e.g. the externalities discussed in the next chapter). They are an indication that markets are then not maximising collective welfare, and hence, the need for governance intervention is advocated. This then means regulating markets by adding institutions, either directly (via command-and-control measures) or by introducing market instruments (property rights, taxes, subsidies, tradable permits, etc.) – with the ultimate end of making government intervention superfluous, once the markets are again functioning in an optimal way. Critics have taken issue with this automatic translation from economics to economic institutions, claiming that it is necessary to distinguish between economics as a discipline and ‘the economy’ as an institution. Indeed, accepting constructivism as a (meta-) theoretical language (as we have done) can only lead to an assertion that there is no such thing as ‘an economy’ (or for that matter, a homo oeconomicus), looked down upon from above by ‘the economist’, but there is indeed progressive economisation of relations (Latour 2004a, p. 135). The ‘pure market commonwealth’ has to be transformed into a ‘market world’ (or ‘an economy’) by ‘economisers’ (accounting systems, modelling scenarios, mathematicians, marketing specialists, statisticians, etc.) through the stabilisation of objects (e.g. the identification of tradable commodities, property rights, market rules, etc.). In other words, markets in their ‘pure’ form are simply an abstraction from one or another concrete realisation in which they inevitably take on social and political biases – an abstraction, critics maintain, that therefore tends to obscure other possibilities for institutional design. This institutional understanding of ‘real-life’ markets provides an adequate entry point for understanding what we have dubbed ‘the institutional critique’. In the literature on institutional design, Ostrom’s (1992) distinction between three kinds of rules is useful. Adopting the metaphor of a game, Ostrom proposes to analyse institutional design on three mutually interconnected levels: 1. the constitutional rules which specify the nature or charter of the institution, or define what the game is about; 2. the collective-choice rules, or rules of the game that characterises the institution; and

This role as an ‘ideological combatant’ in turn calls for particular types of skills in order to penetrate and criticise the philosophical underpinnings of social and political values and theories. Latour (2004a) (and we) clearly take issue with the economist’s role conceived of as an ‘ideological combatant’.

136

Chapter 2

3. the operational or tactical rules, how to act productively within the institution (or game). Market-based governance approaches for instance function within a set of ‘constitutional rules’ (e.g. Rawls’s ‘overlapping consensus’ on fundamental principles of justice) which they do not put into question; as a ‘collective-choice rule’ they (obviously) propose markets; and on the tactical level, it is supposed that people (who play the game) act as rational preference-maximisers. The ‘institutional critique’ then holds that an exclusive reliance on market-based approaches (or aggregative approaches in general) as a ‘dominant cooperative scheme’ will lead to: 1. On the ‘constitutional’ level: inherent conservatism in the sense that the ‘common good’ (or background political culture, overlapping consensus, etc.) or the distinction between the public and the private sphere are a priori bracketed out of the discussion based on a supposed already existing rational consensus. For new policy principles such as sustainable development, this cannot simply be assumed (cf. infra, the discussion on intergenerational ethics) 233; 2. On the ‘collective choice’ level: a strengthening of decisionism as a model for collective choice. Decisionism sees institutions as mere ‘means-to-ends’ (e.g. correcting market imperfections) rather than culture-embedded rules expressing a (desirable) collective identity234. Thus it is blind to social identity as the basis of values, and thus a nonrecognition of the shaping of social identities by institutional structures. Decisionism also tends to isolate from view the methods used to elicit people’s preferences. Furthermore, it is a-historical, because decision issues are insulated from their wider historical background235. This makes market-based approaches profoundly conservative in the sense that it has no tendency to question the desirability of the circumstances in

233

Varner (1994) gives a very interesting example regarding property rights. In Varner’s words, the liberal democratic state – through its legislative, administrative and judicial agencies – has gradually chipped away at the unilateral freedom of private property holders to use or dispose of their property as they see fit. For example, the emergence of rules of strict liability in relation to serious risks absolves plaintiffs from the obligation to demonstrate negligence. Environmental impact assessment (EIA) procedures are another example, insofar as they require proponents of new projects to show that these cause negligible or acceptable damage to the environment. This leads Varner (1994, p. 143) to speculate that the day may come when we “…treat land as a public resource owned in common and held by individuals only in a stewardship (or trust) capacity…”. 234 Wynne (1997, p. 147) gives the example of a coal-mine rescue dilemma. This dilemma involves a decision about a rescue-bid for the trapped survivors in a coal-mine accident, but only by risking rescuers too. A costbenefit analysis (in this case, more correctly a multi-attribute utility analysis) was used to guide policy-makers towards a rational decision, navigating between the different possible outcomes, and weighing possible benefits and costs (including even feelings of guilt, regret, etc.). But as Wynne points out, in order for mining to exist at all as an institution, miners have to be able to trust that if they are ever trapped alive, a rescue will be attempted, virtually regardless of the consequences. 235 A very good example would be a decision concerning the management of radioactive waste. Given that radioactive waste exists, it has to be managed; thus, a waste management facility has to be built somewhere. Arguably, this is a difficult question to tackle, and their will be no single solution without (ethical) drawbacks. An awareness that all possible options have potential drawbacks should then lead to a continuous review of the circumstances that have lead to this difficult position in the first place – i.e. the question whether radioactive waste should be continued to be generated cannot be isolated from the decision context.


137

which the analysis is made, or to suggest that another set of options would be preferable236; 3. On the ‘operational’ level: a strengthening of human behavioural patterns along the lines dictated by homo oeconomicus described above. This is of course not a matter of determinism, but this critique echoes the philosophical constructivist stance that discourses do have material-ordering implications – hence leading people to unreflexively confirm to individual and instrumental commitments (Wynne 1997). Normative debate may of course still take place at the private level, and citizens can take a moral engagement towards (their interpretation of) sustainability as a matter of fulfilling their ‘design for life’. But this moral engagement is not encouraged and the link with the political sphere is severed, because this is dictated by the ‘neutrality’ of public policy. Hence, political learning from (and creative interplay with) spontaneously generated forms of life is made virtually impossible. Furthermore, market-based governance confounds between human ‘needs’ (a notion from the ‘industrial commonwealth’) and human ‘wants’ (a notion from the ‘market commonwealth’). Now an economist could still argue that this might all very well be true, but given the realities of the capitalist market, and given the urgency of some ecological problems, it is very difficult to arrive anywhere near a more desirable solution if you start the discussion on ‘values’. Our best option would still be to revert to market solutions in order to rectify problems created by markets – especially in view of the impulse and opportunity for experimentation and innovation provided by a market economy and a free society. Again (as was the case in our discussion of the expert-based governance scheme), we maintain that there can be no definitive answer in the choice between formalised market-based responses on the one hand and a reliance on the relational negotiation of value (discussed under section 3.4) on the other. And again, as a matter of principle, justification or refutation of a further extension of markets remains a societal choice, which should be considered in due process. A further problem for market-based governance approaches (and aggregative approaches in general), which puts a continuous strain on the ‘rational overlapping consensus’ constitutive of the liberal democratic state, arises in the issue of fairness vis-à-vis future generations. Since future generations cannot express their preferences, all that can be done is for the current generation to agree on broadly shared ethical principles which are fair to future generations. Thus, the question of intergenerational automatically implies the search for a ‘moral’ consensus. Frequently, the Rawlsian approach is advocated. Rawls takes egalitarianism as a baseline against which deviations should be justified. In the case of

236

Schrader-Frechette (1985) has argued that, in a rapidly changing situation, where leaders must act on the basis of likely future events, public welfare cannot merely be the aggregate of private preferences. This is, according to Schrader-Frechette, because present decisions about costs and benefits in part determine future values and preferences. Schrader-Frechette furthermore considers it to be a duty for ‘good leaders’ to attempt to reform public opinion (at least in some instances) – therefore, authentic public welfare and the aggregate of citizens’ preferences are not necessarily identical.

138

Chapter 2

justice between generations, he proposes a ‘principle of just savings’237. Discounted utilitarianism (where time-preferences dictate that present consumption is given more importance over delayed consumption by future generations) of course violates intergenerational equity. But strict utilitarianism (with no preference for the utility of present or future generations), also leads to problems in the equitable treatment of the generations (Deblonde 2001). This is because of the utilitarian defence of the ‘greatest (average or total) wellbeing (however measured) for the greatest number of people’: since we do not know how many generations will follow after us (and, for the sake of simplicity, assuming constant productivity of capital and a constant population), a utilitarian can be expected to adopt a positive savings rate indefinitely (i.e. a never-ending sacrifice on behalf of the future). Accepting a ‘bliss point’ (that is some point beyond someone’s utility cannot be improved anymore) leads to an accumulation phase followed, once the level of saturation is reached, by a steady-state stage with a zero-rate intergenerational saving. A shift to a presumption of a finite number of future generations would still entail a positive savings rate by the earlier generations, which would decline to eventually become negative for the last generation – the latter would consume the whole capital. In any case, a bigger effort would be required from earlier generations, thus raising difficulties for egalitarians. The widely-recognised problem of intergenerational discounting has driven even neoclassical economists to a search for other principles of intergenerational justice. Pearce and Turner (1990, p. 24) for instance define sustainable economic development as increasing the level of welfare while keeping the (natural) capital stock at least constant. This is consistent with Rawls’s ‘principle of just savings’, at least in the steady-state (in a ‘wellordered society’ – cf. footnote 237). In fact, the Rawlsian approach is often construed as consistent with a social welfare approach that constructs the environment as consisting of natural assets which provide a stream of benefits and services for society so long care is taken not to damage them. These assets include the stock of natural resources which 237

Rawls has indeed devoted a full section in “A Theory of Justice” (Rawls 1971, pp. 284-293) to the matter, addressing both procedural and substantive issues. Briefly put, on the procedural side Rawls’s definition and (to some extent) justification of principles of justice relies on a hypothetical construction called the ‘original position under the veil of ignorance’ (this means that we need to define principles while ignoring our specific position in society, or our gender, race, nationality, religious beliefs, etc.). Where intergenerational justice is involved, Rawls asks those put in the ‘original position’ to decide how much they would be willing to save at each stage of advance on the assumption that all other generations are to save at the same rate. For the question of intergenerational justice, Rawls drops the assumption that people are mutually disinterested, and instead assumes that each is a ‘head of family’ with a desire to further the welfare of their descendants. On the substantive side, the ‘principle of just savings’ entails a goal – namely achieving and preserving just institutions and the fair value of liberty – and constraints on the means of achieving this goal – each generation has to carry a fair share of the burden of realisation of the goal. This is in line with ‘justice as fairness’ in the intra-generational case, but, in order to arrive at this conclusion, Rawls has to drop his ‘maximin’ egalitarianism (the people placed under the ‘veil of ignorance’ will chose the set of principles of justice maximising the benefits of the worst-off under each alternative set of principles), since there is no way for the ‘better-off’ future generations to redistribute wealth to the ‘worse-off’ present generations required to save capital. But Rawls also refers to a ‘last stage’ – a stage where ‘just institutions have been firmly established’ (Rawls 1971, p. 289) – and thus, the net accumulation required falls to zero: at this point a society meets its duties of justice by maintaining just institutions and preserving the capital base. Gosseries (2001, pp. 311-317) refers to the Rawlsian point of view as a ‘two-stage approach’, namely an accumulation stage (where net savings are required from each generation) and a steady-state stage (where the savings rate can fall to zero).


139

benefits the economy (e.g. minerals, soils, air, water, wildlife, etc.) and a range of tangible and intangible ‘services’ which are often taken for granted. Obvious services include nutrient recycling, waste decomposition, and climate support; less tangible services include the range of meanings which landscapes, nature and places are invested with by local communities and which society as a whole values. However, it is not clear then how this ‘constant (natural) capital stock’ should be measured and/or implemented238. Interestingly enough, Gosseries (2001), starting from the central ethical tenets of Rawls’s ‘principle of just savings’ in order to develop an intergenerational equity rule, arrives at the same difficulties. Gosseries argues that, as a matter of justice (adopting a maximin egalitarian attitude), well-ordered societies should keep their total capital bases (however measured) constant over time – that is, they should neither save nor dis-save (this comes down to a steady-state economy, thus a departure from Rawls’s view which only prohibits dissavings). This is because societies should not be worried only about intergenerational equity, but also about intragenerational equity. Hence, if there are any ‘surpluses’, they should, according to Gosseries (2001, p. 325), be given to the worst-off in the present generation, instead of being transferred to the next generation(s) (assuming that every subsequent generations respects the same intragenerational equity rules). However, there are exceptions to this general rule (pp. 327-334): an unanimous decision by the present generation on the desirability of positive savings, the uncertainties involved might lead to the adaptation of a more prudent attitude of some positive savings, the risk of future generations falling victim to some exogenous disadvantage, demographic fluctuations, and strict egalitarian considerations could sometimes override the maximin egalitarian position. Moreover, there are also practical issues, related to the implementation of the principle. These practical issues (as was the case with the neo-classical economist Pearce) relate to the definition and measurement of the ‘capital’ that we are supposed to leave for future generations; here we touch again upon the debate between ‘weak’ and ‘strong’ sustainability as set out in section 3.5 of chapter 1. The point is, again, that abstract concepts (such as the ‘function’ of capital) only acquire meaning in concrete social and cultural settings. Abstract ‘rationality’ can guide our reflections, but it can never settle the debate. In this, we agree with Gosseries (2001, pp. 353-354):

238

In another publication, Pearce et al. (1989, pp. 43-53) point out three different interpretations. Firstly, a natural capital stock can of course be considered constant if its physical quantity does not change. But since there is no way of adding different physical quantities, the standard economic approach seeks to value each type of natural capital in money terms (total economic value of natural capital = use value + option value + existence value). The constant capital stock requirement then means that the real total aggregate value of the stock should be kept constant. Secondly, a constant natural capital stock can signify that the market prices of natural resources are held constant in real terms. On conditions that prices reflect absolute scarcity, constant real prices will imply a constant natural capital stock in this sense. Thirdly, maintaining the natural capital stock can be interpreted as a natural capital stock that delivers a constant value of the resource flows. This interpretation allows quantities of natural resources to decline on condition that the prices of these resources rise in such way that the total value is kept constant. The authors admit that there are no absolute rules, and that the choice will depend on a number of considerations, e.g. on the presumed substitutability between natural and man-made capital, on the power of technological progress to reduce the resource input to a unit gain in wellbeing, on population growth, on the ‘economic functions’ the natural capital fulfils (e.g. life-support, climate regulation), on the scientific uncertainties involved etc. (Pearce et al. 1989, pp. 37-43). Even more economic measures of sustainability are explored in Mofatt et al. (2001, pp. 45-74).

140

Chapter 2

…To what extent does this all help in practice though? It forces us to look at the basket of what we transfer to the next generation through the prism of the concept of ‘equivalence’, having in mind the relevance of a choice of ‘metrics’. This entails, for example, that we should not merely focus on ‘environmental capital’ (to conclude hastily that the next generation will be much worse off than us) but that the claims of limited substitutability underlying the notion of ‘strong sustainability’ also need to rest on firmer ground (if any). (…) Drawing precise political conclusions is admittedly not an easier task here than in other areas (…) Uncertainties are huge in practice and we have stressed that prudence should not be excessive…

3.2.3

Conclusion

As was the case for the expert-based governance scheme, governance by aggregation operates according to its own logic of inclusion/exclusion and rules for structuring policy problems. Indeed, once the identity of the participants in the discussion has been specified (the free, equal, autonomous and tolerant individual), both the issues that can be publicly addressed (in contrast to those pertaining to the sphere of private self-determination) and the way in which one’s position can be supported (either by means of ‘rational’ valueneutral arguments or the appeal to discursively insurmountable divisions which accordingly have to be phrased in the language of ‘interests’) follow as logical consequences. In doing so, it tends to be conservative, in the sense that it takes as given (the rational boundaries of the ‘common good’, ensuring the cohesion of society) something which is subject to continuous ethical debates, especially in the case of ‘new’ issues such as finding adequate principles of environmental and intergenerational justice. A strict adherence to ‘neutrality’ implies that certain policy issues cannot enter the political arena or can only be phrased in a particular way (i.e. by reformulating issues in a language acceptable to all reasonable comprehensive worldviews)239. Also, an appeal to abstract ‘rationality’ tends to obscure the fact that (policy) decisions have to be made sometimes (and this is more often than not the case in complex policy issues such as the questions posed by sustainability) in absence of a complete ‘rational’ justification; hence, at a given moment, a sufficiently powerful actor has to cut the knot. By completely removing power from the equation liberal theorists also tend to lose sight of those who stand to gain from a particular way of framing the decision problem. In any case, a reliance on governance by aggregation in itself remains insufficient since it cannot explain how new issues concerning the ‘common good’ appear on the political agenda nor how social and cultural conditions in the public sphere can be ‘shaped’ to provide a more stimulating environment for discussing common values such as sustainable development. In fact, the critique comes down to the statement that liberal pluralism should be radicalised. This does not mean that we have to become ‘illiberal’ in the sense that some

239

Pellizzoni (2003, p. 343) argues that the Rawlsian concept of ‘reasonable pluralism’ leads either to a denial of the plurality of reason or to a deliberate self-containment of discussion. The argument is that either Rawls’s overlapping consensus is either funded on a ‘common human reason’ or the sharing of ‘fundamental intuitive ideas’ – but then there is no actual value incommensurability, or, if incommensurability does exist, then the overlapping consensus is indistinguishable from compromise based either on coincidence or (and Pellizzoni considers this more likely) on deliberate choice. In other words, agreement on the solution to a problem, and hence the nature of the issue at stake (e.g. its public or private status), can be obtained either because there is no fundamental diversity of worldviews, or because discussion is conveniently confined within a ‘safe’ limit.


141

enduring features of political liberalism – e.g. the protection of civil and political rights, the election of parliamentary representatives, the separation of powers, the idea of toleration and respect for moral pluralism and the exercise of state power according to the rule of law – should be abandoned240. On the contrary, there is substantial evidence that the environmental gains that have been made to date in developed countries are largely the result of democratic interaction and the openness of political and administrative decisions to public and media scrutiny (Jänicke 1992). Rather, we take issue with the limited scope and quality of political representation allowed by liberal democracy in general, and governance by aggregation in particular. The remedy we propose will have to incorporate a form of social learning (Section 3.4). This does not imply that both approaches are ultimately irreconcilable. On the contrary, by distinguishing between economics as a method and adding economic institutions, the way is paved for allowing economic insights into a deliberative process. The key contrast, however, is that such an approach sees the formulation of policy for sustainability as driven by ongoing scientific analysis and public debate, rather than simply by attempts to quantify the wishes of individuals. Both approaches are messy, complicated and potentially controversial. But the first one is not necessary and a priori limited to measuring individual self-interest and acting upon the results of this measurement.

3.3

3.3.1

Governance by pacification


The third governance model we will discuss at present – ‘governance by pacification’ – is in a certain sense the mirror image of ‘governance by aggregation’. Whereas governance by aggregation in effect presupposes consent on the nature of the problem at hand (i.e. belonging either to the ‘public’ or ‘private’ sphere, involving questions of either ‘the common good’ or ‘private interests’) in order to function properly (i.e. this consent is needed because conflicts would otherwise be potentially destructive), governance by pacification does not need such consent. In this case, it is openly acknowledged that the policy problem under scrutiny involves competing and often conflicting interpretations of the ‘common good’ and therefore cannot be solved decisively; hence, attention is directed towards mechanisms to guide the ongoing debate on the ‘common good’ (the conflicting interpretations are ‘pacified’)241. Such debate can only result in a ‘compromise’ in Boltanski and Thévenot’s terms – i.e. an agreement on ‘the common good’, without however letting prevail a specific grandeur of one of the commonwealths of the different

240

In a similar vein, Latour (2004a, p. 71) sees his project of ‘the politics of nature’ as a continuation of the liberal tradition: “…I am indeed situating myself in the concatenation of these principles, in the long and venerable tradition that has constantly expanded what was called humanity, freedom, and the right of citizenship…”. 241 Of course, it cannot be excluded that a particular interpretation of the ‘common good’ also serves the ‘interests’ of a party present in the negotiations. Nevertheless, the specific settings of the ‘governance by pacification’ scheme requires of the actors to frame their arguments in notions of the ‘common good’.

142

Chapter 2

actors in conflict. This implies that the ‘common good’ is not further specified in terms of the principles underlying the different commonwealths. Each actor is then able to accept the compromise based on a reasonable argument; however, the reasons are not the same for each actor involved. The difference between ‘governance by aggregation’ and ‘governance by pacification’ is strongly correlated to the traditional difference made in political sciences between pluralistic and (neo-)corporatist political systems, when applied to the level of policy making. Hisschemöller (1993, p. 167) quotes Schmitter (1979), who defines pluralism as …a system of interest representation in which the constituent units are organised into an unspecified number of multiple, voluntary, competitive, non-hierarchically ordered and self-determined (as to the scope of interest) categories which are not specially licensed, recognised, subsidised, created or otherwise controlled in leadership selection or interest articulation by the state and which do not exercise a monopoly of representational activity within their respective categories…

A good definition of corporatism on the other hand is for instance given by Dryzek (2000, p. 91): …a form for the organisation of national political systems (…) as a tripartite concertation of government, labour and business, the last two represented by encompassing associations. All three partners are involved in the making and implementation of policy; institutions such as parliament play a comparatively minor role, as it is the executive that represents government…

Thus, the main difference between both models is that in the case of ‘governance by pacification’, participation is limited to a limited number of compulsory, non-competitive, hierarchically ordered and functionally differentiated categories, recognised and often licensed by the state and granted a deliberate representational monopoly within their respective categories; whereas in the case of ‘governance by aggregation’, participation is usually larger and less formalised (most of the time involving the groups who stand to gain or loose most from the particular policy measure under discussion), both with respect to the interest representation and articulation. Bruyninckx (2002) sees the Belgian policy-making system as an almost perfect example of the corporatist model of state-society relations. This is because in the post-World War II period, a limited number of groups in society (most notably employers and labour unions) were privileged by the government as preferential partners to give input on a number of important societal problems. Bruyninckx considers that, in exchange for their input (and often control) over the content of policies on these issues, these groups actively defended and sometimes even implemented the state’s policies. The Belgian corporatist system is also widely seen as the result of the three ‘fault lines’ that run through the Belgian polity, namely language, religion and economic differences. According to Bruyninckx, this has led to a ‘balkanisation’ of the state into little domains which were divided almost on the basis of a quota system by representatives of civil society coming from the different segments of the three divisional lines. Despite these serious divisions, political conflicts have therefore nevertheless been fought out in a ‘remarkably civilised’ way. Among the many policy problems thus ‘pacified’, energy policy has been a prominent example.


143

Therefore, before turning our attention towards an investigation of the possibilities and drawbacks offered by the model of governance by pacification in the context of planning for sustainable development, we will explore the historical case of investment planning in the Belgian electricity sector. 3.3.2

An example: investment planning in the Belgian electricity sector

Unlike some other European countries, the Belgian government did not nationalise the electricity sector after the end of World War II242. Some parties (i.e. the communist and socialist party) were in favour of such move, but in July 1955 a compromise was reached with the socialist party. The electricity sector remained for the largest part in private hands, but employers’ organisations and labour unions were allowed a part in the supervision of the sector. This agreement led to the foundation of the ‘control committee for electricity and gas’ (CCEG) which became responsible for its implementation. The three labour unions (the Christian ACV, socialist ABVV and liberal ACLVB) as well as the employer’s organisation (VBO) functioned as supervisors in the CCEG, while the ‘management committee of the Belgian electricity producers’ (BCEO) functioned as the controlled party. Originally, government was not represented in the CCEG, but this changed in 1964 with the renewal of the original agreement. The CCEG had to verify whether the technical, economic and tariff structure of the electricity sector (and after 1964 also the gas sector) agreed with the public interest as set out by the dominant lines of energy policy: energy supply was diversified in order to minimise the geopolitical risks of dependence on oil-exporting countries and measures were provided to guarantee a secure and continuous supply of energy at a reasonably competitive price for all concerned. The CCEG ensured that efficiency gains were distributed among capital, labour and customers; costs and tariffs were controlled, and the profits were regulated on a rate-of-return basis to the invested capital (this of course greatly reduced the risk of investment for the electricity sector). The agreement led to a rapid rationalisation and concentration within the electricity sector according to the law of economies of scale: whereas in 1955 36 particular electricity companies were operational, this number decreased to 3 in 1976 (EBES, Intercom and Unerg) (Leroy 1979, p. 137). In 1990, the three remaining private electricity companies merged in Electrabel. Public energy ‘producers’ were since 1978 reunited in SPE. In the new context of the liberalised European electricity market, Electrabel remains the dominant market player in Belgium. In the electricity distribution sector, privately-owned companies equally strengthened their hold on the market. During the ‘60s, a lot of communal companies ceased to exist and instead ‘intercommunal’ companies were established. These are joint ventures between at least two communal authorities, with (the so-called ‘mixed’ intercommunal companies) or without (the so-called ‘pure’ intercommunal companies) private partners. Quickly, the mixed intercommunal companies became the dominant market players: in the mid-’70s, already 75% of the distribution 242

This paragraph is mainly based on Verbruggen (1986) and Verbruggen and Verstappen (2003).

144

Chapter 2

companies were ‘mixed’ intercommunal companies. This enabled the electricity utilities (EBES, Intercom, Unerg, and later Electrabel) to gain a large political influence also on the communal level, because the local authorities became dependent on the profits of electricity distribution. Starting from 1964, both ‘pure’ and ‘mixed’ intercommunal companies were represented in the CCEG. The CCEG was also responsible for the control on the investment plans submitted by the BCEO. The exact extent of this control is difficult to judge, because the documents of the CCEG were not made public. However, there is a strong suspicion that control was rather limited, because the CCEG did not dispose of its own means of expertise (and independent experts were not admitted), and because the committee only looked into the efficiency of the electricity sector (and thus did not take into account the wider perspective of energy policy) (Verbruggen 1986, p. 309). The procedure in vigour was criticised mainly by the ABVV, who boycotted the CCEG in the period 1977-1983. This in turn led to the foundation of the ‘national energy committee’ (NCE), a new consultative body with an advisory function towards government243. Formally, the NCE was installed on 29 April 1976, but it became only fully operational after the state reform (law of 8 August 1980, Art. 73, §1), which laid down the guidelines for the advice procedure on the investment plans submitted by the BCEO244. During the ‘80s, the NCE mainly provided a forum for the debate on the investment plans; and in particular the debate concerning the proposals to build another nuclear power plant in Belgium (‘N8’) combined with the proposed Belgian participation in the French reactors Chooz B1+B2 received a prominent place on the agenda. Formally, the law of 8 August 1980 provided for an inquiry procedure, whereby the investment plans were submitted to the NCE during a one-month period, after which the NCE could formulate an advice or a policy recommendation directed towards government. The competent minister (the Minister of Economic Affairs) then disposed of one month to formulate a decision regarding the desired investments. This arrangement effectively brought about a broadening of the debate on investment planning towards energy policy as a whole and through the involvement of a larger number of stakeholders245. However, another important element in the broadening of the debate was the fact that during the ‘80s, independent (university) research groups (funded by the federal ‘services for the programming of science policy’ (DPWB) under the national R&D ‘Energy Programme’) were active in the fields of energy production and consumption. Independent experts were thus able to provide a counterweight to the experts engaged by

243

Both the CCEG and the NCE were later replaced by the federal and regional control organisms (respectively CREG and VREG in the Flanders region, BIM in the Brussels region, and CWaPE in the Walloon region) in the liberalised electricity market. 244 The NCE was founded by Royal Decree of 12 December 1975. 245 Besides the electricity sector (the centralised producers organised in the BCEO and the decentralised selfproducers), large industrial electricity consumers (VBO-FEBELIEC), other (smaller) consumers (VBOorganisations of tradespeople) and the trade unions were represented. These groups could issue an ‘advice’. In addition, other members of the NCE were entitled to issue an ‘opinion’: representatives of consumer organisations, the federal planning bureau, the national advisory council for the coal industry, and the public electricity utilities. During the hearing sessions of the NCE, independent experts could be invited to respond to the questions raised in the inquiry process.


145

the electricity sector246. Nevertheless, there were also some practical shortcomings which stood in the way of reaping the full benefits of this broadening of the debate: the NCE did not dispose of its own means of expertise, and the law of 8 August 1980 only provided for two months to form an informed opinion regarding the investment plans (starting from the freely chosen moment when the BCEO decided to submit its planning). According to Verbruggen (1986, p. 312, our translation), this time period is …much too short to pronounce a balanced judgment, to formulate alternatives and to come to a cooperation between the parties. This simulated shortage of time leads to a frenetic activity, leading to a decision wanted by the strongest party and not necessarily the best decision…

The historical example of the debates taking place in the NCE is interesting for our purposes since we can learn some important lessons about participation of social groups in decision making and the knowledge base on which the decisions are based. In fact, uncertainty arguments played a major role in these debates (Laes et al. 2004c, pp. 112115). The proponents of new nuclear investment in Belgium (mostly the BCEO, supported by the VBO and the organisations of tradespeople) argued in the successive investment plans (1981-1988) in favour of the construction of one or more new nuclear power plant(s)247. Such arguments were based on the expected growth of electricity consumption (in turn based on general macro-economic forecasts and the expected evolution of economic parameters such as the dollar exchange rate and fossil fuel and uranium prices). Based on these forecasts, the necessary investments in electricity production capacity were calculated248, taking into account a certain (large) ‘safety margin’ (ensuring that the likelihood of production not meeting demand would remain below a certain threshold)249. Investments in new nuclear capacity were mostly promoted with a certain urgency, arguing that a ‘lack of political courage’ would lead to grave consequences regarding employment levels, costs of electricity, etc. 246

In 1987, the then Minister of the State Budget decided to terminate the R&D ‘Energy Programme’ without guaranteeing the continuity of expertise. Austerity measures, combined with the decreasing political importance of energy issues caused by a decrease in oil prices starting from 1985 and the logic of regionalisation (causing competences in the energy field to be spread over the federal and regional levels), certainly played a role in this decision. However, some privileged witnesses also saw this decision as a political move aimed against some independent academics, caught in a growing rivalry with the expert groups engaged by the electricity sector (Laes et al. 2004c, p. 77). In any case, the loss of independent expertise impaired the quality of the societal debate on energy. 247 In 1981 and 1982, two new units were recommended (to become operational in 1990-1991); starting from 1983 the Belgian participation in the French units Chooz B1+B2 was included in the calculations, so that only one new unit (‘N8’, in full ownership of the Belgian utilities) was proposed (to become operational in 1993); in 1985 the anticipated participation of the French utility EDF in N8 was counted in (in exchange for the Belgian participation in Chooz B1+B2; this meant that only half of the N8 production capacity had to be counted in); and in 1988 the BCEO proposed for the last time the construction of a new nuclear power plant (after the pullout of EDF, again in full ownership of the Belgian utilities) with an assumed start of production in 1997, next to the participation in Chooz B1+B2 (these units became operational in 1996-1997). Nevertheless, an eighth nuclear power plant was never built in Belgium, for a large part because of a loss of political support following the Chernobyl accident (in 1986), and because an economically attractive alternative (steam-and-gas fired power plants or ‘STAGs’) became available by the end of the ‘80s. 248 Mostly, a choice had to be made between investing in coal-fired units vs. nuclear power plants. In 1988, STAG-plants also became an option. 249 The Belgian productive reserve margin (maximal power output/peak load) rose from 30.9% in 1980 to 58.8% in 1985, due to the start-up of the nuclear units Doel 3+4 and Tihange 2+3.

146

Chapter 2

Critics on the other hand (mostly the labour unions ABVV and ACV, who often played a catalysing role in the debates), relying mostly on the analyses of the federal planning bureau and the research groups funded by the DPWB, argued that government should not agree with the construction of new nuclear capacity as long as an effective policy for the promotion of the rational use of energy was not implemented. On the production side, these groups were in favour of new investments in coal-fired plants (in the first half of the ‘80s), combined production of heat and power (CHP) and district heating. The argumentation, as said, was based for a large part on uncertainty arguments. Firstly, question marks were put concerning the economic growth forecasts used by the electricity sector. The evolution of electricity demand often turned out to be smaller than predicted on the basis of (optimistic) forecasts. More fundamentally, the critics argued that electricity demand forecasts should be based on a thorough analysis of electricity demand, subdivided into different sectors and end uses, and not just on a historical trend analysis. Furthermore, it was argued that the top-down approach put forward by the electricity producers (presenting electricity demand as an exogenous variable) served to hide a number of real policy options for curbing the growing electricity demand – in fact, they accused the electricity sector of promoting electricity consumption (e.g. by promoting the use of electrical water boilers, electrical heating, etc.) in order to make the existing nuclear overcapacity (by the mid-’80s) more profitable, whereas the critics denounced this use of the electricity vector for non-specific purposes as being inefficient from a global energy use perspective. Verbruggen (1982) argues that by using only the expected value of economic parameters (growth in GNP, growth in electricity demand), the (often significant) influence of deviating situations is underestimated. He characterises investment planning as a typical problem of ‘decision making under uncertainty’, and hence concludes that (Verbruggen 1986, p. 315, our translation) …The more important issue is how mathematical models are used in the decision-making process, rather than their built-in characteristics. When investigating an economically and socially important decision, it is absolutely necessary to multiply a broad spectrum of future scenarios with a broad spectrum of policy options, hence enabling policy makers to make a systematic choice based on a broad range of outcomes. The proposed investment plans are very limited with regard to the number of outcomes as well as with regard to the justification of the chosen parameters…

Secondly, regarding the electricity supply side, it was argued that the BCEO systematically gave a preferential treatment to the nuclear option. Critics saw a systematic practice of ‘underestimating’ nuclear-related costs (investment costs, costs of spent fuel management, cost of spent fuel recycling, etc.) while the costs of other options (e.g. coalfired power plants) were ‘overestimated’, particularly in comparison with data from other countries (Erreygers 1984). The lack of interest in other production technologies (CHP, district heating) was attributed to institutional causes (Verbruggen et al. 1988). Furthermore, in view of the significant uncertainties facing investment planning (e.g. regarding electricity demand, fuel prices, etc.), critics of the nuclear option argued in favour of ‘flexible’ solutions – i.e. production means with a shorter construction time and a quickly adaptable power output (whereas nuclear units provide base load power). Thirdly, a lack of a systematic and transparent investigation of the consequences of the proposed


147

investment decisions was denounced, in terms of the expected environmental impacts, industrial activity (e.g. employment levels), diversification of the means of electricity production, consumer prices, and the impact on the Belgian trade balance. In sum, in the course of the debates taking place within the walls of the NCE, critics saw no decisive reason for investment in new nuclear capacity. They argued that the financial means that would become available by not investing in nuclear capacity would be better spent in the construction of smaller-scale CHP-units (possibly connected to district heating networks), coal-fired power plants and, at the end of the ‘80s, STAG-units, purification of the fossil-fired power plants’ flue gasses of (with a remaining lifetime greater than ten years) and if so desired prolonging the lifespan of the best available fossil-fired units (Verbruggen et al. 1988). The specific settings of the debate (simulated shortage of time, the right of initiative of the electricity sector united in the BCEO, model calculations based on data stemming from the electricity sector itself, etc.) however limited the potential for a science-based conflict resolution (or conversely, for a ‘violent’ eruption of conflicts in terms of intractable politico-scientific controversies); hence, a common recommendation by the NCE was only seldom issued. Conflicts were indeed ‘pacified’ – one should add in spite of the disputes on the adequate knowledge base to take legitimate decisions250. Nevertheless, compared to the CCEG discussions took place in a more transparent manner, supported by independent research, which allowed a more equal defence of the diverging points of view of those implicated in the advice procedure. 3.3.3

The contribution of governance by pacification to the sustainability debate

Reformulating our reasoning somewhat, governance by pacification can be characterised as a compromise resulting from the engagement with the different commonwealths implied in the policy domain in question (in the case of energy policy discussed above, most prominently the industrial, market and civic commonwealth). Such arrangement clearly looks to be promising as a basis also for compromise in the field of sustainable energy policy – after all, as set out in chapter 1 (Section 4), we have defended sustainable development as a ‘legitimacy compromise in the making’. Indeed, literature on sustainable development in general and environmental policy making in particular often refer to ‘green policy making’ as a more participatory and integrated process involving new actors and new forums of decision making251. In this regard, an obvious drawback of the historical example of the NCE as a forum for policy support is of course that access to decision making was allowed only to those groups who were included in the mechanism of exchange between government and civil society. Hence, environmental concerns only 250

Institutional guarantees for the provision of independent expertise were indeed not formally set down in the law establishing the NCE. 251 Agenda 21 (Chapter 8, § 8.2) for instance clearly reads: “…Prevailing systems for decision making in many countries tend to separate economic, social and environmental factors at the policy, planning and management levels. (…) An adjustment or even a fundamental reshaping of decision making, in the light of country-specific conditions, may be necessary if environment and development is to be put at the centre of economic and political decision making…”. Agenda 21 is full of recommendations to change policy making towards a more ‘inclusive’ process, with room for discussion, joint design and joint implementation.

148

Chapter 2

entered into the discussion in the NCE when they coincided with the interests of one of the represented groups (e.g. when used as arguments by the representatives of the labour unions against the representatives of the electricity sector). Environmental groups and grass roots movements thus were obliged to pursue a strategy of sometimes seeking coalitions with the labour unions represented in the ‘dominant cooperative frame’ formed by the CCEG and the NCE, whilst also strongly objecting to the fact that the demands for more environmental justice were discussed and framed in a ‘non-representative’ fashion. In recent years however, the government has reacted to the increasing pressures, although, in the words of Bruyninckx (2002, p. 299) generally rather slowly and without enthusiasm, and mostly driven by international promises. The government has for instance in 1997 established the ‘federal council for sustainable development’ (FRDO) (with a working group on energy issues) with representatives of NGO’s, consumer groups, industrial groups, employers and employees as well as scientists (stemming from the different universities) and public officials, next to ‘older’ socio-economic or environmental advisory councils. Thus, at present, the point of view that participation in sustainable development issues is a matter of representation of traditional representative groups in formal advisory bodies, enhanced by the ‘new’ social movements, seems to be dominant. But, as Bruyninckx (2002, pp. 299-300) also notes, these new demands on the participatory nature of state-society relationships have led to new problems and challenges. Some of the more important problems include: • The new consultative bodies do not have the same standing and possibilities to influence policy making. Their place in the decision-making process is rather insignificant compared to the traditional corporatist bodies (such as the CCEG and NCE) for interest group-state negotiations. In the case of energy (electricity) policy, this coincides with the more general limited action potential of government under the conditions of a liberalised energy market (cf. Chapter 4); • The traditional interest groups (employers’ organisations and labour unions) are still rather dominant in the process, and especially when ‘things really matter’ (e.g. in the negotiations regarding the distribution of climate change commitments over different sectors); • The traditional interest groups might have a difficult time establishing working relations with groups they consider marginal (or not really important at best, as witnessed for instance by membership numbers, financial contributions, actual activities, etc.) previously. It is perhaps even precisely because the new consultative mechanisms have less influence than the old ones used to have that the new social movements are ‘tolerated’; • The ‘new’ participants which are now included – i.e. environmental, development and other groups – have difficulties positioning themselves in these new participatory structures: on the one hand they see the opportunity to be part of the new participatory processes as a way to influence policy making; on the other hand, they have to decide how much they are willing to be co-opted and where they want to


149

allocate their usually scarce resources252. This analysis coincides with a wider analysis of new social movements (a notion which carries with it a focus on the instrumental achievement of particular goals) conducting a ‘dual politics’ that is directed towards the state, on the one hand, while also maintaining critical distance from the state, and consolidating new collective identities in civil society, on the other hand (Habermas 1996, p. 370)253. Bruyninckx’s analysis in fact points to the larger problem of participation conceived of as the formal presence in consultative bodies such as the FRDO, which can be traced back to the implicit presuppositions built into the notion of ‘governance by pacification’. Firstly, as witnessed by the historical example, governance by pacification can only function properly when there is some possibility of appealing to a common problem structure – that is, by using shareable kinds of rational argument referred to as scientific research, past experience, etc. (e.g. the proposed investment plans in the NCE). If no independent expertise is available, this places rather strenuous demands on the cognitive abilities of the participating groups – demands which might be difficult to meet for the less resourceful NGO’s. Furthermore, uncertainty cannot get out of hand (e.g. conditions of ‘ignorance’ are difficult to accommodate): a controversial or flawed definition of the situation, of the nature of the problem at hand and its foreseeable future makes the identification of one’s own interests more entangled than usual. Many of the ‘new’ environmental and technological issues however present themselves under forms of more ‘radical’ uncertainty; and moreover, environmental movements have a history of emphasising these uncertainties and bringing them to the fore in public debates. Thus, they might perceive the common problem structure as unduly constraining the scope of the 252

Dryzek (2000) for one is deeply suspicious of the tendencies of states to seek to co-opt and neutralise oppositional movements. He suggest that ‘benign inclusion’ in the state as a general rule should only be pursued when the movement’s goals can be assimilated by the state and when inclusion does not unduly limit the critical or discursive capacities of these movements (Dryzek 2000, p. 83). This analysis is driven by a deep mistrust of the motivation of governments, which Dryzek sees as favouring state ‘imperatives’ (accumulation of power, seeking legitimation, seeking effective implementation of policy measures by reducing the conflict potential) over the goal of democratisation. Thus, according to Dryzek, when the state ‘invites’ new social movements into policy-making arrangements (such as the corporatist arrangement), it is not because it is ‘goodhearted’ enough to give them a voice, but rather to quell opposition and promote state imperatives. Ironically then, Dryzek considers what he calls ‘passive exclusive states’ to be good for democracy because they are more likely to prompt a vital and oppositional civil society (Dryzek 2000, p. 114). However, the problem with Dryzek’s analysis is that he considers ‘state imperatives’ to be givens, whereas a constructivist point of view sees them as ‘reified social relations, practices and understandings’ (Eckersley 2004, p. 163). Whereas Dryzek’s overriding concern is to maintain a vibrant public sphere (and he rightly points out the dangers of cooptation by the state), our concern here goes farther. The question for us is how the political system can learn from the problem-detection capacities of social actors in the public sphere, thus making the state more responsive to critical argument. This also implies a search for less distorted and more inclusive communicative settings, which are also more likely to conduce towards the engagement of ‘new’ social movements with the political system. 253 Rip (1999, p. 116) gives the example of an (in this case) temporary attempt at pacification (organisation of a consensus conference) concerning a controversy on the introduction of genetically modified crops in German agricultural practices. While the structured discussions taking place within the walls of the consensus conference were productive, the environmental groups decided, at one moment, to step out so as to avoid having the eventual conclusions being attributed also to them, which would hamper their freedom of action in the wider world. In pursuing their own interests in this way, they also (inadvertently or intentionally) undermined the legitimacy of the exercise, which was based on getting the contending parties together.

150

Chapter 2

debate, hence undermining the legitimacy of its results. Governance by pacification by itself is therefore unlikely to be sufficiently flexible in order to address the often ‘surprising’ way in which new environmental or technological issues present themselves (Pellizzoni 2003). Secondly, participation is limited to a few groups, represented by the ‘elites’ taken from their organisation. But whereas for the ‘traditional’ groups such as labour unions representativeness can always be checked (e.g. ‘elites’ have to consult their base, they can be ‘sanctioned’ through social elections), this remains problematic for the ‘new’ social movements. For instance, if ‘green’ NGO’s are supposed to represent ‘the environment’ or ‘future generations’, how can one actually check if they represent them ‘faithfully’ (i.e. they speak as they would speak if they would be present themselves in the political negotiations)254? How are we to know what they truly ‘want’ (cf. Section 3.5 – Chapter 1). Moreover, in the case of environmental or technological controversies, there is always the possibility of previously ‘silent’ subjects (social groups or individual citizens) suddenly demanding to be heard (cf. Beck’s concept of ‘sub-political’ action). The logic of the corporatist arrangement is then at risk of being overtaken by sudden changes in the interpretation of the issues as ‘internally’ agreed by the established community of representative elites, with possible consequent criticisms and unexpected failures in the implementation of regulatory solutions. One might add that ‘new’ social movements are precisely mostly active at this ‘lower’ level of local action, seeking flexible and often fleeting coalitions (and the new communication technologies have greatly enhanced the possibilities of ‘global’ action on ‘local’ issues) with firms, action committees, citizen groups, critical scientists, research institutes, etc., as dictated by their strategic assessments255.

254

Therefore, Dobson’s (1996b) somewhat provocative idea of proxy representation of both non-human animals and future generations in representative assemblies by deputies elected from the environmental sustainability lobby is likely to cause more problems than it pretends to solve. This results from an uneasy reduction of all the possible ways in which ‘the environment’ might be engaged in social coordination (involving all the different ‘commonwealths’) to a mechanism of representation (voting, elections) taken from only the ‘civic commonwealth’. 255 The history of the nuclear controversy in Belgium proves to be a very good example (Leroy 1979; Laes et al. 2004c, pp. 48-69): what began in 1974 as a local conflict over the siting of a nuclear power plant quickly became a much more widely ranging controversy involving other local environmental groups, critical scientists, local representatives of labour unions; and ultimately certain arguments used by the anti-nuclear movement even gained footage on the level of government representatives. Sociology traditionally pays great attention to people’s mobilisation against exclusionary deliberative arrangements. In the case of the sociology of the environmental movement, early analyses (mostly from a neo-Marxist or anti-modernist perspective) saw these as the driving force of transition towards a more ecologically responsible society – a transition that existing institutions were unable and unwilling to prompt (Leroy 2001). More recent sociological contributions (e.g. Torgerson 1999) have argued that the environmentalist movement should become less concerned with an instrumental approach to politics (i.e. strategic action for the furtherance of particular goals) and more with the promotion of the ‘green public sphere’ (i.e. open dialogue centred around sites of environmental concern) as a goal in itself. Yet Torgerson also does not dispense with the idea of a certain instrumentality in the environmentalist movement, if only because the concept of a ‘movement’ carries within itself the idea of goaloriented change, and hence, strategic action. Furthermore, as Haacke (1996) has argued, the ‘green public sphere’ does not always arise naturally: it has to be actively constructed by the use of public power – a move which will invariably be contested, particularly by social forces who stand to loose.


151

A third problem could then be that, despite a commitment towards maintaining ‘open discussion arenas’, the lack of standing or real political influence of the new consultative bodies, combined with the usually limited institutional capital of the ‘new’ social movements, prevents them from being involved more in action-oriented activities at the level of citizens or groups. It is interesting to draw the parallel here with the ‘traditional’ social groups: they gained a seat in consultative bodies only because (and after) they had built up the necessary support, membership, and professionally staffed organisations through action. It remains an open question whether this reversal of the ‘normal’ course of events might not lead the ‘new’ social movements to seek an enlargement of public support through social action rather than ‘sitting in meetings’ or ‘voicing opinions’. 3.3.4

Conclusion

‘Governance by pacification’ has undeniably shown its usefulness in the Belgian policy context characterised by deep divisions. Nevertheless, be it from a moral (e.g. pacifying relations between social groups through more or less ‘distorted’ communicative settings), an empirical (e.g. a possible lack of sufficient learning capacity in the context of environmental or technological controversies) or a pragmatic (e.g. a possible lack of real influence on decision making) point of view, one can, with Bruyninckx (2002) only be sceptical about the potential of such advisory councils for pushing towards a change in the direction of sustainable development – at least when they are not connected to a wider participatory practice. In the concluding chapter of a research project on participatory decision making in the Belgian context, Bruyninckx (2002, p. 299) concludes that …It is safe to say for the moment that there is not enough evidence to substantiate the claim that participation per se has a large positive impact on the effectiveness of environmental and sustainability oriented policy programs. The evidence seems to suggest that participatory bodies do perform their formal functions fairly well and have been effective in that sense but that they are still far removed from being catalysts for farther reaching participatory practices…

According to Bruyninckx, one of the main reasons for failure is that the current dominant conceptualisation of participation is based on a narrow definition which leads to a likewise narrow operationalisation. A broader view of participation centres on a broader definition of social action (not limited to action in political arenas), oriented at the modification of social institutions (including, but not limited to formal political bodies such as advisory councils), linked to the local context of many social actions (and thus not limited to the ‘macro-context’ of the political system) and implied in more phases of political decision making (not limited to the ‘agenda-setting’ phase) Governance by pacification, when taken by itself, entails a logic of gradual transition, an intentional transformation ‘from within’, using a well-known and commonly applied ‘dominant cooperative scheme’, aimed at providing stability also in the new context. But intentional design cannot be viewed in isolation from ‘emergent’ design (i.e. a process of de facto design in which new practices, procedures, norms, and institutions emerge). Emergent design drives on the goals and approaches implicit in the actions and interactions as they occur, and which can be, but need not be, made explicit and reflected upon. But if

152

Chapter 2

the state is really concerned with stimulating social change in the direction of more sustainability, we suggest it should not only be concerned with simply laying down a framework of certain procedures that enable a certain form of participation (i.e. consultation of ‘representative groups’) as an appendage to an otherwise virtually unchanged political reality. The state should then, next to constructing this more cumbersome bridgehead between the public sphere and the state (allowing the heavy traffic of organised interests to pass over), also be concerned with actively building communication lines with society by means of more decentralised action-centred strategies of political empowerment – a point of view generally advocated by proponents of ‘deliberative governance’.

3.4

3.4.1

Deliberative governance


A central element of different views on deliberative democracy is an emphasis on the emancipatory potential of the informal debates in society (the public sphere) that ‘surround’ the more formal political debates, accurately described by Michelman (quoted in Korthals (1994, p. 52)) as: …Much of the country’s normatively consequential dialogue occurs outside the major formal channels of electoral and legislative politics, and (…) in modern society those formal channels cannot possibly provide for most citizens much direct experience of self-revisionary, dialogic engagement. Much, perhaps most, of that experience must occur in various arenas of what we know as public life in the broad sense, some nominally political, some not: in the encounters and conflicts, interactions and debates that arise in and around town meetings and local government agencies; civic and voluntary associations; social and recreational clubs; schools, public and private; managements, directorates and leadership groups of organizations of all kind; work places and shop floors; public events and street life…

As it turns out, many ‘green’ political theorists have lately turned their attention towards the paradigm of deliberative democracy (see e.g. Holemans 1999; Torgerson 1999; Eckersley 2004), as they advocate the rise of a ‘vibrant green public sphere’. The primary reason seems to be that deliberative democracy eschews the view of democracy (discussed under section 3.2) conceived of as a mechanism for the aggregation of private interests or preferences or power trading and bargaining between ‘powerful’ social actors in favour of the paradigm of unconstrained dialogue among free and equal citizens256, and this condition in itself is perceived as being more conducive to the prudent protection of public goods

256

Proponents of deliberative democracy do not, of course, pretend that ‘traditional’ elements of modern pluralist politics (e.g. bargaining between interest groups, majoritarian elections, etc.) can be replaced. Rather, what deliberative theorists reject is the apparently inevitable domination of society by powerful organised interests and the exclusion of any substantive community-based sense of the common good from consideration in decision making. Habermas (1996) for instance situates his account of deliberative democracy between this liberal and republican (which he accuses of posing unrealistic demands of cultural identity upon the public sphere) conceptions of political reality.


153

(such as a healthy environment)257. A recent surge of interest in deliberative forms of technology assessment is also noticeable (see e.g. Renn et al. 1995; Mayer 1997)258. Deliberative democracy thus shares with Rawls the search for a form of normative rationality as a solid base of allegiance for liberal democracies. The insistence on the possibility of grounding authority and legitimacy on a form of public reasoning and a rationality which is not merely instrumental but also has a normative dimension equally results in a strong separation between ‘mere agreement’ and ‘rational consensus’ (in Rawlsian terms constituted by the overlap of ‘reasonable comprehensive doctrines’); and the proper field of politics is identified with the exchange of arguments among reasonable persons guided by the principle of impartiality. As far as Habermas (1996) is concerned, he defends what he claims to be a strictly proceduralist approach (the so-called ideal speech situation – cf. infra) in which no limits are put on the scope and content of deliberation (herein lies one of the main differences with Rawls’s political philosophy, who emphasises the role of the ‘original position’ that forces people to leave aside their particular interests). According to Habermas, the procedural constraints embodied in the ideal speech situation will eliminate the positions to which the participants in the discourse cannot agree. There are many different theoretical versions of deliberative democracy, and the literature is generally concerned with fleshing out the details of a contrast with accounts of democracy focused on aggregative procedures and/or to offer a fuller specification of the reasons to prefer deliberative democracy. However, our aim here is not to go too deep into the more theoretical aspects, as we are more concerned with suggesting ways in which the conception of deliberative democracy may be employed as a critical vantage point or as a model for institutional design fit for our purposes (that is, designing governance structures in the context of (technological) decision making under conditions of uncertainty)259. We will therefore limit ourselves here to a

257

This inference seems to be supported by research in social psychology: solidarity-oriented groups (i.e. groups committed to the deliberative ideals of mutual respect and willingness to listen to each other) seem to be more supportive of equity-based discussion outcomes (for an overview, see e.g. Miller 1999, pp. 61-92). 258 There are several kinds of deliberative practices specifically aimed at technological decision making employed in various contexts, going under the label of ‘participatory technology assessment’ (pTA). Examples include consensus conferences, citizen juries, focus groups, participatory design, etc. (for an overview, see Joss and Bellucci 2002). All aim to bring into contact non-expert citizens with specialists, experts and policy makers. The Flemish ‘office of technology assessment’ (viWTA) is generally committed to developing pTA methods as an input for decision making in the Flemish regional parliament. 259 In particular, it is not our intention here to go into the details of Habermas’ immense work. In view of its cardinal importance for deliberative democratic theory, let us nevertheless give some essentials (for good summaries, see also Webler 1994; Mortier and Raes 1997, pp. 222-235; and von Schomberg and Baynes 2002). Habermas has sought to explain the changes in modern societies as processes of rationalisation, occurring in the three different domains of science/technology, law/morality and art/art criticism. His diagnosis of the current problems of modern society is that of a one-sided and uneven rationalisation through the predominance of strategic agency (based on scientific/instrumental rationalisation) in the state and economic system over communicatively generated rational consensus in the life world (involving the three kinds of rationality). When asked how to rectify these problems, Habermas’ answer is to ‘re-politicise the public sphere’ (a notion Habermas uses to denote the area of public life where intersubjective agreement on values can be reached in order to solve socio-political or practical problems); and he has laid out the requirements of discourse that can fulfil this prescription in a theoretical conception called the ideal speech situation (“Herrschaftsfreie Dialog” in German).

154

Chapter 2

broad characterisation of some ‘family resemblances’ shared by deliberative democratic accounts: unconstrained dialogue, inclusiveness and social learning (see e.g. Dryzek 1990; Habermas 1996; Eckersley 2004; Festenstein 2004). These ideals will be discussed next, before continuing with some of the difficulties of translating them into political reality: • Unconstrained dialogue: this comes down to a requirement that only justified arguments should be allowed to sway participants in the dialogue. This requires that participants give reasons for their proposals, reservations or objections to enable the public testing and evaluation of opposing claims. The requirement of free dialogue also necessarily encompasses the requirement of publicity. Thus, what is meant to matter in deliberation is quite simply the force of the better argument, and not for instance bribery, coercion, bargaining power, the status and/or authority of the speaker, or allowing insufficient time for the dialogue to develop (cf. the debate settings in the NCE discussed in section 3.3.2). Participants in the dialogue are expected to responding to the reasons and arguments of others qua reasons and arguments (and not their bargaining power, status, etc.). The implicit goal of dialogue – mutual understanding – can thus only be reached on the basis of the ‘unforced force’ of the better argument; • Inclusiveness: deliberative democrats usually insist on the quality of impartiality (or better, enlarged thinking) as an essential requirement for a legitimate dialogue, since the point is precisely to weed out purely partial or self-interested arguments in favour of arguments that can be defended as acceptable to all those potentially afflicted by a decision. Participants in the dialogue are under the obligation to offer arguments that are persuasive to all others; this does not mean that they agree with the beliefs of all other interlocutors, but simply that individuals must defend their propositions in terms that they sincerely expect to be persuasive to others who share that commitment (Cohen 1997, pp. 75-76); • Social learning: the social learning dimension of deliberative democracy flows from the requirement that participants are committed to modifying proposals in light of the arguments put forward in the deliberative process. This is, to some extent, a restatement of the requirement of free and unconstrained dialogue, which stated that participants are

These requirements – the roots of social cooperation – are according to Habermas found through the investigation of how language is used (hence, the performative aspect of ‘communicative action’) in everyday situations (i.e. ‘universal pragmatics’). The ideal speech situation is then Habermas’ attempt to describe the presuppositions that discourse participants must hold before communication without coercion can prosper. Thus, Habermas sees emancipatory potential in communicative practices that aspire to (but can never be expected to achieve fully) the conditions of the ideal speech situation. These are (Mouffe 1999, p. 750): 1) participation in such deliberation is governed by the norms of equality and symmetry: all have the same chances to initiate speech acts, to question, to interrogate, and to open debate; 2) all have the right to question the assigned topic of the conversation; 3) all have the right to initiate reflexive arguments about the very rule of the discourse procedure and the way in which they are applied and carried out (Habermas has over the years worked out the ideal speech situation in a number of ways, but the above points seem to capture the essence). Of interest for us is also Habermas’ view on practical discourse (discourse about norms), as one particular kind of speech (next to theoretical, explicative and therapeutic discourse). Habermas maintains that in making evaluative statements, one implicitly relies on a universal principle that applies to the redemption of the statement in question. The principle is: “…no norm can be considered valid unless all those affected can accept the consequences associated, to the extent that the consequences can be known…” (quoted in Webler 1994, p.49).


155

moved to change their positions only according to the force of the better argument. But only under a narrow rationalist interpretation is the commitment of social learning collapsed into the previous ones: such interpretation sees deliberation merely as a process for airing reasons and filtering out the bad ones. Deliberation in this narrow view would come down to patiently listening to other people’s arguments and then explaining why they are wrong, thereby denying the possibility of dissent even after all reasons are sufficiently laid out and discussed. Rather, exchanging arguments must be seen as endowing the outcome with transparency, even to those who are not convinced by the reasons on offer for it (Festenstein 2004, p. 296). Thus, the social learning dimension highlights also one of the claimed advantages of deliberative democracy, that is, the possibility to make decisions that are adaptable and self-correcting in view of new circumstances, information, or revised arguments. These three features of deliberative democracy thus arguable make it a serious candidate for dealing with complex and uncertain sustainability problems and concerns. This is because it invites reflexivity and the continuous public testing of claims on the basis of their legitimacy. Applying a model of deliberative governance would for instance enable the critical and public testing of the normative judgments made also include in scientific assessment (cf. Section 3.1). Deliberative governance is moreover flexible in the sense that it makes no a priori distinctions between expert knowledge or other kinds of knowledge, and hence it is capable of integrating arguments stemming from different knowledge bases. Furthermore, deliberative governance is defended (e.g. by Korthals 1994) as the best model for reaching mutual understandings about common norms, especially concerning longrange and generalisable interests such as sustainable development, in the sense that it at least strives to find ways of mutually accommodating (rather than trading off) the needs of the present and the future and the human and non-human world. Risse (2000) argues that as soon as ‘common knowledge’ is absent – that is, if actors are uncertain about their own identities, interests and worldviews – argumentative rationality becomes necessary (and not merely desirable) for developing trust in the authenticity of the speech acts of other political actors, for advancing an argumentative consensus on the definition of the situation at hand, and for acquiring a collective understanding of the normative framework underlying the policy questions260. Thus, Risse’s analysis suggests that a logic of communicative rationality is likely to arise even in situations which are generally believed to be far removed from the ideal speech situation, thereby adding to the force of the normative argument in favour of truly deliberative conditions an empirical one. Nevertheless, the deliberative democratic model is not without its critics. After all, defending a normative ideal against the imperfect real-world practices of liberal democracies begs many questions, not the least of which is how the ideal might be realised in practice (despite Risse’s assertions). Parallel to our discussion of the ‘governance by aggregation’ model, we will outline two broad categories of criticism: the ‘critique of

260

Risse’s analysis is developed in the field of international negotiations (e.g. about climate change), but the general line of argumentation holds true for any ‘complex’ decision context.

156

Chapter 2

inadequate representation’ and the ‘institutional critique’. The first one addresses the question of the representation of all affected entities (human and non-human) in the deliberation process; while the second addresses the institutional and procedural challenges associated with the move from an unconstrained ideal dialogue setting towards political processes of decision making. This section will not include the discussion of a historical example taken from the Belgian energy policy context since experience with the participatory settings advocated by deliberative democrats is only in its infancy stage261. The governance model we set out in chapter 5 can be seen as building upon deliberative democratic ideas; but, in view of the general lack of experience, it will be of a largely programmatic nature. 3.4.2

The contribution of deliberative governance to the sustainability debate

3.4.2.1 The critique of inadequate representation The critique of inadequate representation is built upon an intuition that – at first sight – resonates largely with the Habermasian concern for an ideal communicative community (when applied in the context of decision making on environmental or technological issues), namely that all those potentially affected by a risk should have some meaningful opportunity to participate or otherwise be represented in the making of the policies or decisions that generate the risk (Eckersley 2004, p. 111). It is also in line with ethical guidelines of ‘free and informed consent’ in the context of an avoidable exposure to risk (excluding risks involving no human agency such as natural disasters) as a baseline against which deviations should be justified. Nevertheless, upon further scrutiny, this claim entails a direct challenge aimed at one of the core ideals of deliberative democracy, namely that (Habermas 1996, p. 127)262 …according to the discourse principle, just those norms deserve to be valid that could meet with the approval of those potentially affected, insofar as the latter participate in rational discourses…

The challenge, first formulated by ecological thinkers (e.g. Eckersley 1990; Goodin 1996) accusing Habermas of anthropocentric prejudices, of course lies in the argument that the opportunity to participate or otherwise be represented should literally be extended to all those potentially affected, regardless of generation or species (instead of being limited to those capable of ‘speech’ or ‘rational judgment’). A stern challenge indeed, since for instance for the proposal to build a nuclear reactor, the spatial community at risk would be an area possibly spanning several continents, and the temporal community at risk extends

261

At the time of publication of this dissertation, viWTA – an institute generally involved with finding and implementing new ways of public involvement in technology assessment – was defining its working programme for the coming years under the overarching theme of ‘Energy and climate’. One of the preparatory reports written for viWTA (Neyens et al. 2004) contains some preliminary findings about participation in the field of renewable energy; the report on the history of the nuclear power controversy (Laes et al. 2004c) gives some insights into local participation in the process of finding an adequate site for the disposal of low and intermediate level radioactive waste. 262 This is the formulation of the discourse principle as found in “Between Facts and Norms”. However, Habermas has reworked his definition of the discourse principle over the years. For a theoretical exploration and critical investigation of the major consequences of this re-formulation, see Kettner (2002).


157

almost indefinitely into the future. A proposal to build a gas-fired power plant puts the whole world at risk, through the emission of greenhouse gases263. And so on – the above mentioned examples largely suffice for conveying the general idea that the extension of the communicative community to a wider involved constituency certainly does not make decision making easier than it was before. A quite reasonable objection would of course now be that it is rather difficult to fathom how all those potentially affected by a risk (including the not yet born, non-human species, etc.) would somehow sit around the table in order to arrive at a consensus on the best possible course of action. And yet the answer to this quite formidable problem, as we will argue in the following paragraph, lies in simply replacing ‘speech’ with ‘representation’… In fact, the crucial problem with the Habermasian account of moral validity – as analysed by Eckersley (2004) from the moral point of view, and Latour (2004a) from the epistemological point of view – is that it rests exclusively on the bedrock Kantian ideal that all individuals ought to be respected as ends in themselves. Habermas’s concern for the safeguarding of the life world (including the mechanisms for generating social cohesion through communicative interaction) from the intrusion of instrumental reasoning leads him to separate ever more strongly the ‘social world’ made up of humans endowed with unquestionable ‘speech’ from the ‘mute’ objective world, whilst locating all emancipatory potential in the former. This leads him to conclude that all of nature (which is part of the ‘objective world out there’) is effectively a passive material upon which humans leave their mark, ideally governed by moral rationality located firmly within the linguistic structures of everyday language. Eckersley’s moral critique on this approach rests on a postulated extension of the Kantian principle towards all ‘differently situated others’, on the grounds that just because they might not be able of providing consent, they should not be subjected to policies and decisions that ‘impede their unfolding in their own distinctive ways’ (Eckersley 2004, p. 120) – an argumentation we have already developed in more detail under section 3.5 of chapter 1, where the idea was developed that the search for the sustainable commonwealth should start from a first step of convoking a collective of humans and non-humans, involved in uncertain associations. Of necessity then, since the first-best solution of letting all non-human others and future generations speak for themselves is impossible, then we should accept the second-best solution of allowing their interests to be represented by others who can speak. This then is the point where the moral challenge meets the epistemological challenge head-on. For, if we allow someone to speak ‘on behalf of’ or otherwise incorporate the interests of the differently situated others, the attention immediately turns towards the question how we can be assured that this person doing the representation really knows what is in the best interest of nature or future generations. Luckily, on this point again we can take away much from our previous analyses elaborated in chapter 1. For the epistemological challenge is precisely answered by the

263

For the moment, we are concerned only with the potential to be harmed, and not with the particular (degree of) damage incurred.

158

Chapter 2

constructivist position we have adopted as an adequate framework for answering the complex problems posed by demands for more sustainability. Let us again briefly summarise the central tenets of the constructivist view here. Constructivism rejects uncritical realism (including the referential view on scientific truth) on the grounds that our knowledge of nature is irrevocably mediated – there is no unproblematic direct access to the ‘real world out there’. It does not deny the existence of ‘nature’ as an extra-discursive reality; it simply acknowledges that we do not have any shared access to this reality other than through a large number of devices (laboratory experiments, measuring devices, discourse, discussions, etc.) – Latour (2004a) speaks of ‘speech impedimenta’264. On the other extreme, there is no question of conflating everything in the category of ‘social representations’ (i.e. social constructivism) and linguistic agency, because such view makes the notion of (human) ‘speech’ unduly unproblematic. Constructivism brackets from critical investigation what makes people ‘speak’ – i.e. the large number of conceptual devices (including the ‘external world’) people share in order to be able to communicate at all. If this is true for everyday communication, it holds a fortiori for communications in political forums. Latour (2004a) cleverly points out the parallels between a political representative claiming to speak for ‘the public’ and an expert claiming to speak for an ‘environmental asset’: both need consultation mechanisms and have to put ‘work’ into assuring that they faithfully represent their constituency (in the case of the politician, by organising meetings, listening to trusted intermediaries, etc.; in the case of the expert, by attending scientific meetings, doing laboratory work, etc.); and one is certainly not more mysterious or unproblematic than the other. In sum, governance for sustainability cannot avoid the category of ‘representation’; therefore Eckersley (2004, p. 112) proposes to replace the words ‘insofar as’ in the above formulation of Habermas’s discourse ethic with ‘as if’: “…just those norms deserve to be valid that could meet with the approval of those potentially affected, as if the latter participate in rational discourses…”265. This modification, however small, proves to be crucial in understanding some of the difficulties theorists wishing to extend the Habermasian framework for the purposes of formulating quality criteria for expert-layman dialogues have encountered. This problem is, as Festenstein (2004, p. 296) puts it, one of conceiving a model for public decision making that is at once sufficiently cognitive to make it truly deliberative (i.e. to bring out ‘the better argument’) and also sufficiently responsive to the positions of individual citizens to count as democratic. In short, in the words of Webler (1994), there is a possible conflict between ‘competence’ and ‘fairness’ (with the latter value stressed by Habermas); and both 264

Habermas has failed to take up these central findings of constructivist accounts of the scientific enterprise, maintaining instead a strong separation between ‘nature’ (known to us through objective science) and interaction (the social realm of public discourse) (Feenberg 1999). 265 Vermeersch (1988, p. 51, our translation) also formulates a version of a Kantian imperative in the context of sustainability: “…For each of my actions, I have to be able to say: if everybody acts in the same way as I do and continues to do so over the centuries, our earth should still be able to unfold in an undisturbed way. Each act which can be shown to transgress an irreversible limit when generalised to large sections of the world population or maintained during several centuries, is an act that goes against the ethic of a responsible view on globalisation...”. This formulation goes too far in the sense that it presupposes a general knowledge of scientific ‘facts’ about ‘the carrying capacity of nature’ – a presupposition that we have already criticised for being unrealistic in chapter 1.


159

Festenstein and Webler (and possibly also others) struggle with finding solutions to the problem of simultaneously upholding both values whilst staying within the framework of communicative rationality (i.e. not imposing external quality criteria upon a dialogue process). Through the combined forces of (critical) ecological ethics and the sociology of science and technology it now becomes clear that the solution to this problem lies in accepting ‘enlarged’ representation as a central category of the discourse ethic. For this forces the discourse participants to think of the ‘best’ way of representing the interests of differently situated others; and this means using some inventive thinking, imagination and state-of-the-art scientific knowledge. This also means making visible (as far as possible) the construction work done by scientists in order to arrive at their solutions, since this work cannot be kept outside of the boundaries of the debate (cf. Section 3.1), and uncertainty arguments will play a pivotal role in assuring ‘faithful’ representation. We should stress here that the argument for an enlarged representation does not come down to the promotion of ‘risk-free’ decisions; rather, the argument is that decision involving possible harm to others should be justified to the satisfaction of all those involved in the deliberation (through their representatives). In general, expanding the range of information relative to the requirement of sustainability (along with the possibilities of critical interpretation) would be the best response to the inevitable limitations associated with any form of representation (cf. the important role of independent research groups in the debates taking place in the NCE discussed under section 3.3.2). However, ‘pure’ scientific understandings should also be placed alongside the more vernacular understandings of risk or environmental impacts, and adequate mechanisms of representation should also be foreseen for this kind of knowledge (e.g. one can think of various pTA tools). It is clear that this view is politically and institutionally challenging. For instance, it is unlikely that the uncertainties, controversies and interdisciplinary struggles within and between scientific and vernacular understandings of environmental problems can be solved to the general satisfaction of all those involved; and hence, the question of how to arrive at decisions under conditions of complexity and a pluralism of values will typically arise. Also, demanding that the parties proposing a risk-generating activity – whether private property holders or public authorities – justify their decisions to potential victims or their representatives in an open and critical communicative setting will require specific institutional provisions which might run counter to some powerful vested interests. These issues will be raised in the following section; a more detailed proposal for solving them will however be postponed until chapter 5. 3.4.2.2 The institutional critique Let us take as a point of departure for developing the institutional critique the deliberative theory of law, democracy and the state developed by Habermas (1996) in order to spell out the consequences of his discourse ethic (originally intended to explain communication in the public sphere outside the institutions of the state) for present-day democratic politics266. Put (very) briefly, in this work Habermas opposes the idea of deliberative democracy as an 266

For a discussion of some of the central themes of this work, see also von Schomberg and Baynes (2002).

160

Chapter 2

ideal procedure that should be mirrored in all social institutions (and even in all government institutions) as much as possible. He reserves the actual realisation of the ‘discursive mode of socialisation’ for the legal community (a notion including the constitution, legislative processes and the actions of the judiciary), while nonetheless continuing to point out the more spontaneous associations and movements within the public sphere as being able to both utilise and radicalise political communication structures, through their unique capacity for indicating and dramatising problems in ways that can be taken up by political parties and legislators. According to Habermas, legal processes precisely only step in when other integrative mechanisms (e.g. social integration in the public sphere, functional integration in the economy and administration, etc.) become ‘overburdened’; in doing so, they solve the same kind of problems as the processes they replace, but on a more abstract ‘reflexive’ level. Therefore, the validity claims made in legal norms fundamentally remain rooted in the same linguistic structures of ‘normal’ everyday language. In Habermas’s understanding, deliberative politics remains a part of a complex society, which however, as a whole, resists the normative appeal practiced in the legal community (Habermas 1996, p. 302). A common and recurring argument against the feasibility of this account of deliberative governance has been that it is too idealised, by failing to take into account the political realities of present-day society – most notably the dominant role of mass communications and functional exigencies of the economy; and a reliance on the delegation of political power to representatives, who deliberate in institutional settings often far removed from the ideal communicative forums envisaged by the proponents of deliberative democracy (e.g. relationships based on power and influence, a lack of other-regarding orientation by the actors involved, etc.). Habermas (1996, pp. 321-323) has responded to this critique by primarily pointing out the strength of the deliberative model in providing a counterfactual ideal (of necessity somewhat removed from ‘real’ conditions) for exposing (Habermas 1996, pp. 327-328) …how much of the normative countersteering of constitutional institutions can compensate for the communicative, cognitive and motivational limitations on deliberative politics and the conversion of communicative into administrative power. One must ask how much the social facticity of these unavoidable inertial features, even when they are already taken into consideration in the formal structure of the constitutional state, provide a point where illegitimate power complexes that are independent of the democratic process can crystallize. One must investigate the degree to which, in particular, the power concentrated in social subsystems, in large organizations and public administrations, inconspicuously settles into the systemic infrastructure of the normatively regulated circulation of power, and one must investigate how effectively the unofficial circulation of this unlegitimated power encroaches on the constitutionally regulated circulation of power…

Real-world politics according to Habermas remain caught in the tension between ‘facts and norms’ (hence the title of his book), and his main goal has been to point out that the laws and norms enacted by democratic states have a legitimating force so long as the possibility remains of discursively redeeming them (Habermas 1996, p. 145). Ecologically inspired critiques in response to Habermas’s reconstruction of modern democracies in terms of communicatively generated power have generally proceeded along two lines of argumentation. In her discussion of what Eckersley (2004, p. 150) calls “…Habermas’s highly idealised yet deeply ambiguous reconstruction of democracy…”,


161

she provides us with a concise version of these arguments: firstly, she says, on the one hand Habermas has tracked the decline of the public sphere (which is, in present-day societies, more and more heterogeneous, with no shared class consciousness or general belief system), yet, on the other hand he keeps pointing to its possibilities to ‘shift the entire system’s mode of problem solving’; and secondly, Habermas has argued that the discourse principle is rooted in universal linguistic structures of everyday communication in the life world, and yet it can only be upheld and realised within the framework of a democratic legal system that is necessarily limited in terms of its temporal, spatial and ethical boundedness. As Eckersley summarises, the problem is that Habermas seems content to resign himself to a sociological analysis – a reconstruction from the perspective offered by the implicit presumptions of communicative action – of the inherent tensions and limitations of democratic states, without seeking to explore how these tensions might be alleviated – i.e. by adopting the perspective offered by the discourse principle as a constructive vantage point. Reformulated in a more positive sense, Eckersley’s critique on Habermas’s theory comes down to two challenges: firstly, ‘given that a flourishing public sphere is crucial for democracy, then how might this potential be furthered?’; and secondly, ‘how the democratic will-formation, despite its inherent limitations, might more closely approximate moral and ethical rather than merely pragmatic modes of reasoning?’. Each of these questions will now be taken up in turn. As it turns out, the first question largely resonates with the concern for a ‘production of voices’ representing differently situated others explored in the previous section. On this point, Habermas remains remarkably complacent: he merely stresses the importance of the existence of a system of civil and political rights and a pluralist and tolerant political culture. Furthermore, he offers little assurance that the arguments emanating from the public sphere will be heard and/or acted upon, as he leaves this up to ‘traditional’ influencing mechanisms on political opinion-formation such as elections and the role of political parties. This is somewhat strange in view of Habermas’s sustained emphasis on the crucial role of dialogues going on in the public sphere as a critical resource for the political system – a role which cannot simply be taken for granted, since in modern Western democracies, it is typically the executive or powerful lobbies which exercise considerable influence over the direction of parliamentary debates267. Thus, Habermas’s reliance on the practical judgements of political leaders and/or policy brokers to enact upon discourses originating in the public sphere is likely to fall short of the requirement to mobilise society towards a more sustainable future. As variously argued by Pellizzoni (2003) and Mouffe (1999), this is a result of Habermas’s conservative view and definition of the public sphere. As these authors argue, the existence of a public sphere should not be conceived of as guaranteed merely by the extension of civil rights, but by the availability of

267

As we will for instance show in our reconstruction of the policy development cycle in the context of the Belgian decision to phase out nuclear power (Chapter 4). The history of the nuclear power controversy in Belgium furthermore shows how in the past a powerful lobby of electricity utilities, associated engineering firms, financial holdings and construction industry enjoyed a large discretionary power over energy policy decisions (Laes et al. 2004c).

162

Chapter 2

an ensemble of practices that make the constitution of democratic citizens possible268. Pellizzoni (2003, p. 348) convincingly argues that the basis for social integration does not lie in the rational sharing of political values, but rather in the joint deployment of individual resources in response to common problems. What is crucial for a lively public sphere is the ‘experience of cooperation in the division of labour’, as this leaves more room for developing forms of life that foster identification with democratic values (rather than abstract subjectless rational consent), predicated on socially recognised contributions to cooperative processes269. Thus, the crucial role of the state would precisely lie in promoting such deliberative practices and discursive arenas for common problem-solving and policy implementation. Undoubtedly, this is a challenging endeavour, but there seems to be at least a growing awareness that more ‘narrow’ approaches increasingly pose difficulties270. The ‘production of voices’ (also of the ‘non-human others’ and future generations) can be implemented through formal (e.g. environmental impact assessments, provision of rights of environmental information, local and regional environmental councils, environmental monitoring, etc.) as well as informal (e.g. consultative platforms, information campaigns, etc.) mechanisms. Taken by themselves, the mere provision of these mechanisms of ‘diverse representation’ do not however guarantee a solution to the second question raised by Eckersley, that is, assuring that the voices emanating from the public sphere will be effectively taken into account by the responsible decision makers – after all, there might be many reasons why political representatives may find it difficult to engage in ‘enlarged thinking’ (this is yet another form of the ‘problem of representation’, as lack of motivation, experience, information, etc. all might play a role). In chapter 5, we will argue that the precautionary principle can be enlisted as a means for minimising (not of course eliminating) these problems of representation, as it offers the possibility for 268

On a theoretical level, this is also where constructivists disagree with Habermas’s point of view. Habermas sees the real-world conditions of social and power relations, language, culture, limits in time and resources for communication, etc. as empirical obstacles to the establishment of an ideal communicative community, whereas constructivists see these real-world practices precisely as being constitutive of identity and community. Thus, for instance, procedures for democratic interaction cannot simply be derived from pure communicative structures and then applied to particular cases. Rather, constructivists maintain that procedures only exist as a complex ensemble of practices. One can reflect on these practices and distil from them a ‘rational’ procedure of achieving cooperation, but this can only be a local and temporal solution; fundamentally, it is because the procedures are inscribed in shared forms of life that they can be followed and accepted. Now, the difference is perhaps not as dramatic as it sounds, since in our context, we need not ask the question whether Habermas’s view on linguistic structures having an implicit orientation towards mutual understanding has been present in human communication since times immemorial or whether it is a particularly modern expectation resulting from the ‘democratic life form’. It can safely be accepted (even for constructivists) that the principles of political equality and autonomy Habermas derives from his linguistic investigations have indeed become general normative expectations or aspirations. The constructivist however is more interested in other questions: given an acceptance of these principles, how are they being played out in reality? 269 Compare to De Dijn (2003, p. 61, our translation): “…If abstract ideals and principles would really gain the upper hand over actually experienced values and norms, then humanity and society would disintegrate. For ideals and principles are much too deficient in ‘instinctive’ elements, i.e. elements that are related to the intimate familiarity with complex symbolic meanings…”. 270 One can think of local Agenda 21 initiatives, local partnerships with citizens for the siting of low-level waste repositories, the ‘Publiforum’ organised by the ‘Flemish institute for technology assessment’ on GM food (viWTA 2003), participatory experiences in local renewable energy projects (Neyens et al. 2004), etc. However, we are far from claiming that participatory practices are already common practice. On the contrary, these examples seem to be still rather marginal, and research on the actual effectiveness or added value of these practices is still in an infancy stage (Rowe and Frewer 2004).


163

design of a number of institutional innovations which force decision makers to confront scientific uncertainty and long-range environmental problems head-on. Now that the precautionary principle has emerged in international, community and national legislation, it can be reflected upon and tried out as a design principle. We are here in agreement with de Sadeleer (1999), who speculates on the possibility of a shift in the dominant political cooperative scheme concerning environmental and/or risk issues leading to a new design hierarchy with a precautionary principle as a charter and/or a culture-embedded guideline. Of course, the precautionary principle needs to be applied discursively in particular cases: the design hierarchy needs to be articulated, which gradually becomes more concrete as it is embedded in concrete practices. The goal can of course not be to eliminate the (necessary – see the next paragraph) tensions between the public sphere and the state, but rather that the tensions can be played out in a more creative way, allowing social learning to occur. Let us turn our attention now towards Eckersley’s second challenge, namely how democratic decision making might tend towards moral or ethical reasoning rather than mere pragmatic considerations. The problem here is how the values of unconstrained dialogue, inclusiveness, and social learning (Section 3.4), developed in the philosophical laboratory of deliberative democratic theorists, might be upheld in real political decisionmaking settings. Again, Eckersley accuses Habermas of being to complacent in his answer to the problem. Indeed, Habermas has over the years worked out his discourse principle in more flexible terms in order to confront empirical political realities, as he states that the negotiation of legal norms may accommodate moral-political, ethical-existential and pragmatic modes of argumentation271. Whereas moral deliberation requires a universal perspective (freed from personal, egocentric, ethnocentric, etc. perspectives) and regards questions of justice (the search for a generalisable interest), ethical deliberation is oriented towards the principles and common values of particular forms of life or communities (the search for a shared value). Pragmatic arguments, in contrast, take as their starting point the preferences and goals of particular agents (hence, the discourse principle is now also able to accommodate the ‘aggregative’ view on governance for sustainability discussed under section 3.2) which serve as ‘limiting conditions’ for other participants in the discourse. Now, the point is that it is not clear how Habermas’s solution might work in the context of the often intractable controversies on technological developments involving risks to health and/or the environment. Reverting to a ‘moral’ solution in terms of a ‘generalisable interest’ will fail because in the case of ‘radical uncertainty’ (involving a debate on both ‘facts’ and ‘values’) there is no way of establishing with mutually acceptable 271

Habermas’s point in making these analytic distinctions is to point out that the practical test for political validity is not as stringent as the moral test for moral validity: while moral norms are based on a rationallyshared consensus (according to a universal standard), legal norms are of a more historically concrete character and need to be based on a rationally motivated agreement. However, moral discourse is still bearing indirectly upon the political process through the procedures that regulate it. Thus, Habermas is adamant in upholding the primacy of the moral point of view, since for him, a stable and well functioning democracy requires the creation of a polity integrated through rational insight into legitimacy.

164

Chapter 2

approximation what the outcomes of a choice will be – thus preventing a justification in terms of its being in the interest of everyone affected (even if all the participants are sincerely committed to finding a shared justification for a course of action) (Pellizzoni 2003, p. 344) 272. ‘Ethical’ solutions (in Habermas’s terms) turn on the revelation, through critical reflection, of the consonances in a deeper form of life that can bridge differences of opinion. Unlike a moral consensus, such an ethical consensus cannot be entirely rationally motivated, as participants in an ethical dialogue cannot work themselves entirely free from their boundedness to a particular form of life. An ethical consensus can however only be found within a particular community or form of life; between different communities holding different visions of the good life, what remains for Habermas is a ‘dialogue’ of reciprocally deaf voices (for a similar argument, see Bohman and Rehg 2002). This is particularly relevant for our purposes, since research on technological controversies strongly indicates that such controversies can only be understood fully if one also takes into account the clash of different ‘cultural repertoires’ that try to make sense of the technology in question. According to Rip (1999, pp. 110-114), who gives an overview of research findings in the issue, it is useless and furthermore counterproductive to try and do away with these repertoires or ‘cultural myths’, because they fulfil and essential function of uncertainty reduction in society273. Thus, we are left with Habermas’s third alternative –

272 Moreover, as De Dijn (2003) argues, seeking for a strictly ‘rational’ justification for a moral course of action (a ‘view from nowhere’, which, according to De Dijn, is an impossible endeavour) can turn out to be counterproductive, since it might disrupt any cooperation actually occurring between people based on a not entirely reflected ‘intuitive’ shared form of life. Similarly, Boltanski and Thévenot (1991) have developed their ‘commonwealth model’ on the premise that there is no encompassing rational point of view dictating to people which mode of coordination should be chosen under which circumstances. 273 In view of its importance to our subject, we will paraphrase Rip’s overview here. According to Rip, when new technological developments present themselves, at first, there is recognition of novelty, and there are attempts to ‘name’ it. Labels, like ‘the atom’, around which actors can assemble and with which they can link up, are the main route through which new technology acquires a social and cultural presence. Such labelling occurs at an early stage, before there is much experience with the new technology. The process of naming sets the scene, creates associations, and shapes learning about the new technology. Obviously, different interests are involved (even within the group of those who introduce a new technology there will be such differences). But one also sees a particular pattern emerge: a difference between ‘insiders’ and ‘outsiders’. The people who introduce the new technology see themselves as insiders who know much more about the technology and therefore position themselves as also more knowledgeable about its potential embedding in society. At first, actors not involved in the new technology need not consider themselves excluded, or as being outsiders – but insiders will nevertheless define them as such. The combination of labelling and diffuse group formation leads to situations where stereotyping and inclusion/exclusion behaviour becomes self-reinforcing. Such situations can give rise to a further, and reflexive, type of labelling, that of ‘proponents’ and ‘opponents’ of new technologies. As Rip maintains, neither ‘proponents’ nor ‘opponents’ are simple categories, but to think in those terms seems natural. The opponent/proponent dichotomy has become a cultural pattern (reinforced by the media), and it serves actors in their attempts to order a complex environment. Other patterns occur, because concerns about one technology, and promises about the same or another technology, are not separate issues anymore. They are connected, for example through shared labels of ‘risk’ and ‘promise’ of technology. Promises and risks of a new technology can be contrasted, and are, in fact, pitched against one another. According to Rip, in some cases, weighing promises against risks leads to mutual articulation and a better understanding of the value of a new technology. Over time, repertoires of promise and risk have emerged which allow such articulation of the value of a technology, without necessarily producing a consensus. In particular, a socio-cultural pattern for addressing novel technologies has become established since the ‘70s. Rip identifies this as the issue of ‘control’, and partly the specific responses of the technology promoters to the possibility of control. The new risk repertoire is said to be rooted in the novelty of the technology (and thus unpredictability of the dangers), and in the irreversibility and macro-character of the effects once they occurred (and thus in their essential ‘unmanageability’) – nuclear energy of course being the topos par excellence.


165

pragmatic compromises – but this is hardly likely to be a satisfactory solution, since in ethical matters, one cannot bargain or make concessions easily without compromising oneself and one’s integrity, and thus it is likely that the policy response to these difficult ‘taboo trade-offs’ will be of the kind discussed under section 3.2.2.1 (procrastination, buckpassing, etc.). Furthermore, if it is accepted that technological controversies imply a multiplicity of genres of speech and argument (traced back to the different ‘cultural repertoires’), then one might also challenge the implicit normative presumption of a shared commitment towards mutual understanding. Thus, caught between the Scylla of universal morality and the Charybdis of pragmatic compromise, deliberative governance seems fundamentally impotent when confronted with the realities of controversies on technological developments274. Should we, after all the hard work done in the previous sections, leave it at this rather gloomy conclusion? Fortunately for us, the answer to this (we admit, rather rhetorical) question is ‘no’. Recent work on deliberative democracy in general (and deliberative technology and/or risk governance in particular) can come to our aid, as it has been increasingly preoccupied precisely with conditions of often intractable disagreement. The resulting new insights are all more or less predicated on, at the one hand, a recognition of pluralism, and the ineradicable dimension of antagonism this entails, should be taken seriously275 (rather than being ‘massaged away’), and on the other hand, coming to terms

Furthermore, Rip identifies another (less coarse) pattern of ‘rules, routines and regulations’ in order to reduce uncertainty about new technology, as well as the effect these rules have on the cultural shaping of daily life. He asserts that rules are not only justified in terms of scientific knowledge (e.g. acceptable daily intakes of pesticides) or rational moral justification, but also in terms of effectiveness through relevant cultural transformation (e.g. the habit of washing fruits and vegetables). Hooking up with the cultural dimension of daily life thus might be an important element in the reduction of overall uncertainty or ‘intractability’ surrounding a certain technological development. Rip speaks of the ‘domestication’ of technology. He sees some possibilities for genetic engineering (e.g. labelling of GMO’s promotes the freedom of choice consumers highly value, possibility to connect to ‘myths of purity’ for food, etc.), but not for nuclear power: there will always remain a ‘yawning chasm’ (p. 114) between the practices of those involved in managing this technology on a daily basis, and the individual electricity-consuming ‘concerned’ citizen (electricity being an ‘invisible’ – in cultural terms – element of daily life). These cultural practices then tend to reinforce the way society handles ‘dangerous’ technologies in general: there is a diffuse political support for stringent measures (even if there is ‘no danger’ for a particular chemical). This pattern tends to reinforce another structural pattern of contemporary society, that of a separation between ‘promotion’ and ‘regulation’ of technology. In sum, Rip stresses the importance of agonistic interaction between different ‘cultural repertoires’ as part and parcel of societal reduction of uncertainty on new technological developments. This is achieved through the adoption of dominant cultural patterns: ‘proponents vs. opponents’, ‘risk vs. benefit’, and ‘promotion vs. control’. Controversies die down, not only through scientific analysis (although this might be a part of the solution), but through evolving practices and domestication of new technology. However, closure of the debate is always provisional and can be challenged through new scientific discoveries, new developments in related domains, etc. 274 Mouffe (1999, p. 754) summarises the problem of deliberative (liberal) democrats as follows: “…In a very systematic fashion liberal thought evades or ignores state and politics and moves instead in a typical always recurring polarity of two heterogeneous spheres, namely ethics and economics…”. 275 Again, on a theoretical level, this is where constructivists and Habermasians disagree. Rather than looking for a pure rational agreement where every trace of power would be effectively erased, constructivists accept that power (in its many forms, including e.g. persuasion, rhetoric, etc.) forms a constitutive ingredient of human interaction, communication and identity (rather than an impediment). In Latour’s somewhat idiosyncratic terms, ‘actants’ (humans and non-humans taken together) are always implied in uncertain forms of association – it is a priori unclear who is doing the ‘acting’ (i.e. exerting power).

166

Chapter 2

with the dimension of politics, which cannot be simply reduced to a choice between either pragmatic compromise or moral consensus276. To get out of the impasse, the crucial opening move offered by various authors (Mouffe 1999; Rip 1999; Latour 2004a) is to redefine the goal of democratic politics as a transformation of ‘antagonistic’ interactions into ‘agonistic’ ones. In the words of Mouffe (1999, p. 757), antagonism is a struggle between enemies, whereas agonism implies a struggle between adversaries – the difference being that an adversary is recognised as a legitimate enemy, i.e. someone’s ideas you combat but whose right to defend those ideas is not put into question. In the same vein (albeit in a different context of the study of socio-technological controversies), Rip (1999) points out the dangers of antagonistic struggles leading to an impasse where parties regard each other as enemies which should be destroyed (that is of course in a symbolic way – e.g. by mutually labelling the other as ‘contemptibly wrong’, ‘irrational’, ‘unethical’, ‘merely self-interested’, etc.). Antagonistic interactions do not spring into being overnight, but rather result from the cumulative effect of many individual events and experiences, as shown in our historic reconstruction of the nuclear power controversy in Belgium (Laes et al. 2004c). Under such conditions, the possibilities for a constructive interplay between perspectives are practically zero. To be sure then, developing governance structures on the basis of agonistic interaction most likely is a very challenging endeavour which at the very least requires a certain attitude of the interacting parties (Mouffe (1999, p. 756) even speaks of a ‘conversion’). For instance, they have to accept that they are competing for primacy within the same universe of discourse with others who cannot be a priori branded as unreasonable – implying an awareness of the naivety of dogmatic beliefs, a recognition of the fallibility of the one’s own perspective, and leaving room for ‘reasonable dissent’ (Keulartz et al. 2004, p. 23). Of course, one cannot sit back and merely wait for the actors involved in a particular policy domain to show enough goodwill and patience to sit back and listen to each other. What seems to emerge from the more ‘empirical turn’ in deliberative democracy is that such attitude has to be actively cultivated and even imposed in carefully designed settings. For instance, in a thoughtful analysis of the potential for deliberation and social learning in the context of sustainable agriculture, Vandenabeele (1999, pp. 79-80) points out the crucial role played by independent process supervisors in assuring the quality of ongoing dialogues. Success also seems to depend on the participation of actors to the dialogue who are able to bridge different perspectives. Existing situations involving considerable differences in political (and cultural) power also seem less conducive towards qualitative dialogues. And so on; in short, ‘controlling’ the dynamics of the ‘sway of power’ (whether it be economic, cultural, political, technological, etc.) is certainly not something that can be

276

The point is here quite simply that the conceptual instruments offered by Habermas are not ‘fine-tuned’ enough to register what is going on in political processes. Although Habermas distinguishes between different types of practical reasoning, he makes no distinction between different types of pragmatic compromises, nor does he offer any criteria for distinguishing between more or less acceptable compromises.


167

managed merely by designing good procedures277. However, design principles such as the precautionary principle at least offer the perspective of making the context more conducive towards the enactment of the agonistic interaction between different perspectives we envisage (cf. Chapter 5). The good news is that, even setting aside unrealistic expectations of unconstrained dialogue and enlarged thinking, some measure of social learning can be expected to occur in particularly designed settings that manage to keep the conflict within certain boundaries278. Moreover, as Meadowcroft (1997) and Rip (1999) argue, there is an intimate link between conflict and learning – the reason being that social learning takes an effort, and individuals and groups will generally try to avoid heterogeneous interactions with other groups which will potentially challenge them. Thus, for social learning to occur, there has to be a ‘forceful focus’ (i.e. a government declaration that it is planning to introduce new regulations in a particular sector) to put the actors in motion. In any case, even with procedural safeguards in place possibly leading to some measure of social learning, one should not expect consensus, or even genuine understanding of each and every perspective among the involved actors. One cannot simply do away with the dimension of ‘politics’ so pithily described by Barber (1984, p. 121)279: …To be political is to have to choose—and, what is worse, to have to choose under the worst possible circumstances, when the grounds of choice are not given a priori or by fiat or by pure knowledge (episteme). To be political is thus to be free with a vengeance—to be free in the unwelcome sense of being without guiding standards or determining norms yet under an ineluctable pressure to act, and to act with deliberation and responsibility as well…

The essential contribution of ‘politics’ (not available to ‘economics’, ‘ethics’ or ‘morality’) to the ‘construction of the common world’ lies in the fact that only politics is able to change the identities of entities demanding to be heard (rather than simply representing the interests of pre-constituted entities as is done by economics – Latour (2004a) calls this the ‘possibility of betrayal’), and only politics is able to institute what Mouffe (1999) calls a ‘hegemonic order’. A hegemonic order should be understood as an ensemble of practices, discourses and institutions aiming at the creation of a unity in the context of conflict and diversity, which can only be obtained through the creation of an ‘us’ by an expulsion of ‘them’. In other words, hegemony is always composed of an often uneasy mixture of

277

This would certainly be a rather foolhardy view in the context of Belgian energy policy, characterised by a large disparity between the different actors involved in the opportunities to influence decision making. For an ‘insider view’ (the author of the book, Luc Barbé, was the chief of cabinet of the former state secretary of energy and sustainable development (1999-2003)) on the close relations between the political sphere and private firms in the context of Belgian energy policy, see Barbé (2005). 278 Bennett and Howlett (1992), in an overview of the literature on policy learning perspectives, distinguish between three forms of learning: ‘government learning’, which involves state officials learning about processes and resulting in organisational change; ‘lesson drawing’, which involves policy networks learning about particular policy instruments and resulting in a change in the policy programme; and ‘social learning’ (used here in a more restricted sense), which involves policy communities learning about ideas and core values and resulting in a paradigm shift. However, they conclude that learning seldom alters the core values of a particular group. 279 ‘Politics’ here should not be understood as an activity undertaken by professional politicians; any ‘profession’ (scientists, philosophers, economists, etc.) can engage in politics.

168

Chapter 2

‘rationality’, ‘legitimacy’ and ‘power’ (and everything in between)280. Once this is accepted, the main question becomes not how to eliminate power but rather how to constitute forms of power more compatible with the requirements of sustainable development. Left to itself, politics is tempted into naturalising its frontiers by the creation of essentialised identities. However, if we want to actively produce ‘voices’, allow enough room for social identification and stimulate social learning, then this tendency should be held in check by the requirement that no ‘enemies’ are formed but rather ‘adversaries’ – the hegemony should always be provisional. Whereas a decision would simply be impossible without some form of exclusion, the traces of the act of exclusion should be made visible as far as possible. 3.4.3

Conclusion

The preceding sections have taught us the great value of the contribution of deliberative democratic theories to governance for sustainability. These theories – and in particular the discourse ethic as developed by Habermas – have constantly sought to improve the rationality of public debate and decision making (and continue to do so). They constantly keep us on our toes to look out for ‘voiceless entities’ (though this required a slight modification of the Habermasian perspective in order to include the ‘non-human others’) and conditions of ‘unrepresentative representation’. As a device for exposing such conditions, one can think of stakeholder analysis (who stands to gain or loose from a decision and should consequently have a say in the debate?) and institutional analysis (to what extent is the institutional setting of the debate in line with the ideal speech situation as characterised by Habermas?). However else one wishes to defend deliberative democracy, we take the argument of its use as a ‘critical vantage point’ to constitute its unimpeachable core. However, from a constructive point of view – i.e. enlisting theories of deliberative democracy for a design of procedures that might better resemble the ideal communicative settings – the perspective offered by Habermas shows its limitations. Largely absent from this perspective is a more active view of participatory involvement as the basis of common problem-solving, which also offers possibilities for identification and a more ‘passionate’ commitment to democratic values (rather than a rational consensus). This implies a much stronger institutional support for participatory processes, and a corresponding investment in training, administrative support, communication, etc. We have suggested (though not yet worked out) how the precautionary principle might be enlisted for this purpose. A second deficiency of the Habermasian perspective lies in its inability to deal with the truly political dimension of governance and the ineradicable conditions of conflict, tension, contrast, etc. Here we have argued that the solution does not lie in seeking for an (unattainable) consensus, but rather in guaranteeing that, despite the conflictual character of interaction, some measure of learning will still be possible. ‘Antagonism’ should be transformed into

280

Thus, the core of our argument comes down to the observation that Habermas’s conceptual apparatus is not suited for an adequate description of this mixture, much in the same way as an engineer would not turn towards the fundamentals of chemical kinetics in order to build a chemical reactor.


169

‘agonism’. This in turn will require careful procedural design – a task we set out for ourselves in chapter 5.

4 Summary and conclusions In conclusion to this chapter, it might be useful to remind the reader of our initial point of departure. In our introduction, it was set out that we would study governance mechanisms as ways to attenuate and/or reduce the impact of the often inevitable uncertainties on the policy choices to be made in view of sustainable development. Thus, in retrospect, it becomes clear that all mechanisms discussed each present a specific possibility for achieving tractability, ranging from the fact/value separation in the expert-based governance scheme up to the potential of broadly conceived ‘risk repertoires’ as a general societal mechanism for uncertainty reduction. As becomes clear from the previous sections, tractability is achieved not only by defining problems in a specific way, but also by assigning social roles to actors. Expert-based governance (Section 3.1) operates according to a (deep-seated) logic that political authority and decision making can be separated from the scientific authority provided by experts, based on disciplinary competence. It is an efficient scheme for arriving at solutions when ‘facts’ are generally uncontested. However, difficulties arise when value choices are (often tacitly) delegated to expert committees, or when abstract scientific insights (and the models of human behaviour they often incorporate) prove to be ineffective when applied to multi-dimensionality of practical contexts. Governance by aggregation (Section 3.2) seeks a resolution of policy problems through negotiations between different ‘interests’, but is mostly fraught with difficulties when applied beyond a short-term perspective. Governance by pacification (Section 3.3) represents a widespread political strategy to tackle (politically organised) value positions of a different kind, and has shown its value in the past, also in Belgian energy policy. Nevertheless, be it from a moral (e.g. pacifying relations between social groups through more or less ‘distorted’ communicative settings), an empirical (e.g. a possible lack of sufficient learning capacity in the context of environmental or technological controversies) or a pragmatic (e.g. a possible lack of real influence on decision making) point of view, one can only be sceptical about the potential of such advisory councils for pushing towards a change in the direction of sustainable development – at least when they are not connected to a wider participatory practice. A last scheme – deliberative governance (Section 3.4) – is strong in its insistence on giving a ‘voice’ to the voiceless and promoting the values of ‘unconstrained dialogue’ and ‘social learning’. However, this scheme is deficient in the sense that it attempts to ground the search for solutions to governance problems in terms of generalising and abstracting processes, which tend to deny the true nature of a political conflict, characterised by ineradicable conditions of tension, conflict and contrast. What needs to be done in this final section is to draw together the understanding of the different governance schemes, and consider possibilities of turning such insights into suggestions, advice or even guidelines how to do things intentionally. Of necessity, we

170

Chapter 2

will remain rather global here, but further specification to concrete approaches will be made in chapter 5. Perhaps in spite of the critiques raised on the different governance approaches, several promising features seem to arise from the analysis of expert-based, aggregative, pacifying and deliberative approaches to governance alike. A clear first lesson seems to be that, when setting up governance structures for sustainability, one should give attention to both the ‘intentional’ and ‘emergent’ design of interaction procedures and practices. These are two sides of the same coin: it would be no good to develop an ideal procedure which simply would not work in ‘the world out there’ because it leaves no room for identification or mobilising collective passions. A problem that has become ‘tractable’ within the walls of a consultative body needs to address also the ‘intractability’ in society at large, and this means taking into account the mobilising force of ‘risk repertoires’ (even if they could seem biased at first sight). Thus, it would seem that there are good reasons to prefer a two-tiered governance structure for sustainable (energy) development. A first tier would then consist of a formal deliberative body located in the ‘political system’ (consisting of different groups which are sufficiently representative of the different discourses on the problem at hand). The necessity of a ‘central body’, consisting of a rather small number of representatives, operating in a somewhat protected setting is predicated on the grounds of the demanding requirements for social learning to occur, whilst keeping conflicts within bounds. But, as discussed in section 3.3, the mere formal provision of such a deliberative mechanism is unlikely to initiate any substantial changes by itself. Two further requirements thus seem absolutely necessary: firstly, the ongoing work in the deliberative body has to have a connection to formal government initiatives (thus creating a ‘forceful focus’ setting the social groups represented in motion). Central government will still have to assume an orienting role, e.g. by elaborating national energy policy strategies, establishing targets and frameworks, and mediating the interface between national performance and international agreements (e.g. the rise of the precautionary principle as an international norm for decision making) (Meadowcroft 1997, p. 451). Secondly, the deliberative body, through its work, needs to be connected with ongoing local initiatives; by building upon emerging patterns of local interaction in view of sustainability, it could be possible to establish a web of overlapping groups (some of them focused on a particular regional environmental problem, or on a particular issue (e.g. local initiatives on low-level radioactive waste management) or technology, or on a particular siting discussion (e.g. wind farms), etc.) which can explore the practical implications of sustainable development in the various dimensions of actual social life. In any case, discussions of planning processes and outcomes will need to be iterative and ongoing – no single centre could in any case possibly ‘plan’ all the necessary adjustments sustainable development will entail. Within each tier then, it is clear that a lot of attention should go to interaction processes and quality control (hence, there is a need for a neutral ‘guardian of the process’). Within the deliberative body, a first concern should be to look out for a ‘production of voices’ –


171

i.e. by actively seeking out previously excluded perspectives on the problem at hand (hence, the importance of a good connection with the debates going on in the public sphere). Next, the deliberative body should address the question how these perspectives could be represented most effectively – i.e. by convoking a ‘jury’ most capable of speaking in the place of the relevant perspectives (work done by theorists of deliberative democracy will be very helpful here). And finally, the deliberative body also needs to come to a ‘closure’ of the debates, but this ‘closure’ can only be considered after the previous two demands of problem structuring are met (and consensus should generally not be expected). Conditions of an a priori closure (e.g. in terms of included points of view, ‘facts’ rather than ‘values’, ‘experts’ rather than ‘laymen’, etc.) should be avoided as far as possible – on the contrary, the processes adopted should make the ‘construction work’ done by experts in order to arrive at an advice more visible to all the actors involved. However, as we have discussed, ‘closure’ should always be provisional – hence again the importance of the twotiered structure, allowing residual problems to be taken up at the lower tier for further articulation. In any case, the goal should not be to eliminate the unavoidable (and necessary !) tension between the public sphere and the state apparatus. Rather, we seek to explore how the tension might be played out in more ‘productive’ ways, particularly for those ‘voices’ (e.g. ‘non-human others’, future generations) that have historically been rather marginalised in processes of governance. It is clear that this is a very demanding task, which will be taken up in chapter 5. The next chapters (Chapter 3 and 4) will however be concerned with the reconstruction of actual policy processes and practices, in order to provide more practical insight in the perhaps sometimes abstract reasoning developed here.

CHAPTER 3 THE EXTERNE METHODOLOGY AS A DECISION SUPPORT FOR SUSTAINABLE ENERGY POLICY This chapter will be devoted to a first practical case-study of decision support for sustainable energy policy, as we will be concerned with an investigation of the merits and drawbacks of external cost calculations. Such calculations are often presented to policy makers as ‘objective’ tools for the comparative assessment of different electricity production chains in view of sustainable energy provision, and their use is often strongly advocated on this basis. In particular, we will focus on the ‘ExternE’ methodology developed under a large-scale research project funded by the EC. Building on our discussion on the merits and drawbacks of the model of governance by aggregation in the previous chapter, our main goal in this chapter is to find the ‘right’ application of external cost calculations in the policy development cycle. Rather than highlighting the limitations and weaknesses of the concept of externalities with a view to attacking the general limitations and weaknesses of economic theory in general, we take a look at the actual numbers coming out of the ExternE work (Section 3), and try to put them in an applicationoriented context (Section 5). Our analysis will be guided by a critical review of some of the commentaries on the ExternE methodology (and external cost calculations in general) as found in the literature. The external costs of the nuclear fuel cycle will of course retain our attention somewhat longer (Section 4). In the course of this chapter, we focus especially on two interrelated questions: a) the wide disparity in external damage costs reported in previous studies, and the extent to which this represents a problem; and b) the usefulness of previous valuation efforts for policy purposes. One of the main theses of this chapter will be the distinction we introduce between a broad cost-benefit analysis aimed at setting up a political framework that distributes rights and responsibilities among the involved stakeholders, and a more narrow form of quantitative cost-benefit analysis aimed at choosing among a range of possible policy measures for the implementation of the broad political framework 281 . Thus, our analysis aims to show that although quantitative methods can never carry the full weight of a justified policy decision at the level of institutional support or reasonable choice, it is nevertheless practically required to quantify benefits and risks of proposed policy measures for carrying out a technical verification 282 . Furthermore, we maintain that if economic quantification is attempted at this level, then this quantification should be carried out in full. Therefore, our point of view makes a fundamental distinction between the choice to apply a method and the methodological choices made within the framework of the chosen

281The

difference will become more clear in Chapter 5, where we will set out a new proposal for a sustainable governance structure. 282 The notions of ‘institutional support’, ‘reasonable choice’ and ‘technical verification’ refer to the different levels of political judgment expounded in chapter 2 – section 2.3.

174

Chapter 3

method. Our recommendations on the use of external cost calculations are thus clearly of a limited, conditional, and pragmatic-strategic order (Section 6). Various objections to quantification are best interpreted and addressed not as obstacles to numerical representation of ‘qualitative’ parameters, but rather as pleas for more reasonable uses of such representations.

1 Introduction While the concept of ‘externalities’ is central both to markets and economics 283 , traditionally only few studies have considered the evaluation of externalities associated with energy use 284 . Throughout the ‘70s and ‘80s, the dominant analytical approach to the environmental appraisal of electricity supply options was provided by comparative risk assessment 285 . However, the end of the ‘80s saw the rise of neo-classical environmental economics research on the external costs of electricity provision, with major studies being commissioned by influential bodies such as the EC, the US Department of Energy (USDOE), the German electricity industry, and numerous other industry, state, national and non-governmental bodies 286 . The relevance of recognising, assessing and internalising external costs was quickly enlisted in the emerging political discourse on sustainability, for instance in the EC’s Fifth Environmental Action Programme (EC 1993), which required the integration of the environmental dimension in other policy areas. One of the key elements of this programme was ‘to get the prices right’ and to ensure that environmental externalities were accounted for in market mechanisms. Internalisation of external costs and benefits was highlighted again in the EC’s White Paper on energy policy (EC 1995b). This expressed political need seems to be predicated also on a desire to transform the ‘fuzzy’ concept of sustainability into an objectively measurable quantity (Stirling 1999b) 287 .

283

According to NEA (2003, p. 15), the concept of externalities has been referred to in the economic literature since the beginning of the 20th century, most notably by the economist Pigou. At the turn of the previous century, Pigou launched the visionary proposal to tax pollution as a means to combat the London smog. At the time Pigou’s proposal was regarded as an academic curiosity, but several decades later it was rejuvenated as the core of the ‘Polluter Pays Principle’. Starting from the early fifties, externalities were explicitly related to the detrimental impact of economic activities on the environment. Political recognition of the existence of externalities however remained forthcoming until the early ‘70s with the rise of the above-mentioned ‘Polluter Pays Principle’, defined as a means to allocate costs of pollution prevention and control measures to polluters, and thereby to consumers of their products, rather than to society as a whole. 284 In the Belgian context, one of the earliest efforts at economic quantification of the impacts of different options for electricity generation (a comparison of coal vs. nuclear power) was furnished by Hecq and Vouche (1984). However, at that time many major impact categories could not be quantified easily; to fill in the gaps, rules of thumb, simplifications and/or expert judgments were used rather extensively – a procedure which gave rise to criticisms (e.g. by Erreygers 1984). 285 For an overview of such studies, see Stirling (1997). 286 For an overview of such reports, see OTA (1994). 287 Schrader-Frechette (1985) refers to this tendency for policy makers to search for technical solutions to essentially political problems as ‘the fallacy of unfinished business’ (i.e. the expectation that, in the future, some technical solution will become available for the present ‘subjective’ choices).

The ExternE methodology as a decision support for sustainable energy policy

175

As there is quite some literature which critically reviews the concept of (energy-related) externalities and the general usefulness of basic framing assumptions on a conceptual level (see e.g. Schrader-Frechette 1985; Stirling 1997) – and indeed, much of our discussion of the governance by aggregation model is concerned with such a conceptual investigation (cf. Chapter 2 – Section 3.2) – the present chapter does not attempt to add much further to this theoretical discussion. Neither will we engage in a detailed methodological discussion – for this, we refer the interested reader to the methodological volumes of the ExternE project (EC 1995a – Vol. 1; EC 1999a – Vol. 7) which set these out in great detail. Rather than highlighting the limitations and weaknesses of the concept of externalities with a view to attacking the general limitations and weaknesses of economic theory in general, we take a look at the actual numbers coming out of the ExternE work (Section 3), and try to put them in an application oriented context (Section 5). Our analysis will be guided by a critical review of some of the commentaries on the ExternE methodology (and external cost calculations in general) as found in the literature. The external costs of the nuclear fuel cycle will of course retain our attention somewhat longer (Section 4). But first, we will put on our economist’s hat in order to briefly address the way in which externalities are defined, together with the practical and theoretical implications (Section 2).

2 Theoretical background

2.1

Definitions

Within neo-classical economics, the theoretical basis for including external costs in economic analyses is well understood. Neo-classical economics analyses interactions between productive activities and the preferences of individual buyers constrained by a feasible range of choices and income. In other words, ‘traditional’ neo-classical economics is concerned with the investigation of price determination and market structures subject to both a formal constraint (i.e. an analysis of human interaction characterised by a competition between different human desires, limitation of the actual goods and labour forces which are useful for the satisfaction of desires, a coexistence of people afflicted by the same or similar desires competing with each other for the means of their satisfaction – cf. Chapter 2 – Section 3.2.2) and a domain constraint (i.e. an analysis of human interactions that occur in an institutional context that is deliberately created or used for economic ends – ‘the market’). Environmental economics proposes to expand the traditional domain of neo-classical economics to the analysis of ecological problems. This expansion in turn rests upon the central concept of ‘externalities’. Put briefly, an externality is “…a service or disservice rendered to persons other than the contracting parties…” (Pigou 1952; quoted in Dietz et al. 1994, p. 44). This concept is perhaps most easily explained by an example. A coal-fired power station produces (besides the desired electricity) harmful fumes. The fumes spread out over the countryside, representing a threat to (amongst other things) livestock and crop farmers. In order to combat or eliminate the noxious effects of these fumes (which might for example cause reduced crop yields),

176

Chapter 3

the affected farmers must make certain investments. In the absence of any incentives to do so (e.g. legislation setting emission standards), the company operating the coal-fired power plant however will not take into account in its economic calculations the costs that it is imposing upon other agents (e.g. the neighbouring farmers), who thus remain ‘external’ to the sphere of economic relationships in which the company itself operates 288 . The farmers’ interests are thus compromised in the sense that they are unable to assert their preferences since they must make investments for which they cannot negotiate any compensation. In this example, the externality imposed upon the farmers is thus negative; but externalities can also be positive, for instance if we assume that the reliance on coal for electricity production makes the national economy less vulnerable to potential future price shocks on the international oil or gas markets. However limited in nature, we believe the above discussion of externalities serves to clarify a few essential characteristics of the concept. On a conceptual level, the analysis of externalities has been extended by constructivist scholars (e.g. Callon 1998; Latour 2004a) beyond the formal domain of economic interactions. According to this constructivist reading, externalities arise through the systematic (intentional or unintentional) negation of the ‘costs’ (used in a more figurative sense) borne by other agents required to make possible an interaction between people which subsequently appears as ‘naturally given’ (Callon 1998). Since externalities are manifestations of (physical, social, cultural, etc.) interdependence, and since it is simply impossible for actors engaged in a certain situation to foresee the full consequences of such interaction, externalities will surface in all types of interaction 289 . In turn, the existence and magnitude of externalities necessarily depends on a particular framing of the interaction under consideration (e.g. the different ‘dominant cooperative schemes’ discussed in chapter 2 provide examples of ‘framing’). This framing does not abolish all links with the ‘outside world’, but rather puts it between brackets. As a result of this ‘bracketing’, inevitably, an ‘exterior’ is created. Now, according to Callon (1998, p. 250), it is possible to respond to this concept of framing (essential to the understanding of any kind of interaction), by adopting one of two diametrically opposed attitudes: the first one emphasizes the closure of interactions and the mutual agreement of the actors engaged in this closed situation; the second one reacts by highlighting the possible links and ‘forgotten’ connections with the ‘outside world’. A parallel can be drawn here with the concepts of ‘agonistic’ and ‘antagonistic’ interactions which proved to be central to our discussion of the requirements for governance structures for sustainable development in chapter 2: the first point of view corresponds to an antagonistic attitude – it creates ‘enemies’ which it tends to suppress with physical or symbolic devices; while the second one tries to transform the antagonistic interaction in an agonistic one – it tries to make visible and measure (up to a certain extent) what has been ‘forgotten’ or expelled to the ‘exterior’.

288

‘Internal’ costs include e.g. the construction cost of the power plant, the fuel costs, operating and maintenance costs, dismantling costs (that is, if fully internalised – cf. Section 4.2), distribution costs of electricity, etc. 289 Therefore, in itself, the existence of externalities is not morally outrageous, nor can it be expected that externalities will ever be fully internalised.


177

This last sentence already points at the crucial contribution of economics in the creation of the agonistic (instead of antagonistic) interactions we envisage. Economics as a scientific discipline has precisely been obstinately concerned with finding ways of making interactions calculable, and with thinking up devices that encourage market conditions to emerge 290 . Out of a professional concern for collective efficiency and the optimisation of resource allocation, neo-classical economists have traditionally been on the look-out for ways to extend or restore the logic of the ‘pure market commonwealth’ in the case of ‘market failures’. Market failures arise when, in terms of efficiency of resource allocation or in terms of the provision of socially desirable goods, the best result (in an economic sense) that could have been obtained was not obtained in practice. The concept of ‘externalities’ is intimately related to this general category of ‘market failures’: externalities – whether positive or negative – render the market (at least partially) inefficient, because they are responsible for a gap between private marginal income and marginal social costs. It is important to note that externalities (in the economic sense) are legally legitimate interferences imposed upon others: they arise out of a particular institutional arrangement of property rights. If we take the example of a perfectly competitive electricity market, according to the standard economic hypotheses, the electricity price (and the total amount of electricity produced) will be determined by the marginal cost of electricity production – i.e. the additional cost incurred by selling the last kWh of electricity. If externalities are present however, this private calculation – which, under a number of hypotheses (e.g. perfect information, people acting as rational utility maximisers, etc.), is supposed to guarantee a social optimum – is biased: it does not take into account the investments which the farmers in our example need to make in order to protect themselves and their crops from the effects of the harmful fumes. In this case, in the absence of proper incentives, the market is deficient as a vehicle for providing the optimal solution. This, in turn leads the neo-classical economist to recommend instruments (e.g. property rights, taxes, subsidies, etc.) that allow for the internalisation of the external costs 291 . At first sight, this definition of externalities seems clear and unproblematic enough. However, there are some major difficulties involved in identifying and assessing externalities, which we will discuss in the following section. As we did in the previous chapters, we will continue to address these issues from the point of view of constructivist

290

Economics as a discipline has shown a strong ‘imperialistic’ tendency towards other disciplines, e.g. in social sciences (‘rational choice theory’); political sciences (‘public choice theory’); legal theory (‘law and economics’), etc. 291 The difference between a broad and narrow definition of ‘externalities’ now also becomes clearer. Let us take an example provided by Deblonde (2001, p. 163). Suppose that the environmental damages caused by car traffic can be calculated in economic terms. Public authorities, according to economic theory, should then decide to levy a tax on private car use equivalent to the amount of pollution (in monetary terms) caused by this activity. But because of the non-elasticity of private car use and despite the taxes levied, people continue to drive their cars as before, producing nearly as much pollution. Now in economic terms, the externality no longer exists, since the costs of private car use have been internalised. In physical terms however, the amount of interference is practically unchanged; hence, the physical externality does not disappear.

178

Chapter 3

sociology. According to us, our particular vantage point has the advantage of offering a perspective from which the ‘framing’ actions by economists can be made visible. Hence, the constructivist point of view allows us to keep track of the ‘maps of the exterior’ produced by economists as well as the progressive deployment of instruments that make the different world states calculable. This seems important because economists, left to themselves or aided by policy makers, have an inclination towards the assumption that economic framing is the norm – in the double sense of something that is desirable and statistically predominant – and therefore that ‘leaks’ towards the exterior are merely accidental and subject to amendment by the use of (tried and tested) economic tools. Thus, a strong convergence of interest could develop between those seeking to promote a particular academic discipline (e.g. in need of large resource inputs in terms of money and manpower), and those seeking to promote or justify difficult or potentially unpopular decisions. In this regard, the siren song of the single result or ‘analytical fix’ (i.e. ‘energy option x has lower external costs than technology z, hence...’) may sound particularly tempting for policy makers wishing to advance particularly preferred political options 292 . Indeed, as Krewitt (2002, p. 840) proclaims …the ExternE label became a well recognised ‘brand’, the scientific quality of the work was well accepted on the international level, national and international organisations got used to referring to ExternE numbers as a standard source for external cost data, and also industry expressed increasing interest in ExternE results…

Whether or not this convergence between academic and industrial or political interests indeed takes place, the potentially intoxicating mixture of research interests and high decision stakes serves as a statutory warning for a ‘correct’ communication of research results towards broader, non-technical audiences (most notably policy makers) and implied use of the results.

2.2

Major difficulties

Despite being since long firmly established in economic theory, it seems difficult to establish an unequivocally agreed methodology for external cost calculations. The basic requirements are simple: in essence, for coordination by market mechanisms to occur, it is necessary for actors to 1. have preferences; 2. hierarchise and rank them (i.e. quantification is required); and 3. reveal and negotiate them (i.e. an actual market space has to exist).

292

In the foreword to a summary of the ExternE results destined to a broad audience (EC 2003), Philippe Busquin (the then EC commissioner responsible for research) claims that “…the ExternE results allowed different fuels and technologies for electricity and transport sectors to be compared. Policy actions could therefore be taken to tax the most damaging fuels and technologies (like coal and oil) or to encourage those with lower socio-environmental costs (such as renewables or nuclear)…”. However, as we will try to show in the present chapter, careful examination of the ExternE results suggests a much more cautious attitude towards the unequivocal encouragement of nuclear power…


179

Within the market sphere, this is relatively unproblematic: individuals have fairly clear information on which to base their decisions. Commodities sold and bought on markets tend to be visible, their characteristics are generally well-known, and they have a price. Expanding the domain of neo-classical economic theory to ecological or social phenomena therefore seems to be a completely different matter altogether. Application of economic analysis in domains where no market exists then requires the identification of ‘hypothetical markets’. In order to calculate the economic value of externalities on these hypothetical markets in turn implies that 1. the ‘leaks’ from commercial interactions are actually identified (i.e. identification of priority impacts, description of impact pathways); 2. the impacts of the ‘leaks’ are identified (i.e. quantification of burdens, description of the receiving environment, quantification of impacts on the receiving environment); and 3. the impacts are measured (i.e. economic valuation of impacts) (EC 1995a, Vol. 2). Each of these steps can provoke difficulties, as each of them can be tainted by the various degrees of uncertainty set out in chapter 2. In the following sections, we propose an analysis of the major difficulties encountered in this endeavour. This analysis proceeds along analogous lines as our more general discussion of the ‘governance by aggregation’ scheme in chapter 2 (Section 3.2.2). Firstly, we will investigate whether typical studies on externalities of electricity production give an adequate account of the identification of the ‘leaks’ and what they represent in terms of the ‘interests’ of the affected parties (Section 2.2.1). Secondly, we will look more specifically at some common problems of quantifying these impacts (Section 2.2.2). The following section will then go into the specifics of the ExternE methodology (Section 3). 2.2.1

Representing the interests of affected populations

Perhaps the most fundamental difficulty (shared with other forms of quantitative risk analysis) facing the various techniques of economic valuation of social and environmental costs is that in order for people to negotiate their interests, different impacts have to be made comparable on one scale for all relevant actors (i.e. those who are affected by the externalities). However, it is only very rarely the case that an individual technology is seen to present only one specific form of impact. Normally, the characterisation of risks associated with any individual technology requires the aggregation of a series of different magnitudes, each corresponding with a particular form of impact. The conventional analytical response to this breadth and diversity of issues in risk assessment implies adopting a single major yardstick of performance and seeking to measure all the various aspects of risk using this as a metric (e.g. external costs in environmental valuation; or the number of expected deaths, decrease in life expectancy, etc. in comparative risk assessment). Setting aside for a moment the methodological difficulties involved in deriving these indicators (some of which will be discussed in the next section), it is hoped that in this way the multiplicity of impact factors typically confronted in technological risk may usefully be reduced to a single key factor, thus significantly simplifying the process of

180

Chapter 3

The form of different aspects of sustainability Severity

Do the options differ in the ratios of risks of death to risks of injury or disease which they pose? How much illness or how many serious injuries equate in severity with one death?

Immediacy

Are the effects associated with the different options equally immediate in their manifestation or do they differ in the degree of latency between the initial commitment of a burden and the eventual realisation of an effect?

Gravity

Are the effects associated with some options dominated by low probabilities of large impacts, while those of other options are characterised predominantly as high probabilities with relatively low impacts? To what extent are impacts the result of single or repeated events?

Reversibility

Are the effects associated with different options all equally reversible after they have been committed?

The distribution of different aspects of sustainability Spatial distribution

Are the effects associated with different options identical in their spatial extents? Is it better that impacts of a given magnitude are geographically concentrated or dispersed?

Balance of benefits and

To what extent is the social distribution of the environmental burdens caused by

burdens

each option balanced by the distribution of any associated social or economic benefits?

Fairness

To what extent do the distributions of burdens imposed by the different options act to alleviate or compound pre-existing patterns of privilege or social disadvantage? To what extent should exposure to other (unrelated) factors be taken into account in the assessment of sustainability?

Public or worker

To what extent do different options impose different distributions of effects across

exposure

workers and the general public?

Intergenerational equity

Do the effects associated with certain options present risks to future generations to a degree not associated with others? How should these effects be balanced?

Human or non-human

Do the options differ in the degree to which their impacts affect the well-being of humans and non-human organisms?

The autonomy of those affected by different aspects of sustainability Voluntariness

Do the environmental effects of different options vary in the degree to which exposure may be considered to be ‘voluntary’ prior to the commitment of an impact?

Controllability

Once committed, are the impacts associated with different options all equally controllable from the point of view of the affected individuals or communities? Do certain effects require efforts at control which are perceived to pose a threat to democratic institutions or processes?

The ExternE methodology as a decision support for sustainable energy policy Familiarity

181

Do the effects associated with different options differ in terms of the degree to which they are familiar to individuals, communities and established social institutions? Do responses to the different effects involve equally disruptive changes to normal routines and attitudes?

Trust

Do options differ in terms of the degree of trust enjoyed in the wider society by the institutions and communities charged with evaluating and managing their associated impacts? Does the appraisal of certain options tend to be more a specialised undertaking than that of others?

Table 5. The multi-dimensional character of sustainability (Source: Stirling 1997 and 1999b, pp. 122-123)

appraisal. This approach however is far from unproblematic when used as a representation of the interests of populations afflicted by the risks in question. Drawing upon one of the longest-established fields for the comparative appraisal of technological risks – that concerned with energy technologies – Stirling (1997, 1999b) identifies a large number of such discrete dimensions, relevant for the appraisal of the sustainability of different energy production technologies (Table 5) 293 . The dimensions represented in Table 5 have been replicated across groups of lay people and experts judging large and diverse sets of hazards, and thus at first sight provide an opportunity to describe impacts of energy technologies as they are conceived by ‘the affected public’. As such listings seem to capture essential elements of understanding risk (and sustainability), any attempt to faithfully represent the ‘public interest’ should take these dimensions into account in some way 294 . However, this rather general statement 293

More aggregate sets, consisting of overarching higher-order factors representing a set of correlated dimensions, have been suggested (see e.g. NRC 1996). One such factor, labelled ‘dread risk’, groups dimensions of perceived lack of control, dread, catastrophic potential, fatal consequences, and the inequitable distribution of risks and benefits. Nuclear weapons and nuclear power usually score highest on the characteristics that make up this factor. A second factor, labelled ‘unknown risk’, is defined at its high end by hazards judged to be unobservable, unknown, new, and delayed in their manifestation of harm. Chemical industry and technologies involving interference in the genetic structure of beings score particularly high on this factor. A third factor, reflecting the number of people exposed to the risk, has been obtained in several studies. However, this more abstract ‘risk triptych’ is of less importance to us, since at the present moment we aim to be as specific as possible about relevant risk dimensions. 294 This seems a convenient moment to do away with one possible major misconception: we are not pitting an ‘expert understanding of risk’ (‘risk = probability x consequences’) against the ‘public risk perception’ (‘risk is perceived through multiple dimensions’). Psychological theories throw an interesting light on this presumed opposition. Such theories basically suggest that there are two fundamentally different ways in which human beings process information about the world when they make judgments or arrive at decisions (Epstein 1994, Sloman 1996). One processing system is older (in evolutionary terms), fast, mostly automatic, and hence not very accessible to conscious awareness and control. It works by way of similarity and associations, including emotions, often serving as an ‘early-warning system’. The other processing system works by algorithms and rules, including those specified by normative models of judgment and decision making (e.g. the probability calculus, formal logic), but is slower, effortful, and requires awareness and conscious control. For the rulebased system to operate, one needs to have learned the rule. The association/similarity-based processing system requires real world knowledge (i.e. more experienced/expert decision makers make better decisions using it than novices in a domain), but its basic mechanisms seem to be ‘hard wired’ (i.e. a result of the basic

182

Chapter 3

needs to be qualified in some important ways. Firstly, such listings are derived from a particular theoretical approach to risk. The risk dimensions mentioned in Table 5 are of a cognitive nature. They report how people regard and assess technological risk from a detached and isolated point of view. Typically, a large number of people are probed with questionnaires, asking many different questions about many different risks. The risk dimensions withheld for analysis are the result of reconstructions made by the researchers. This might be a drawback, as they do not necessarily represent how people perceive risks on an individual basis 295 . Social aspects of attitudes towards technological risks, such as the commitment of a certain person in a particular risk issue, are not taken into account, even though they are shown to be very significant (Sjöberg 2003). This cognitive aspect of the approach is a limitation, but not if the aim is merely to provide an indication of which kinds of interests could be advanced in the debate. Secondly, no attempt is made to provide any kind of ordering in these risk dimensions. The aim is simply to cover as much as possible the different possible perspectives on technological risks, thus suggesting some issues around which a public appraisal of technologies could evolve, without actually predicting the exact sequencing or force of the separate dimensions. In fact, only one dimension could already result in a substantial controversy. A significant limitation then arises from the fact that an attempt is made to identify ‘timeless’ dimensions, whilst social views are always culturally determined and subordinate to history and unpredictable future events.

structure of our neural networks). These two processing systems often work in parallel and, when they do, more often than not result in identical judgments and decisions. One becomes aware of their simultaneous presence and operation in those situations only where they produce different output. Experiential thinking is intuitive, automatic, and fast. Psychologists believe that the beneficial aspects of experience or associationbased processing in the context of risk have enabled us to survive during the long period of human evolution and remain the most natural and most common way to respond to threat, even in the modern world. It relies on images and associations, linked by experience to emotions and affect (feelings that something is good or bad). This system transforms uncertain and threatening aspects of the environment into affective responses (e.g., fear, dread, anxiety) and thus represents risk as a feeling. The psychological risk dimensions identified by the psychometric paradigm described above clearly are mostly affective in nature and the likely result of association-based processing. Proponents of formal analysis, as a result of an intellectualistic attitude (the primacy of the cognitive over the ‘manifest image’ of everyday practice) tend to view affective responses to risk as irrational. Current understanding thus suggests that nothing could be further from the truth. The rational and the experiential system not only operate in parallel, but the former seems to depend on the latter for crucial input and guidance. Looking at Table 5, can we say that the risk dimensions omitted (or at best tacitly incorporated according to an implicit value judgment) by the expert approach are ‘irrational’ (e.g. questions of social justice, intergenerational equity, etc.)? Is it irrational to prefer a risky activity over another on the premise that you trust the institutions involved in managing the first activity more than the others, ceteris paribus? Again, to avoid yet another misconception, we want to emphasize that scientifically based assessments about selection rules for decision options with varying degrees of uncertainty as well as approaches inspired by the public understanding of risk are both rational procedures of selection despite remaining uncertainties and ambiguities. Both procedures cannot be substituted by intuition, public opinion or political pressure. Regardless whether a ‘science-based’ or ‘public understanding’ approach to sustainability evaluation or management is applied in public policy making, different actors (proponents of technology, regulatory agencies, etc.) need an ethically defensible and consistent set of procedures. Thus, rather than reverting to polarised positions, we suggest it would be more constructive to recognise that both systems have their own set of advantages, as well as biases and limitations. 295As a matter of fact, Sjöberg (2003) argues that much of the explanatory power of the psychometric paradigm derives from a) shared semantics of the risk dimensions used, b) the use of relatively few words and c) sampling from very restricted domains.


183

Thus, bearing in mind these essential qualifications (which essentially come down to the statement that by no means are we claiming that we have cast the full breadth and scope of the ‘public understanding of risk’ in some iron psychometric law – a claim which would of course violate the central tenets of a constructivist position), pointing out these different risk dimensions (albeit in a stylised way) nevertheless allows us to derive some insights relative to the present discussion. The most important lesson seems to be that, while experts may often reasonably claim greater authority with respect to the likely probabilities and magnitudes of individual effects, one should clearly recognise that expert (or, for that matter, any other person’s) judgments are essentially context-specific, political and ethical in character when it comes to a relative prioritisation of different effects. Each of the individual risk or sustainability dimensions thus represents an area where analysts (be it experts, policy makers, laypeople, etc.) must adopt extraneous choices which fundamentally involve certain value judgments. Furthermore, each of these choices potentially alters the relative ranking of different energy options, as the potential for pursuing technological options which are manifestly superior on almost all sustainability dimensions is most of the time rather limited in political reality. Significant trade-offs have to be made – for instance, the relative importance accorded to the dimensions of ‘gravity’ and ‘intergenerational equity’ (see Table 5) might seriously affect the relative performance of nuclear power. Seen in this light then, aspirations to construct a robust and generally unambiguous analytical characterisation of sustainability solely on the basis of quantitative indicators looks not only potentially futile, but it also might serve to hide the fact that a lot of the impacts simply cannot be scored quantitatively. But perhaps the most serious problem of the ‘analytical fix’ is pointed out by Stirling (1999b, p. 125): …None of this is new. Much of what has been discussed here is well known to many of those involved in the development of sustainability indicators and in the practice of the various disciplines of environmental appraisal. The problem is not that the difficulties are unknown. The issue is rather that the implications remain on the whole relatively under-explored by theoreticians and under-addressed by practitioners. More importantly, they tend to be under-communicated to the sponsors of the analysis, to the wider discourse on sustainability and – most seriously – to the stakeholders and other interested parties who stand to be affected by any individual decisions which are informed by appraisal under sustainability indicators…

All of the above objections can only lead to the conclusion that political decisions can (and should) not be based solely on a cost-benefit analysis informed by calculations of externalities (indeed, this should come as no surprise after our (meta-)theoretical explorations in chapters 1 and 2). Now of course, a quite reasonable remark here might be that these objections, while perhaps being of some interest to armchair philosophers, might be of less practical relevance for the harsh realities of real-world political decision making confronted with constraints in terms of time and resources. Robbing the political sphere of the certainties offered by the analytical fix might even be seen then as offering a licence to arbitrariness. In any case, the challenge of coming up with a better alternative to the ‘rational’ approach offered by comparative risk assessment or external cost calculations is a

184

Chapter 3

legitimate one 296 . Fortunately for us, meeting this demand (while avoiding the dangers of the ‘analytical fix’) seems to be relatively forthcoming in principle. Firstly, if different (electricity-generating) options cannot be meaningfully compared in terms of a single scalar ‘sustainability’ number, then we should rather see sustainability as a vector with as many dimensions deemed necessary. Secondly, based on the insight that, in the context of ethical and moral uncertainty over the priority of the different dimensions summed up in Table 5, sustainability cannot be laid down once and for all (even in principle) in one overriding statutory consensual basis, policy analysis should clearly not be concerned with a forlorn search for such indisputable base, but rather with mapping out the performance of different (electricity production) options under different (but equally legitimate) ethical assumptions 297 . The role of policy analysis then becomes the systematic and transparent exploration of the implications of different perspectives. In the future, we would like to refer to any form of policy analysis based on these premises as a ‘broad cost-benefit analysis’. What form could such broad cost-benefit analysis of strategies for sustainable energy policy actually take? This question will be at the centre of our attention in chapters 6 and 7 of our dissertation. For the moment, let us confine ourselves to an illustrative example, provided by Klinke and Renn (2001). These authors, based on an investigation of regulatory practices in a number of EU countries, have attempted to cast such practices into different policy-making scenarios 298 . Their point of departure is the observation that, in general, risk evaluation in a policy context starts from a broad classification of risks into three ‘domains’: the ‘normal’ area (risks with little statistical uncertainty, low catastrophic potential, little damage when the product of probability and consequences is taken, nonpersistent and local effects, and reversible consequences) where ‘normal’ risk-cost-benefit approaches can be applied; and the ‘intermediate’ and ‘intolerable’ area which cause more problems on any of these dimensions. For risks falling into the ‘intermediate’ or ‘intolerable’ area, Klinke and Renn (2001, p. 162) propose some additional criteria of risk evaluation (again predicated on actual regulatory practices): • Probability of occurrence (p): from 0 (absolute impossibility) to 1 (absolute certainty); • Extent of damage (d): from 0 (no damage whatsoever) to ‘infinite’ (catastrophic consequences); • Certainty of assessment: confidence interval on probability and extent of damage; 296

As demonstrated conclusively by Schrader-Frechette (1985), who systematically assesses and rejects arguments directed against the value of explicit ‘rational’ policy analysis. According to us, the most pervasive reason in favour of such analysis is that it at least helps to make political choices visible. Visibility in turn is an absolute precondition for making politicians more accountable for their choices (e.g. it enables investigation whether the cost-benefit analysis is applied consistently between similar cases). 297 We are referring here to our discussion of Arrow’s ‘impossibility theorem’ (Chapter 2 – Section 3.2.1). However, we do not mean to imply that there can be no consensus whatsoever on the priority of certain impact dimensions over others. In particular, as mentioned several times earlier in this text, the rise of the precautionary principle at the international level points at an increasing political recognition of anticipatory action in the face of uncertain and irreversible risks (cf. Chapter 5). 298 In chapter 5, we will undertake a more profound discussion on the role of scenarios in the governance process.


185

• Ubiquity: defined as the geographic dispersion of potential damages (i.e. relating to questions of intragenerational justice); • Persistency: defined as the temporal extension of potential damages (i.e. relating to questions of intergenerational justice); • Reversibility: defined as the possibility to restore the situation to the state before the damage occurred (possible restoration measures are e.g. reforestation, decontamination or cleaning of polluted water); • Delay effect: characterising a long time of latency between the initial event and the actual impact of damage; and • Potential of mobilisation: understood as violation of individual, social or cultural interests and values generating social conflicts and psychological reactions by individuals or groups who feel inflicted by the risk consequences. These risk criteria largely resonate with a selection of the dimensions taken up in Table 5, with potential of mobilisation functioning as a kind of ‘container concept’ intended to capture social risk dimensions such as the perceived level of personal or institutional control, equitable distribution of the impacts, voluntariness of the exposure, etc. Theoretically then a large number of risks classifications could be derived from a permutation of all eight of these risk dimensions. However, based on the insight that in reality not all possible combinations of risk dimensions are actually realised (e.g. some risk dimensions are tightly coupled, other combinations have no empirical counterpart in reality, etc.), Klinke and Renn (2001) propose a classification of practically occurring risks into six different classes 299 . Klinke and Renn’s main objective is not only to provide a consistent and comprehensive risk classification, but also (and more importantly) to gain an effective and feasible policy tool for the evaluation and the management of risks. Therefore, each risk class forms the base for designing specific political strategies and measures tailored to the specifics of the risk dimensions in question. These strategies pursue the goal of transforming ‘unacceptable’ into ‘acceptable’ risks, i.e. the risks should not be reduced to zero but moved into the area of ‘normal’ risks, in which routine risk management becomes sufficient to ensure safety. For instance, in the case of the Sword of Damocles risk class (which includes nuclear power), these strategies include a reduction of the damage potential (e.g. research to develop substitutes, technical measures to reduce the

299

Each of these classes is named after a character taken from Greek mythology. These include the Sword of Damocles (p low, d high, confidence interval of p and d low – e.g. nuclear power, chemical plants), Cyclops (p uncertain, d high, confidence interval of p high, confidence interval of d rather low – e.g. floods, earthquakes, volcanic eruptions, triggering of nuclear, chemical or biological weapons systems), Pythia (p uncertain, d uncertain (but potentially high), confidence intervals on p and d high – e.g. BSE, release and spread of transgenic plants), Pandora’s box (p uncertain, d uncertain (only presumptions), confidence intervals of p and d uncertain, persistency high (several generations) – e.g. persistent organic pollutants, endocrine disruptors), Cassandra (p high, d high, confidence interval of p rather high, confidence interval of d rather low, delay effect high – e.g. anthropogenic climate change, destabilisation of terrestrial ecosystems, health impacts of atmospheric pollution) and Medusa (p rather low, d rather low, confidence interval of p rather high, confidence interval of d rather low, potential of mobilisation high – e.g. electromagnetic fields). For our illustrative purposes, we sidestep the issue of who gets to decide which risk belongs under a certain risk category – an issue which Klinke and Renn (2001, pp. 168-170) do address.

186

Chapter 3

catastrophic potential, stringent liability rules, international safety standards, subsidies for alternatives with the same use, containment), increasing resilience (e.g. capacity building (training, monitoring), building in technical resilience (e.g. through redundant systems), international control (e.g. under the auspices of the ‘International Atomic Energy Agency’ (IAEA)), international liability commitment) and emergency management. All strategies and respective measures are arranged by Klinke and Renn (2001) according to priorities (e.g. reduction of the damage potential should have a higher political priority than emergency management – without of course neglecting measures of the latter kind). Now, it is clear that in practice – e.g. when designing measures for sustainable energy policy – risks falling into different risk classes have to be addressed and balanced at the same time. As resources are generally limited, priorities have to be set. On this level of judgment (deciding which risk class demands political attention first), it is clear that there can be no single ‘rational’ answer to the broad cost-benefit analysis involved, be it for the simple reason that the benefits of most policy measures can hardly be calculated (since for many risk classes probabilities and consequences are highly uncertain). However, classifications such as the one provided by Klinke and Renn (2001) do at least offer the opportunity for initiating a deliberative procedure, based on a rational common understanding and explicit formulation of the criteria used (hence ensuring the transparency of the decisions taken). Furthermore, these criteria can be used as norms for the (equally broad) evaluation of the policy measures (e.g. ‘do the measures taken succeed in moving a risk from the ‘intermediate’ to the ‘normal’ area, or – perhaps as a side-effect of a too restrictive focus – do they result in an aggravation of risks from another risk class, or do they have no verifiable result whatsoever?’). The next step in the decision-making process (deciding on the appropriate measures to be taken once the political priorities have been set) might already be more conducive to more ‘formal’ forms of cost-effectiveness or even cost-benefit analysis, since this step involves comparisons between different ‘risky’ technologies within the boundaries of one risk class. Hence, impacts will be already more comparable. As Klinke and Renn (2001, p. 168) suggest, if resources are limited, strategies and measures should be taken in line with the priority list they have developed for each risk class. This in turn implies that these authors have some idea of the (cost-)effectiveness of the different measures involved, although they do not explicitly reveal the reasoning or methods used in their priority-setting 300 . In the following sections, we will investigate to what extent the explicitness offered by external cost calculations can be helpful in this respect. 2.2.2

Quantifying the interests of affected populations

Even within this more circumscribed domain of priority-setting within the confines of one risk class, external cost calculations can only be helpful if they are based on scientifically sound insights – i.e. the insights derived from them should not result from arbitrary methodological choices. As mentioned in the introduction to this chapter, specific studies 300

One can only presume that, in line with the general argument in their article, this priority list reflects a ‘political’ cost-effectiveness judgment based on observations of a number of different regulatory settings.


187

on the (economic) externalities of energy use started in the ‘80s (e.g. Hohmeyer 1988; Ottinger et al. 1990; Pearce et al. 1992 – reviewed in EC 1995a). Since then, substantial progress has been made in estimating the monetary value of the impacts of different energy systems. Most of these ‘first generation studies’ are considered now to be of a very preliminary nature, far from the current state of the art, and highly aggregated and approximate (Eyre 1997). However, this methodological progress, which of course results in a change of externality estimates over time, is sometimes used as a pretext for launching a general critique on the methodology of externality calculations as such. For instance, in an oft-cited paper published in Energy Policy, Stirling (1997, p. 531 – cf. Figure 2 and notice the logarithmic scale in this graph!) concludes on the basis of a review of externality studies (32 different government and industry-sponsored studies dating from the period 1979 – 1995!) that …the degree of overlap in this aggregate picture would accommodate any conceivable ranking order for these eight options. One of the most basic tasks in environmental appraisal is to achieve, for any given set of assumptions, some notion of the ordinal ranking of the various options under consideration. Without this, then there must be serious questions over whether a particular appraisal methodology is of any practical policy use at all (our italics)…

Figure 2. Ambiguity in the ranking of electricity supply options in the literature on externalities (Source: Stirling 1997, p. 532)

We believe this to be a rather partisan point of view. Stirling, seemingly overeager to throw away the baby with the bathwater, fails to give any consideration to the possibility that ‘any given set of assumptions’ might be better (from the point of view of economic

188

Chapter 3

science) than another 301 . The disparity of results among studies will only be a serious problem if it can be shown to arise from inherent problems with the various methodologies available to researchers. If there is no meaningful basis for selecting between different methods of calculating externalities, and if these different methods cause the large disparity in appraisal results, then of course no reliable policy conclusions can be drawn from then. Furthermore, the comparison of externality estimates across studies will become equally meaningless (except for the purposes of invalidating the methodology as such). Stirling (1997, p. 529), aware of these issues, points out that, based on the observed evolution of externality estimates in his overview of the externality literature, there appears to be no obvious process of convergence of the sort which might be expected in a ‘maturing paradigm’ under an established set of assumptions. Indeed, it seems quite reasonable to expect that some amount of learning should occur among researchers in the field of externalities of electricity production – i.e. repeated assessments will likely lead to higher convergence as new methods are developed and older ones improved. But Stirling overlooks the fact that newer (and better) studies also will extend their analysis towards ever more impacts for monetarisation. Thus, overall, the ‘learning effect’ will likely work in two ways: the convergence will tend to decrease the variability of externality estimates, while incorporating more and more impacts will tend to produce higher total estimates with cumulative higher uncertainties. In essence, Stirling’s problem is that he fails to make the distinction (in constrast to what we did at the beginning of this chapter) between the choice to apply a method and the methodological choices made within the framework of the chosen method. On the first point, his arguments raise a valid critique; on the second, they do not. The second point can only be addressed by a proper meta-analysis of externality studies, involving a statistical effort to explain (and learn from) the variance in impact estimates, instead of simply surveying and juxtaposing different results. Such learning experience is precisely undertaken on some very specific (and technical) issues in the still ongoing ExternE project with the ultimate aim of becoming the ‘benchmark’ in externality studies. To what extent this goal is reached will be discussed under section 3. Let us just conclude this section with a brief overview of some of the methodological uncertainties which had a major influence on the variability observed in the ‘first generation studies’, but which can now be considered as ‘settled’. These include: 1. The inclusion or exclusion of different impact factors; 2. The selection of boundaries of the system under scrutiny; 3. Site-dependence of the results (most notably for local and regional air pollution – e.g. larger populations down-wind from a power plant will be subject to larger externalities);

301 Stirling fails to make a distinction between ‘the context of discovery’ (where intra-scientific validity criteria should be applied) and ‘the context of justification’ (where criteria for the translation of scientific results into policy recommendations should be applied). He is however right in pointing out the many value-based choices that have to be made in the context of externality calculations (and we have greatly benefited from his thoughtful exposition of these choices – see Table 5), but he goes too far in denying externality studies practically all policy-making relevance.


189

4. Methodological questions about appropriate economic valuation tools (and often a lack of suitable economic valuation studies) 302 . On the first issue, it is clear that, working within an economic framework, ‘all’ impacts should be included insofar as they have an unpriced and/or uncompensated impact on another actor. Some impacts can of course hardly be quantified in economic terms (e.g. aesthetic impacts, proliferation issues), but this does not imply that such impacts cannot be addressed in qualitative terms in externality studies (as is common practice in the ExternEproject). In fact, this issue then comes down to an adequate and correct communication of the results and inherent limitations of externality studies towards larger audiences rather than an inherent flaw. Every theory has its particular blind spots, and economic theory is no exception. Of course, what is considered to be an ‘important’ impact can vary from study to study (Sundqvist 2004). But this observation appears to be of only marginal practical importance, since for most technologies a large consensus exists on the ‘important’ impact factors (cf. infra). Essentially the same observations hold true for the second issue (system boundaries). While it is true in principle that “…the scope for regression into successive depths of the economy renders decisions over system boundaries an essentially subjective judgment…” (Stirling 1997, p. 527), it is clear that analysis should at least address all fuel cycle stages (from mining of the resource to final waste streams), while a qualitative judgment regarding the order of magnitude of other impacts (e.g. the environmental effects of the day-to-day commuting of workers to power plants) compared to the other impacts along the fuel cycle generally suffices for an informed (albeit still intersubjective) decision whether or not to take up these impacts (or a decision to initiate further research into the issue). The point here is not that impacts considered to be ‘negligible’ from the point of view of externality calculations (e.g. local impacts will often be small in economic terms when normalised over the entire lifetime of electricity production and compared to the other impacts of a fossil-fuelled power plant) should be considered to be ‘negligible’ in general, but rather that these impacts should be addressed through other instruments than purely economic ones. On the third issue, it is quite clear that site-dependent evaluations will provide more accurate pictures of externalities then do evaluations based on ‘hypothetical’ power plant locations. Finally, concerning the last issue of impact valuation, two principal methods generally used are the ‘bottom-up’ calculation of damage costs and the calculation of abatement (or control) costs. Although abatement costs are often (mistakenly) seen as estimates of

302

Some possibilities are: the ‘mitigation cost’ approach (expressing costs in terms of the alleviation of environmental damage once committed), the ‘hedonic market’ or ‘contingent valuation’ approach (expressing costs through an examination of prevailing property or wage markets, or responses to questionnaires), the ‘abatement cost’ approach (taking the costs for controlling the contamination at the source as a proxy for the social costs of the impacts themselves), and the ‘bottom-up’ approach (expressing costs associated with a physical dose-response relationship).

190

Chapter 3

damage costs, conceptually they are very different. Damage costs are a measure of society’s loss of welfare resulting from the damage arising from a specific adverse environmental impact. Abatement costs are what it costs society to achieve a given standard that restricts the extent of the impact to an acceptable level, and are thus likely to be only tenuously related to damage. Abatement costs were often used as a surrogate for damage costs as they are a relatively straightforward concept, are relatively easy to derive, and can be applied to most environmental impacts. Essentially, abatement costs can be calculated simply by dividing the cost of mandated controls by the emissions reduction achieved by the controls. In general, however, existing abatement costs (i.e. the costs for meeting certain imposed standards) must be viewed as a poor substitute for estimating damage costs. The implicit assumption in using abatement costs as a substitute for damage costs is that society controls pollution until the benefits of additional controls would be outweighed by the costs of their imposition. But using the cost of regulation to estimate the benefits is rather a meaningless and circular procedure, given that a cost benefit ratio of unity will always be achieved: since society ‘knows’ the correct extent of damages, any restriction on pollution it imposes on itself will automatically be the most ‘rational’ one possible. A further flaw is that use of abatement costs to value externalities implies that legislators are able to make optimal decisions when imposing policy instruments to modify polluting behaviour to achieve such an ‘optimal’ outcome. However, in practice, epidemiological studies of e.g. cost per life saved have indicated large variations in the values implied by the costs and benefits of different regulations (Owen 2004). Estimation of ‘bottom-up’ damage costs is thus clearly the preferred option according to standard economic theory 303 . It focuses directly on explicitly expressed preferences as revealed by willingness to pay to avoid environmental damage or by stated preferences in either real or simulated markets. Clearly, however, a major disadvantage is the scale of the data requirements for deriving estimates of these damage costs 304 . Sundqvist (2004), using the factors of ‘method’ (i.e. abatement vs. damage costs), ‘fuel type’ (i.e. the different technologies for electricity production), ‘fuel stages’ (i.e. the scope of the study) and ‘site dependence’ as explanatory variables, has performed a statistical meta-analysis of the disparity of results among a large sample of electricity externality studies 305 . He found out that the choice of method (with the ‘bottom-up’ damage approach resulting in lower externality estimates) and the scope of studies (with the studies restricted to the generation stage resulting in lower externality estimates) were major systematical explanatory variables for the observed disparity of results; while ‘site dependence’ did not prove to have a conclusive impact. Investigation of the ‘fuel cycle’ variable confirmed the rather uncontroversial overall impression that renewable energy options (hydro, wind,

303 Hence, although the work of Hohmeyer (1988) and others advanced the debate on externalities research considerably, the style of analysis was too simplistic for adoption for policy analysis. In particular, no account can be taken of the dependence of damage with the location of emission, beyond minor corrections for variation of income at the valuation stage. 304 The lack of data is often a reason for reverting to the less precise estimations based on abatement costs. 305 Sundqvist (2004) analysed 63 externality studies conducted over the last 20 years.


191

solar) 306 display a generally better environmental performance than do the fossil fuels (especially coal and oil), while for gas, nuclear and biomass no statistically strong correlation with low external costs could be established 307 . Still, this result for biomass is somewhat surprising as the use of biomass for power generation is often actively promoted through different subsidy schemes (also in Belgium). Sundqvist’s analysis – while not sufficient to explain all of the variability in externality studies – does indicate that the disparity in externality estimates is caused for a large part by systematic (rather than arbitrary) methodological reasons and to problems in the application of methods, and thus, that the search for more robust methods (within the economic paradigm) makes sense. This still leaves unanswered the question of where one should draw the line between ‘political’ judgment on the one hand and economic valuation on the other (Section 5). However, before tackling this subject, we cannot avoid going somewhat deeper into the ‘state-of-the-art’ in externality studies – the ExternE project.

3 The ExternE project In response to the general theoretical difficulties experienced in attempting to quantify the external costs of electricity production, the EC together with the USDOE in 1991 launched a joint research project to identify the appropriate methodology for this type of work. After the first phase of the project, which ran over 4 years and established an operational accounting framework for the assessment of external costs of energy production (named ‘ExternE’ in Europe) (EC 1995a), the EC’s Directorate General XII continued an independent programme of follow-up activities. The expressed principal objectives of ExternE were: 1. to develop a unified methodology for quantifying the environmental impacts and social costs associated with production and consumption of energy from different fuels; 2. to use this methodology to evaluate the external costs of incremental use of different fuel cycles in different locations of the European Union; 3. to identify critical methodological issues and research needs. More than 50 teams from 15 countries in Europe participated in these follow-up activities on the improvement, dissemination and application of the ExternE accounting framework (EC 1999a). In the meantime, the increasing acceptance of the ExternE methodology went along with a broadening of the scope of application: ExternE results were used as input in the cost-benefit analysis of European environmental policy measures, green national accounting, and the methodology was extended to include the external costs of transport 306

The Belgian team responsible for the implementation of the ExternE methodology however came to quite different conclusions concerning the external costs of photovoltaics (although these results are classified as preliminary) – see section 3. 307 In the Sundqvist (2004) paper, the dependent variable is the ‘worst case’ total external cost estimate produced by the individual studies. External costs are then categorised as ‘high’/’low’ if they exceed/are lower than the average ‘internal’ production cost of the eight technologies taken up in the overview (set at 3.97 US cents per kWh).

192

Chapter 3

(see the ExternE website ). The latter project (ExternE Core/ Transport) resulted also in a major review of the general methodology (as reported in Rainer and Bickel 2001). Work in the ExternE project in the more recent past included the ‘NewExt project’ (2001-2004), which aimed to improve the assessment of externalities by providing new methodological elements for integration into the existing accounting framework that reflect the most important new developments in the assessment of external costs 308 ; while the ‘ExternE-Pol project’ (2001-2004) aimed to provide an assessment of externalities that have evaded quantification so far (e.g. energy supply security) and extend the methodology to new technologies (e.g. fuel cells, residential heating technologies) and other countries in the enlarged European Union (e.g. Czech Republic, Poland, Hungary). Recently, a new methodological update has been published (EC 2005) 309 . Ongoing work focuses on disseminating the finding towards policy makers (the ‘Maxima project’, 2002-2006) and harmonising and sharing of estimates for environmental and health externalities developed also in other fields (e.g. agriculture, industry, waste, etc.) with the aim of arriving at best practice guidance (the ‘Methodex project’, 2004-2006). In any case, the work done so far over more than a decade has thus been reviewed extensively through the expert networks engaged within the project and by comparison of the work of different teams. It is safe to say that ExternE represents the largest and most thorough body of work in the field of environmental economics of energy production and use. The following sections give a brief overview of the core concepts underpinning the ExternE methodology before going into the remaining uncertainties and general robustness of the method 310 .

3.1

Life cycle analysis

When comparing the environmental footprints of alternative energy technologies, it is important that the power generation or combustion stage of the technology not be isolated from other stages of the cycle. To give a particularly relevant example, fuel cells emit virtually no greenhouse gases in their operation if they run on hydrogen. However production of their fuel (hydrogen) from fossil fuels may involve increases in greenhouse gas emissions in excess of those that would arise from using current commercial fossil fuel technologies to meet the same level of energy requirements (an oft forgotten issue). To avoid such distortions, the concept of life cycle analysis (LCA) has been developed since the beginning of the ‘70s. LCA is based upon a comprehensive accounting of all energy and material flows (from ‘cradle to grave’) associated with a system or process. The

308

These include a better understanding of the monetary valuation of the increased mortality risks from air pollution (based on empirical investigations of the ‘value of a statistical life year lost’), monetary valuation of ecological and CO2 emission impacts based on the preferences revealed in political negotiations, an assessment of environmental impacts and externalities resulting from multi-compartment (air/soil/water) impact pathways, an assessment of the externalities of major accidents in non-nuclear fuel chains, including a methodological testing of the new methods and calculated externalities (see the ExternE website ). 309 Since this methodological update was not yet published at the time of writing this chapter it is not taken into account in the following sections. 310 For a very concise review of the ExternE achievements, see also Vanooteghem (2005).


193

approach has typically been used to compare the environmental performance of different products that fulfil similar functions. In the context of an energy product, process, or service, an LCA would typically analyse the emissions and energy use associated with fuel extraction, transportation and preparation of fuels and other inputs, plant construction, plant operation, waste disposal, and plant decommissioning (see e.g. IEA 1993). The ExternE methodology has numerous links to LCA. The central concept of fuel cycle or fuel chain analysis, in which all components of a given system are analysed from cradle to grave, corresponds with the LCA framework. There are however also some major differences between a typical LCA analysis and an externality analysis. In an LCA no attempt is made to prioritise between the different impacts, as such analysis is concerned only with mapping out the different material and energy streams associated with a certain process or product. Furthermore, in an LCA only material and energy flows are assessed, thus ignoring some externalities (such as the fiscal externalities associated with supply security). As with an LCA, the starting point for fuel chain analysis is the definition of the temporal and spatial boundaries of the system under investigation, and the range of burdens and impacts to be addressed. The range of burdens and impacts addressed in the ExternE project is very broad. The first step in the project consisted in simply listing all (according to present expert knowledge) possible burdens; initially no account has been taken of the likelihood of any particular burden actually causing an impact, whether serious or not. The first series of ExternE reports (EC 1995a, Vol. 1-6) provided comprehensive listings of burdens and impacts for most of the fuel chains considered. The purpose of the exercise was simply to catalogue everything to provide a basis for the analysis of different fuel chains to be conducted in a consistent and transparent manner, and to provide a firm basis for revision of the analysis as more information on the effects of different burdens could become available in the future. In keeping with this broad view on burdens, the ExternE project also attempted to address the full spatial (from local to global) and temporal (from immediate to 100,000 years) range of impacts. The consideration of such long time intervals clearly raises important methodological questions, which will be addressed in section 3.4.3, but the basic idea is nevertheless sound: no impact should be removed a priori from consideration on an arbitrary basis, or just because it will be difficult to quantify in economic terms. Also, in the presentation of the results ExternE clearly indicates the spatial (local/regional/global) and temporal (short term/medium term/long term) characteristics of impacts, which is a sign of good practice.

3.2

The impact pathway approach

Having identified the range of burdens and impacts that result from a fuel chain, and defined the technologies under investigation, the analysis typically proceeds with a prioritisation of impacts, a description of priority impact pathways, quantification of burdens (e.g. emissions levels), description of the ‘receiving environment’ (e.g. a doseresponse function describing a human health impact), quantification of impacts (e.g. the number of additional cancers caused by the emissions), economic valuation (e.g. economic valuation of cancer treatment) and finally an assessment of uncertainties. Prioritisation of

194

Chapter 3

impacts is done with an eye on keeping the number of impacts to be considered at a reasonable level, whilst however taking care that the analysis covers those impacts which represent the greatest externalities (cf. supra). Next, the ExternE project uses the impact pathway approach for the assessment of the external impacts and associated costs resulting from the supply and use of energy. An ‘impact pathway’ systematically traces the link between a burden (i.e. something that potentially causes an impact) and the economic valuation of the final impact. The analysis proceeds sequentially through the pathway, as shown in Figure 3 for the case of acidifying emissions. Emissions and other types of burden such as risk of accident are quantified and followed through to impact assessment and economic valuation 311 .

Figure 3. The impact pathway approach (Source: EC 1995a – Vol.1)

An important aspect of the impact pathway methodology is the accumulation of uncertainties in each step of the pathway. For instance, the uncertainty interval on the emission levels of point sources is usually between 10-50% in the European context (Torfs 2001); and dispersion modelling further adds to the uncertainty of the final damage estimate. One of the most important conclusions of the ExternE project is precisely that uncertainties are very large in some cases. Therefore, in the 1999 ExternE methodology

311

There exists an extensive literature on different environmental valuation methods, and to review this in detail here would be beyond the scope of the present chapter. For an excellent overview however see Garrod and Willis (1999).


195

update, a lot of attention was given to the development of a scientifically sound method for dealing with uncertainties. The danger was recognised that earlier estimates, which offer apparently precise results on the basis of perhaps not so reliable data, could be misleading. Uncertainties arising within the ExternE methodology can take many different forms (cf. Chapter 2). In some cases the uncertainty is statistical in character, i.e. numbers show a statistical spread around a mean value (e.g. in the case of some technical data, the use of dose-response functions). Uncertainty can also arise due to imperfect modelling, i.e. as a result of model limitations or imperfect knowledge of the phenomena to be modelled. Other uncertainties are even less amenable to mathematical formalisation. For many longterm effects, such as global warming or the security of nuclear waste repositories, the impacts are very scenario-dependent. Other inputs to the assessments (e.g. the discount rate and the statistical value of life in different countries) have inevitable political or ethical components which make the construction of probability distributions highly controversial. As these choices tend to have a large impact on the final results, we will discuss them separately under section 3.4.3. For the statistical uncertainties, the ExternE team has found a very elegant technical way of indicating confidence in results. Since the impact pathway analysis is typically multiplicative, the distribution of outcomes from such a multiplicative analysis is typically lognormal (in other words the logarithm of the outcome follows the normal – i.e. Gaussian – probability distribution) (EC 1999a) 312 . With the lognormal distribution the confidence interval is predicted from the geometric mean (µg) and the geometric standard deviation (σg) 313 . The 68% confidence limits (i.e. the outcome has a 68% chance of lying between these limits) are then defined by the range µg/σg to µg.σg; the 95% confidence limits by the range µg/σg² to µg.σg², and so on. Now, due to the specific properties of logarithmic calculations, the square of the geometric standard deviation for the entire impact pathway is simply found as the addition of the squares of the deviations of each separate step. Thus, each separate element of the impact pathway (emission levels, dispersion modelling, doseresponse functions, etc.) is characterised within the ExternE project with a geometric standard deviation, making it easy to derive the composite number for an entire pathway. To clarify things even further, the concept of a ‘confidence band’ was introduced. Estimations of the geometric standard deviation were grouped into three broad classes: ‘high’ reliability (letter code A; 2.5 < σg < 4); ‘moderate’ reliability (letter code B; 4 < σg < 6) and ‘low’ reliability (letter code C; 6 < σg < 12). For instance, for externality estimates marked with the letter code A the ‘true’ value of the externality could be 2.5 to 4 times smaller or larger than the median estimate (with an 68% confidence limit). Externality estimates marked with the letter code C have an 68% chance of being 6 to 12 times larger or smaller than the median estimate, and a 95% chance of being 36 to 144 times larger or smaller. The latter result implies that the ‘true’ externality estimate lies somewhere in an

312

For example, air pollution effects on health are calculated as: ‘damage’ = ‘pollution concentration’ x ‘population’ x ‘exposure-response function’ x ‘valuation’. 313 For the lognormal distribution, µ is approximately equal to the median value. g

196

Chapter 3

interval possibly spanning 4 orders of magnitude around a median estimate, which is considerable 314 .

3.3

Results

All in all then, the following results are obtained (see Table 6 and Table 7). For the fossil fuel technologies, the impacts of public health damages of air pollution are, together with the impact of global warming, the most important impacts that have been quantified. This statement needs to be qualified however because the ExternE estimates (along with other studies) do not provide external cost estimates for ecosystem impacts beyond crop damage 315 . However, both health impacts and impacts of global warming are still subject to large uncertainties as witnessed by the large ranges in the reported external costs (cf. next section) 316 . For the nuclear fuel cycle, the priority pathways considered are those concerned with the radiological impacts on human health and occupational hazards. The indicators used to assess human health impacts of radiation include fatal cancers, non-fatal cancers and severe hereditary effects in future generations, estimated according to the guidelines of the ‘International Commission on Radiological Protection’ (ICRP). Renewable energy sources assessed in the Belgian context are wind energy, biomass and photovoltaics. Photovoltaics show very high external costs associated with the production process of the modules. However, damage costs for pollutants emitted in Belgium were used, which is an overestimation, since most components are fabricated outside Belgium. For biomass, many options are possible (only one of which is shown here); all options have the advantage of being CO2 neutral. Wind energy has very low external costs, associated mainly with the construction of the wind turbines. The effects of visual pollution have not been assessed; however, they are believed to be small if the process of site selection is carefully managed.

314

I.e. the interval between median/12² and median x 12² (in the worst-case scenario). The tools developed under the ExternE project (the Ecosense model) do allow the assessment of the change in ecosystem area in which ‘critical loads’ for acidification are exceeded per unit emission, but the current state of knowledge does not allow to calculate the exact physical damage to ecosystems resulting from such exceedance of critical loads (Krewitt 2002, p. 843). 316 Today, a majority of the scientific community agrees with the hypothesis that the increasing atmospheric concentration of greenhouse gases, caused mainly by a burning of fossil fuels, will lead to a warming of the earth’s atmosphere. While observed changes of the global mean temperature first had been accounted for as a natural phenomenon due to normal climate variability, it was only in the mid-’80s that, based on the first results of the ‘World Climate Program’, climate change science started to focus on the additional anthropogenic greenhouse effect. The central question became to what extent human activities contributed to an increasing concentration of greenhouse gases, which in turn, through the resulting global warming, could inter alia lead to a rise in sea levels, far reaching modifications of weather patterns and nearly unpredictable feedbacks with other climate parameters (IPCC 2001). 315


Externality

Coal

Gas

197

Nuclear (open fuel cycle)

Public Health

Mortality, morbidity


Radiation and non-

impacts of production

impacts of production

radiation impacts:

(PM, ozone, SO2, NOx);

(ozone, SO2, NOx)

collective dose (over

transport impacts

entire fuel cycle); mortality, morbidity impacts (PM, SO2, NOx) over fuel cycle (excl. transport and production)

Occupational health

Diseases from mining

Accidents during

Radiation and non-

and accidents during

transport of gas

radiation impacts:

mining, transport,

(pipelines), construction

collective dose (over

construction and

and dismantling

entire fuel cycle) and

dismantling

accidents (over entire fuel cycle)

Major accidents

NQ

NQ

Major nuclear accidents

Crops

Crop and soil impacts


No impact

(sulphur, acidification

(sulphur,

and ozone)

and ozone)

acidification

Ecosystems

Only impact quantified

Only impact quantified

NQ

Materials

Sulphur and acidification


No impact

damages on surfaces

damages on surfaces

Operational impacts of

Production

NQ

Noise

traffic Visual impacts

NQ

NQ

NQ

Global warming

CO2, CH4 and N2O

CO2, CH4 and N2O

CO2, CH4 and N2O

damages

damages

damages during transport and construction phases

Total estimate (Euro/MWhe)

0.5 – 49.0

No FGD nor SCR

9.8 – 322.8

With FGD and SCR

1.8 – 125.5

0.1 – 3.7

FGD = Flue Gas Desulphurisation; SCR = Selective Catalytic Reduction NQ = Not quantified Ranges in economic impact valuation include uncertainty of global warming impact (low and high estimate) and include an uncertainty interval (68% confidence level) Value reported for the nuclear fuel based on 0% discount rate over a time horizon of 10,000 years; damage costs of severe nuclear accidents are not included are generally reported separately (but see Section 4.1) Table 6. External costs of non-renewable electricity production technologies in Belgium (Source:Torfs 2001, p.34; based on ExternE Core/Transport methodological update)

198

Chapter 3

Externality

IGCC biomass

Wind

Photovoltaics

(2005 Technology) Public Health




impacts of emissions



during culture and

(PM, SO2, NOx) in the

(PM, SO2, NOx) in the

transport of biomass

production phase of

production phase of

materials for wind

photovoltaic modules

turbines Occupational health

Accidents during

Accidents during

transport of biomass

transport of materials,

No impact

construction and operation of wind turbine Major accidents

NQ

No impact

No impact

Crops





(sulphur,


and ozone)

and ozone)

and ozone)

Ecosystems

NQ

NQ but very small

NQ

Materials




damages on surfaces

damages on surfaces

damages on surfaces

Operational traffic

Electricity generation

No impact

Noise

acidification

impacts Visual impacts

NQ

NQ

No impact

Global warming

CO2 damages in transport

CO2 damages in the

CO2 damages in

of biomass

production phase of

production phase of

materials for wind

photovoltaic modules

turbines Total estimate (Euro/MWhe)

3.6 – 7.0

100

NQ = Not Quantified Results for photovoltaics are preliminary; more precise production data for the Belgian context are needed. Data for wind energy do not take into account public perception of local effects (e.g. aesthetic impacts) Table 7. External costs of renewable electricity production technologies in Belgium (Source: Torfs 2001, pp. 40-63)


3.4

199

Major uncertainties and general robustness of ExternE results

From a quick investigation of the ExternE results, it becomes apparent that global warming and public health impacts account for virtually all the externalities from all fuel cycles. The fact that health and global warming effects dominate is important since both are very controversial both in terms of ‘dose-response’ literature and in terms of economic valuation (Pearce 2002). We will give here just a cursory look at the major uncertainties; the nuclear fuel cycle will be dealt with in more detail in section 4. 3.4.1

Health impacts of air pollution

The assessment of the health impacts from air pollution has turned out to be one of the most successful activities of the project. In particular, ExternE quite early pointed out the importance of the exposure to fine particles for human health, at a time when this was not generally believed to be a significant threat. The ExternE team has been closely linked to leading epidemiologists in this field, which facilitated an intense review and discussion process which resulted in generally more robust insights (Krewitt 2002). The health impacts of fine dust particles have been demonstrated on an epidemiological basis, and acceptance of the dose-response relationships is growing internationally (e.g. by the ‘World Health Organisation’ (WHO)). It is nowadays considered to be a high priority impact, and air quality criteria setting limits on the amount of particles figure in most policy documents (at least in industrialised countries). There is still disagreement however on a number of topics: whether the health impact of fine particles is caused by the number of particles, their mass concentration, or their chemical composition; whether there exists a threshold value below which no more health effects can be observed; whether the health impacts of fine particles and ozone are additive; whether the same dose-response relationships can be used for the whole of Europe; whether results of North-American studies can be used in the European context; etc. In terms of the risks of death, there are two types of literature. The first relates acute episodes of pollution to life risks and the second (far smaller) relates chronic exposures to air pollution to life expectancy. Although it is becoming increasingly clear that the chronic exposure epidemiology is more important (Torfs 2001, p. 66), quantification of these impacts is also still very much uncertain 317 . Furthermore, economic valuation of these impacts also turns out to be quite controversial 318 . All in all then, it

317

Torfs (2001, p. 65) reports how a panel of British experts was asked to rank 13 different impacts according to their confidence in the assessment of these impacts. Chronic mortality risks were ranked at the lowest place (low confidence in the assessment); acute mortality was ranked third (high confidence). 318 Roughly, the debate turns on the following issues. Firstly, for economic calculations, one needs to know the actual period of life that is foreshortened by the acute exposures to air pollutants, and not just the number of deaths caused by a period of acute pollution. Pearce (2002, p. 28) tends to believe that this is a matter of days (and not years). Furthermore, the evidence tends to suggest that the deaths tend to occur in the older agegroups (over 70 years old). Thus, the relevant economic value for valuing acute mortality will be the willingness to pay of people older than 70 to avoid days rather than years of life lost. However, most valuation studies base the ‘value of a human life’ on the willingness to pay of median age groups involved in accidents, which will give higher economic values (commonly around 3 MEuro). For chronic exposures, the relevant economic value would be the amount of money one is willing to spend to extend one’s life by 6 months – 1

200

Chapter 3

seems that Pearce (2002, p. 31) gives a fair appraisal of the general robustness of health impact evaluation in ExternE: …While the usual academic conclusion that ‘more research is needed’ always seems frustrating to policy makers, the fact is that we do not know enough about the epidemiology and we certainly do not know enough about the economic valuation of life risks to be confident about the kinds of adders being produced in externality studies. In some cases, being more certain of the absolute magnitude of the adders may not matter too much. For example if we simply wish to prioritise investments by social cost, a ranking may not be affected by what values we use. But if we wish to use the values to set, say, energy taxes, then the absolute magnitudes do matter…

3.4.2

Global warming

In the first phase of ExternE, the existing literature on the environmental damage costs from greenhouse gas emissions was reviewed, and it was concluded that “…all attempts to value these impacts require important normative judgements, and therefore the potential for synthesis or consensus is remote…” (EC 1995a). The main problems related to the quantification of climate change might be summarised as follows: • Climate change impacts are very complex, with a very large number of different physical endpoints possibly affected. One has to rely on complex climate computer models for an estimation of the impacts of climate change; however, the understanding of many mechanisms is still poor. As a consequence, the models studied in ExternE only cover a time horizon of 100 years (up to 2100), which is fundamentally inappropriate in view of the long-term nature of the effects. Furthermore, most damage estimates based on these models do not take into account possible (and likely) adaptation behaviour to global warming – Pearce (2002) calls this the ‘dumb farmer syndrome’; • Strictly following the theoretical foundations of monetarisation, climate change impacts should be valued at regional or national prices. This may be objectionable for ethical reasons, since a life lost in a poor country will ‘cost’ less than a life lost in a rich country. To compensate for this effect, in the ExternE work, the global warming damages are ‘equity-weighted’ 319 (EC 1999a – Vol. 8). But then the actual value of this equity weight and its unique application to the problem of global warming becomes the subject of some controversy, further complicated by the fact that it will be the descendants of the ‘poor’ generations of today who will benefit from present efforts to limit impacts of climate change (Pearce 2002); • Climate change is a long term problem, so from the economic point of view discounting becomes a very important issue (cf. infra). Damage estimates vary considerably with the discount rate used. Furthermore, while economists are usually able to estimate the social discount rate for a single national economy, the relevant discount rate for the global warming damages would be one that would apply to the whole world. ExternE consequently in view of these difficulties gives results for three values of the discount rate: 0%, 1% and 3%. year in the age group of 70 – 80 years old. The ‘NewExt’-project within ExternE performs empirical valuation studies in three European countries to clarify the issue. 319 Predicated on the fact that one Euro of damage to a low-income country will represent a larger disutility than one Euro of damage to a high-income country.


201

In view of these uncertainties, it is perhaps a bit surprising that starting from the 1999 methodology update (EC 1999a), ExternE recommends the use of a ‘central’ damage estimate of 2.4 Euro/ton CO2 (with a ‘minimum’ of 0.1 Euro/ton CO2 and a ‘maximum’ of 16.4 Euro/ton CO2), based on a discount factor of 1% (but only taking into account damages occurring within a horizon of 100 years). The environmental economist faces a dilemma here: is it better to leave out potentially important external damages form the valuation and present biased estimates or should one make use of ‘rough estimates’ so as to provide (or at least approach) some kind of ‘full cost’ estimate? 3.4.3

Discounting

Discounting is the practice of placing lower numerical values on future benefits and costs as compared to present benefits and costs 320 . In the context of externalities it is an important issue because many of the environmental damages of present actions will occur many years from now (e.g. global warming, loss of biological diversity, storage of nuclear waste, decommissioning of power plants, etc.) and the higher the discount rate, the lower the value that will be attached to these damages. The practice of discounting in the context of environmental costs and damages is criticised by environmentalists (using different arguments, such as taking into account risk and uncertainty, rights of future generations, irreversible damages, etc.), who argue in favour of a 0% discount rate (or even sometimes a negative discount rates, as they argue that in the future, environmental assets will become even scarcer, and hence, people will be willing to pay more for them). Economists however are adamant in upholding a positive value for the discount rate (Pearce et al. 2003) 321 . Recent proposals for the ‘correct’ discount rate in the context of long-term environmental damages include a proposal by Rabl (1996) for a two-step discounting procedure, using a conventional social discount rate (3-8% is usually recommended) up to a certain time, determined by the average duration of long-term commercial loans (Rabl proposes 30 years), followed by a reduced rate for intergenerational effects (equal to the long-term growth rate of the economy, Rabl proposes a value of 1-2%). Pearce et al. (2003) on the other hand, based on empirical findings of actual consumer behaviour (most notably the phenomenon of ‘precautionary saving’), argue that discount rates to be used by governments in investment and policy appraisal should decline over time, up to a low value of about 1%.

320

Suppose an investment project entails a single benefit B (net of investment) at time t=0 and a single cost C at time t=N. Discounted at rate r, the net present value P = B – exp(-r.N).C. This equation highlights the traditional justification for the discounting of long term costs: the benefit B can be invested at rate r and the resulting future value is available to pay for the future cost C. 321 The principal argument being that the effect of lowering the discount rate towards zero is to increase the amount of saving that the current generation should undertake. With a zero discount rate, further reductions in consumption would always be justified in the name of increasing future generations’ consumption. This would effectively reduce the present generation to a subsistence level of consumption, sacrificing their interests to those of future generations.

202

3.4.4

Chapter 3

Conclusion

How do these uncertainties affect the general robustness of the ExternE results? Krewitt (2002, p. 841) shows how, in the past, the assessment of the external costs for a coal-fired power plant (costs of climate change not included) rose from 15 Euro/MWh in 1994 to 45 Euro/MWh in 1996 (due to the inclusion of chronic mortality impacts of fine particles); decreasing again to 15 Euro/MWh in 1997 (due to the introduction of valuation based on the years of life lost) and to 6 Euro/MWh in 2000 (due to the adaptation of dose-response functions for chronic mortality derived from epidemiological studies in the US to European conditions). The fact that the results change over time is not in itself not unduly problematic, and should rather be seen as a ‘normal’ result of scientific learning. In fact, it seems fair to say that the ExternE project has been very quick to react on and learn from scientific developments in related fields. That the scientists are not to blame for the rapid and sometimes quite drastic changes is witnessed by the fact that these changes nevertheless have stayed within the professed uncertainty range 322 . The political implications are however quite sensitive to changes much smaller than the indicated level of uncertainty, which suggests a careful use of such estimates in the political arena. Furthermore, changing background assumptions as a result of the strive for providing the most up-to-date scientific information could have had an effect on the general public and political trust in the reliability of the estimates. It is of course difficult to predict to which extent the ongoing research might influence the external cost estimates in the future. However, in contrast to the climate change problem, for the health impacts even the ‘extreme’ possibilities provide a reasonable range of uncertainty – thus enabling at least an ‘order of magnitude’ estimation; and furthermore consensus seems to be growing as the scientific network of the ExternE project is expanding (Torfs 2001).

4 Externalities of the nuclear fuel cycle For the nuclear fuel cycle, the implementation in the Belgian context is closer to an indepth aggregation / transferability exercise than to a real implementation of the ExternE methodology in the sense that no specific dispersion modelling has been undertaken specifically for Belgian conditions. Impacts have been quantified through transferring impacts and damages of each fuel cycle stage for the mass or energy streams (e.g. radiation dose received per unit of enriched fuel or per unit of electricity produced) from the French case (discussed in EC 1995a – Vol. 5; Schieber and Schneider 2002; NEA 2003) to the Belgian fuel cycle. All steps of the fuel cycle were taken into account, from mining and milling of uranium ore to waste disposal (existing surface disposal site for low and intermediate level waste, hypothetical underground disposal for high level waste), via conversion of ore, enrichment (gaseous diffusion), fuel fabrication, electricity generation

322

Public health impacts from air pollution have been classified in the ‘moderate’ uncertainty class (letter code B), which according to the ExternE nomenclature corresponds to an uncertainty interval spanning about 1 to 1.5 orders of magnitude.


203

(and possibly also reprocessing of spent fuel), construction, dismantling of nuclear power plants and the transport of radioactive materials (by rail and road). The priority impacts of the nuclear fuel chain to the general public are radiological health impacts due to routine and accidental releases to the environment 323 . The source of these impacts is the release of radioactive materials through atmospheric, liquid and solid waste pathways. Occupational health impacts, from both radiological and non-radiological causes, were the next priority. These are mostly due to work accidents and radiation exposures 324 . However, there is some debate on the extent to which occupational health impacts can be considered as externalities (e.g. the risks involved might have been compensated for already by higher wages in the nuclear sector) (Schieber and Schneider 2002, p. 113). Including occupational health effects as externalities therefore represents a conservative estimate. Impacts on the environment of increased levels of natural background radiation due to the routine releases of radionuclides have not been considered as a priority impact pathway, except partially in the analysis of major accidental releases. This reflects the general anthropocentric perspective taken in the regulation of the risks of ionising radiation. The analysis of impacts from severe accidents involves complex issues; therefore, this impact is generally treated as a distinct category in the presentation of results (not included in the overall assessment of the externalities of the nuclear fuel cycle). Overall, the external costs of the nuclear fuel cycle in the Belgian context are represented in Table 8, which conveys the general message that the external costs of nuclear power are low compared to fossil fuel-powered stations, and of the same order of magnitude than renewable electricity generation options (wind energy, biomass). However, as the assessment of the nuclear fuel cycle seems to be plagued by many of the issues exposed in earlier sections of this chapter, this message can be seen as potentially misleading. In particular, the most important choices for the nuclear fuel cycle concern the definition of spatial and temporal boundaries. In addition, there are some aspects of the fuel cycle which prove to be very resistant to objectivation in terms of external costs. Some of these aspects will be discussed in greater detail in the following sections. Concerning the temporal and spatial boundaries of the analysis, a clear distinction has to be made between local impacts (0-100 km.), regional impacts (100-1000 km.) and global impacts (above 1000 km.); in the short term (< 1 year), the medium term (1 to 100 years) and the long term (100 to 100,000 years). The major contributors to health impacts at the different spatial and temporal scales are set out in Table 9.

323

The radiological health impacts – i.e. the number of potential fatal cancers, non-fatal cancers and severe hereditary effects in future generations – were estimated using the dose-response functions recommended by the ICRP (see EC 1995a – Vol. 5). 324 Non-radiological health impacts – i.e. number of deaths, working days lost and permanent disabilities for workers – were based on published statistics for the French nuclear sector (see EC 1995a – Vol. 5).

204

Chapter 3

Euro/MWh

Uncertainty rating

0.3

C

0.0008 – 0.35

C

Production

0.02

A

Other fuel cycle stages

0.08

A

Production

0.02

A

Other fuel cycle stages

0.11

A

Public health Collective dose Nuclear accidents Occupational health Industrial accidents

Collective dose (workers)

Greenhouse gases Low estimate

0

High estimate

0.13

0.1 – 3.7

Total

Ranges in economic impact valuation represent uncertainty of global warming impact (low and high estimate) and include an uncertainty interval (68% confidence level) Value reported for the nuclear fuel based on 0% discount rate over a time horizon of 10,000 years; damage costs of severe nuclear accidents are not included are generally reported separately (but see Section 4.1) Table 8. External costs of the nuclear fuel cycle in Belgium (without reprocessing) (Source: Torfs 2001, p.34)

Local

Regional

Global

0-100 km.

100-1000 km.

> 1000 km.

Short term

Non-radiological impacts

(.

A reconstruction of the policy development cycle in the case of the Belgian nuclear phase-out law

245

preservation of nuclear know how. No adequate assessment of socioeconomic costs of the phase-out decision.

Goal and strategy formulation

What should be the goals?

Prime goal is to phase out nuclear power.

Who is involved?

Parliament has voted on the proposal of law. Law can only be revoked by the Council of Ministers when security of energy supply is put at risk.

How to enhance co-operation?

Government depends on electricity generators (to implement technological alternatives), grid operator (to ensure sufficient exchange capacity) and regional governments (rational use of energy, permits for production units, etc.) for the effective implementation of the phase out.

How are uncertainties dealt with?

Period considered long enough (without further assessment) to develop alternatives or to ensure capacity for electricity import.

What is the link with other policy issues?

Relevant policy areas include climate and energy policy.

Implementation

Which policy instruments are most effective? What are side effects?

Ban on industrial generation of electricity based on nuclear fission. Limited information on possible side effects.

Monitoring

Are the policies really implemented?

No guarantee; very strong political signal could ensure a de facto implementation of the phase out. Dangers to security of energy supply are the only recognised form of a ‘force majeure’.

Is there room for new scientific insights?

Nuclear option is ‘kept open’; it is unclear which research lines will be favoured in the future. Government spending on national nuclear R&D remains high, but mainly in the areas of nuclear safety and waste management.

Are the goals likely to be met?

Monitoring of security of supply by the Commission for the Regulation of Electricity and Gas markets (CREG).

Evaluation

Table 10. Summary of the reasoning behind the Belgian nuclear phase out

246

Chapter 4

Apparently, the government was not willing to organise a large societal debate on the issue, as an emergency treatment of the proposal of law by the parliament was demanded393. In Table 10, an overview is presented of the policy development cycle, with the relevant policy questions and the corresponding answers taken from the above-mentioned official documents. Some selected issues from this table will be developed further in order to demonstrate how in this case the policy problem was transformed in a tractable technical form. 3.3.2

Recasting the problem in a technical mould

Technical or expert-based approaches to policy problems are supported by certain policy claims (cf. Chapter 2): • The problem can be circumscribed within the limits of one or a few academic disciplines (most notably engineering and economics); • Within these academic disciplines, there is a general consensus on the applicable methods (or new methods are developed within the disciplinary limits); • Experts are easily recognisable based on academic qualifications; • Participation by other actors (non-specialists) is very limited; • Transparency is very limited; • Decisions are preferably taken by one political body (which appears as a ‘monolithical block’ – e.g. ‘the legislator’); • The problem is tractable within the mandate of recognised bureaucratic institutions (the hierarchy of existing institutions is respected); • Furthermore, a technical approach can only function properly within the limits of given policy goals. In the following paragraphs, we discuss each of these bullets in turn. Let us start with the policy goals directly related to nuclear policy. First of all, it is remarkable that the real motives of the nuclear phase out are hardly discussed. No explication of the motives is found in the explanatory annex of the proposal of law (Parl. DOC 50 1910/001), and, as mentioned earlier, the announced explicatory note in the federal plan for sustainable development was most likely never published. It is clear that the legislator wants to phase out the existing nuclear power plants for electricity production, and that new power plants cannot be built394. This decision only applies to nuclear fission technology, not to nuclear fusion. It is only in response to one of the hearings organised in the course of the discussions on the proposal of law (and also in some interviews in newspapers), the state secretary specified that the phase out is justified because of:

393

In accordance with Article 80 of the Constitution. Possible future applications in the field of desalination of sea water or hydrogen production are not prohibited by the law, adding further to the confusion over the motives of the legislator. 394


247

• The risk of catastrophic accidents, which are admittedly very low; but, in the words of the state secretary, entirely unmanageable in a densely-populated country such as Belgium; • Proliferation risks, where the greatest area of concern is the fabrication of a so-called ‘dirty bomb’ (i.e. spreading radioactive materials with a conventional explosion); • The risks posed by the management of nuclear waste, and, most notably, irreducible uncertainties linked to the management of high-level waste over very large time spans. In chapter 2 we argued that, on account of its own self-understanding, goal-setting could not be discussed in a technical discourse on governance, as this task is left to the discretion of the political world. As a matter of fact, this is not entirely true. This approach may also be applied in goal-setting, but for this, the goals in question have to be transformed into ‘technical’ standards. The transformation of the concept of ‘uncertainty’ into a technical issue served as an excellent example here. In view of practical applications in the policy sphere, risk was defined as the probability of an adverse event multiplied by its (quantitative) consequences (e.g. number of casualties, number of injured, etc.). As a consequence, according to a certain (technical) interpretation, different risky activities become comparable on an equal basis. One might also be tempted to set technical standards to the societal acceptability of risky activities, e.g. a risk of 10-6 per year (i.e. one casualty in a million years of activity) is often proposed as a general ‘acceptable risk ceiling’ for society. With this brief reminder of our earlier arguments, we want to make the point that by interpreting the goals of the phase-out law as the state secretary did (i.e. catastrophic risk, proliferation risk, and long-term risks of high-level radioactive waste), these goals are clearly set beyond the limits of a technical risk discourse (thus asserting the exclusive powers of the sovereign to decide on these issues). The goals are said to be ‘ethical’, meaning, in this case, beyond (a strictly technical) discussion. After all, how can one reasonably calculate the risk that someone will construct a ‘dirty bomb’ in the years to come? And is it not true that the consequences of a serious reactor accident are nearly ‘incalculable’ (true, the probability is very low; so, in mathematical terms, the problem is the unstable mathematical case of ‘zero x infinity’)? But while the goals of the phase-out law are effectively placed beyond the limits of a technical debate, the means are not. And then, in light of reaching these goals (and the broader goals of energy policy, cf. infra), many questions can be raised regarding the efficiency and effectivity of phasing out nuclear power plants once they have reached a 40-year lifetime. For instance, why is the operational life span limited to 40 years? In the US, life extensions have already been granted by the USNRC to a number of nuclear power plants for an additional 20 years of operation beyond the current 40 year license (Hashemian 2002). In the Belgian context, detailed calculations (subject to the boundary condition of safety rules and radioprotection norms) would have to indicate whether this is a conceivable option from a socio-economic point of view. However, the opportunity to at least stage a debate on this issue was not taken, contrary to France for instance, where a report (Bataille and Birraux 2003) thoroughly analyses lifetime extensions and comes to quite favourable conclusions. Furthermore, regarding proliferation and safety, should we infer from the rationale of the

248

Chapter 4

phase-out law that the reigning international and national safety and proliferation regime (conventions, laws, etc.) are considered to be insufficient as a guarantee? If safety is the issue, will further research into reactors offering features of passive safety be encouraged? And will the reduction of risks in one area (i.e. catastrophic risk, proliferation risk, and long-term risks of high-level radioactive waste) not lead to an increase of risks in another? These are all reasonably conceivable questions in light of the above-mentioned goals which simply have not been addressed in the policy-making process. Concerning the broader goals of energy policy, it is fair to say that these can generally be summarised as guaranteeing a secure and reliable energy supply, with a high level of safety, with minimal impact on the environment, providing large opportunities for a qualitative employment and leading to acceptable prices for both industrial and private consumers. In the explanatory annex to the law, and in the parliamentary hearing, an attempt is made to show that all of these objectives can be reconciled with the objective of phasing out nuclear power. Only for the issue of energy security explicit (but nevertheless still non-binding) provisions were made: this has to be monitored by the CREG on a yearly basis (starting from 2015) and can be invoked by the government as a cause of ‘force majeure’395. Some interpretations of the law draw the attention toward this ‘force majeure’ clause, which is seen by some as an opportunity for future governments to escape from the consequences of the law. However, the state secretary for energy and sustainable development repeatedly argued that the phase-out law would be de facto irreversible, because potential investors in nuclear power plants would be scared off by the prospect of political insecurity: every time a ‘green’ political party would take part in government, it would most certainly advocate a nuclear phase out again. The argument is interesting, because it holds an implicit recognition by a member of government that the state apparatus (or parliament) can no longer muster enough ethical or moral consensus for its laws to be upheld; instead, it has to rely on jamming interventions in the economic sphere to be able to uphold its goals. Besides this, one of the issues which has drawn a great deal of attention is the compatibility of the phase out with the (post-)Kyoto commitments. Let us take this as just one example for the question of justification. In this discussion, the justification of the law relies on the CO2 emission results of a purely hypothetical and exemplary reduction scenario of electricity demand developed by the AMPERE Commission. In fact, this commission points out that between a high-growth scenario and a low-growth scenario, 17 Mton of CO2 could be avoided. But these are purely theoretical figures, i.e. there is no certainty that either the high or the low-growth scenario is more ‘real’ (actually, recent increases in electricity consumption were of the order of magnitude of 2-3% per year). Firstly, the explanatory annex gives no details about how this reduction of 3% per year could be reached. Secondly, it will already be necessary to advocate a lowgrowth scenario to be able to reach the (post-)Kyoto reduction with nuclear power still present. No matter how it is explained, an overnight nuclear phase out will result in an increase of CO2 emissions of about 16-17 Mton. This is not to say that government should

395

A juridical notion meaning that the law can be revoked by unforeseeable and compelling circumstances beyond the will of the party invoking the ‘force majeure’.


249

not take active measures to promote the rational use of energy pending more detailed analysis of the reduction potential396. We only point out that the justification given in the proposal of law is, at best, based on shaky foundations. The justification given in the policy documents quite frequently quotes the conclusions of the AMPERE Commission, and thus gives the appearance of relying heavily on expert opinion. However, it is often unclear whether this is really the case. For instance, when the explanatory annex to the law announces the government’s intention to ‘keep open the nuclear option’, it is not certain that this means quite the same as the AMPERE Commission’s recommendation to ‘keep open the nuclear option’397. The consequences of the phase-out law can also be dealt with within the mandates of existing institutions; no additional institutional structures need to be created. The mandate of the CREG is slightly changed, in the sense that starting from 2010, the commission is required to undertake a yearly review (instead of a three-yearly review) of the indicative investment planning (covering a period of 10 years) in view of ensuring enough replacement investment for the planned nuclear phase out as a guarantee for security of supply. Regarding broad participation, we already mentioned the fact that the government demanded an emergency treatment of the proposal of law, a fact which inspired some members of parliament to criticisms of ‘dogmatism’ and a ‘lack of respect for the parliament’ (Parl. DOC 50 1910/004, p. 10). This being said, participation in the drafting of the federal plan for sustainable development was encouraged (although it did not result in the promised explanatory note as mentioned earlier). Summing up, we believe it is fair to say that an attempt was made to recast the nuclear phase out as a technical or well-structured problem. But this could only be done by leaving a great amount of ‘white spots’ in the justifications given. The state secretary did refer to the scientific opinion expressed in the advisory process (thus showing a willingness to enlist ‘science’ as an aid), but in a rather ambiguous way (and, in the case of using scenario results discussed above, even bordering on manipulation). Most commentators on the law agree that it is very poorly argued: e.g. in the memorandum attached to the proposal of law it is not even mentioned why nuclear power is not consider to be a valid option any longer; the legislator merely asserts but gives no evidence that the phase out of nuclear power does not conflict with earlier engagements to reduce the amount of greenhouse gas emissions; it is not clear why the law is restricted to the ‘industrial production of electricity’; the

396

Such analysis was undertaken after the phase-out law had been voted – see Fraunhofer Institute (2003). E.g. the AMPERE Commission recommends participation in public and private R&D initiatives concerning future fuel cycle concepts, while the official position of the state secretary has always been that public funds for nuclear R&D (e.g. in the Belgian nuclear research centre) should only be used for research in nuclear safety and waste management issues. 397

250

Chapter 4

interpretation of the ‘force majeure’ clause remains unclear, etc.398 Parliamentary scrutiny did not significantly challenge the reasoning leading to the decision399. It appears then that the nuclear phase-out law is an example of what von Schomberg (2002) has called a ‘typically knee-jerk’ restrictive technology policy that is characterised by the selective and arbitrary perception (as opposed to a systematic perspective we intend to offer in chapter 5) of ethical and moral problems. This is not to say that political decision makers have been ‘wrong’ in choosing the reduction of the risks represented by a reliance on nuclear power as a political priority. However, by failing to frame these political priorities in a broad riskcost-benefit analysis (as advocated in chapter 3), the law risks to show a lack of ‘ethical robustness’ needed to secure long-term goals such as sustainable development400. In the remainder of the present chapter, we check this robustness against the viewpoints on a wide range of issues relevant to the question of nuclear power in a sustainable development perspective.

4 Interviews with members of the FRDO In our selection of respondents, we chose not to limit ourselves to the ‘traditional’ actors implied in energy policy matters. A number of problem-oriented criteria (representativeness, interest in long-term societal problems, insight in political decision making, distance from centres of power) led us to select members of the FRDO for this exercise. The selected people to be interviewed received a written invitation (by mail and e-mail), with a short memorandum containing the goal of the interview and a general outline of the type of questions to be asked. In the end, the people who chose to participate were representatives of the following organisations: labour unions (2), NGO’s (3), advisory bodies (3), public administration (1), universities (5)401, electricity generators (2), employers’ organisations (2), and business federations (1). However, all interviewees revealed their personal opinion (as opposed to the official view of their organisation). The interviews were conducted during the months of April and May, 2002, before the treatment

398

For a juridical-institutional commentary, see Neuray (2003) and Michiels (2003); for a more technically and economically inspired critique, see D’haeseleer (2003). Our constructivist approach has been published in Laes et al. (2002a, 2004a). 399 Barbé (2005), the former chief of cabinet of the state secretary for energy and sustainable development, disagrees, stating that the phase out has been thoroughly discussed and that the record of the discussions in the parliamentary commission should be considered as a ‘reference document’ on the subject. If such thorough discussion was truly the intention of the then government, it would be hard to explain why at first an attempt was made to introduce a prohibition on new licenses for nuclear power plants in an implementing order of the Law of 29 April 1999 concerning the organisation of the electricity market (Michiels 2003), thus effectively bypassing the ‘cumbersome’ process of parliamentary debate... 400 We are not claiming here that governments should completely ignore acceptance of technology by the population. However, basing policy decisions solely on acceptance (again, we do not accuse the government of having done so – we merely intend to point out some of the dangers of ‘governance by acceptance’ as a limiting case which will never be realised fully in reality) renders the decisions made particularly vulnerable to possibly rapidly changing rates of acceptance, anti-innovative impacts (the danger of being too precautionary already in early stages of technological developments) and possible loss of trust in a government that no longer seeks to define the ‘common good’. 401 Representing expertise in the field of law, human ecology, energy economics & policy, climate science and environmental sciences.


251

of the proposal of law by parliament. This precision is necessary with respect to the changing salience of some policy issues. As there exists no standard methodology for the analysis of semi-structured interviews, we had to develop our analysis grid in function of the interview results. We have done this through the definition of a set of categories. With respect to content, each category defines a cluster of arguments revealed by comparing and contrasting different responses. A single category is determined by the responses to a number of questions in the interview grid: indeed, as the interview grid was designed to facilitate the interaction between the interviewer and the participant, relevant information could not be extracted directly from any given question. The categories relate to different aspects considered to be relevant in the context of governance for sustainable energy, e.g. motivations for market liberalisation, policy instruments for market regulation (e.g. energy taxes, subsidies, etc.), long-term perspectives on the Belgian energy sector, etc402. More details on specific questions of methodology of case-study research can be found in Annex 1. Before going into the details of the interview results, we wish to draw the attention to some general observations403.

4.1

Observations

The interrogated persons often accepted to present a personal point of view and dared to express views that do not necessarily correspond with the views defended in public. Consequently, it was important to guarantee the confidentiality of the resulting information. Neither the name of individuals nor organisations will be used in this chapter. With regard to the information obtained, it is interesting to note that the participants did not answer in the same way to the same questions, e.g. some participants were able to go into much greater detail on some questions. Therefore, it is impossible to make a synthesis through a representation of the ‘average result’. The opinions were asymmetric to the extent that they cannot be compared on some particular issues, either because each participant had a very different domain of competence, or a desire to emphasise one precise factor above others. A sociological analysis of the interviews shows that the different participants did not share a common language, common interests or a common vision. The capacity to expand on some information sources (e.g. the AMPERE report) or certain topics was often very divergent. For instance, it appears that the involvement of important social groups in the management of high-level waste has been minimal for the time being. 402

It is however obvious that the interview approach cannot guarantee absolute causal correspondence between the actors’ statements and their actual decision-making behaviour. Our emphasis on the interpretation and reconstruction of argumentation chains also means that other (typical) facets of policy analysis, most notably social network mapping of ‘key players’ in energy policy through ‘reputational’ (identifying actors who are perceived to be influential in energy policy issues), ‘positional’ (identifying actors who have energy-related issues as a mission statement) or ‘decisional’ (identifying actors who have actually participated in decision making on energy policy issues) approaches (Jegen and Wüstenhagen 2001) – though important – was not undertaken in the context of this dissertation. 403 These observations were for a large part made by my colleague Michel Bovy, who was present in some interview sessions as an observer.

252

Chapter 4

Participants were very hesitant to express opinions, and then only in the broadest of terms. For these reasons, we have chosen to leave this topic out of the interview results, which does not imply that it is not important in the overall discussion on the sustainability of the nuclear option (as mentioned several times earlier in this text). It is difficult to reveal a shared rationality – rather we faced different juxtaposed strategic visions under the umbrella of ‘sustainable development’. There is no ‘metalanguage’ and the interview grid does not produce such a language, but rather a common approach to gather information. The different positions are as well based on political arguments as on scientific argumentation. In fact, the participants were only brought together by these interviews, and not by a common point of view as a result of each participant’s involvement in the FRDO. The questions raised during the interviews are a result of the choices made by us within the very broad framework of sustainable development. As such, they do not necessarily represent the full opinion of the participants on the topic of nuclear energy in a sustainable development perspective or on strategic guidelines for sustainable development in general. Thus, we saw it our task to develop certain perspectives further based on a literature search. Also, we must keep in mind that energy, and a fortiori nuclear energy, is not necessarily at the heart of the sustainable development discussion for every involved actor – a plethora of other issues can be imagined, such as poverty, social security, ageing of the population, food and health risks, etc. – and this observation also reflects on the resources that each actor is prepared to spend on the development of a point of view on the issue. But as was phrased by Smits and Leyten (1991, p. 25) in their seminal historic analysis of technology assessment: …The positions and interests of groups which have few opportunities of their own to produce scientific foundations, can be sufficiently interesting and important to explore and develop more fully (our translation)…

Participants were often forced to occupy a double position during the course of the interviews. On the one hand, as experts, they referred to scientific argumentation they master very well. On the other, they expressed views as a layperson or based on a political conviction. Scientific knowledge is not necessarily decisive, neither is the political vision; they seem to be mutually reinforcing. This situation was often paradoxical and confusing for participants. As a scientific expert, they have to be prudent, reserved, and clearly delineate the limits of their scientific knowledge. They have to show the uncertainties and are reluctant to take position in more complex or value-laden general questions. As a layperson or civilian, they take position on a political level and express more ‘easy’ ideas – i.e. without wanting to base them on very precise scientific facts. Several participants, mainly representatives from the scientific world, pointed at this difficulty during the course of the interviews, and in fact made it very clear when they expressed views as ‘experts’ or as ‘civilians’, often in the course of the same answer. As is explained in the theoretical part of this dissertation (Chapter 1), this ‘forced’ reference to ‘easy’ ideas, ‘ideology’, ‘visions’,


253

however fragmentary they might be, was entirely intentional, as reference to them can simply not be avoided.

4.2

4.2.1

Interview results

Economic issues

4.2.1.1 Liberalisation of the electricity market The liberalisation of electricity markets is recognised by almost every interview participant as one of the most important driving forces behind the restructuring of the energy sector. As such, liberalisation and its relation to sustainable energy and nuclear power occupied a central position in a large part of the argumentations collected during the interviews. Liberalisation of the electricity market was generally welcomed, albeit for different reasons and providing certain conditions are met. For example, representatives of environmental NGO’s pointed out that the ‘real’ costs of different electricity supply options will at last become transparent, without governments intervening in favour of one technology over another. In any case, according to this point of view, the opening of the energy markets offers the prospect of fundamental change – a prospect that appeared to be very dim indeed during decades of government intervention. For a number of reasons (cf. infra), environmental NGO’s prefer that the future energy (electricity) system would be based on a configuration where local production for local needs is assured by the operators, selling energy services (i.e. heating, lighting, etc.) instead of just energy vectors (i.e. electricity, fuels, etc.). Moreover, they show a strong preference for small to medium-sized operators, i.e. that the means of production are owned by a local community, municipality or town, with citizens as shareholders. Energy ‘production’ and distribution companies would have to assure the planning, installation, management and maintenance of apparatus for decentralised energy ‘production’. Such a solution could prove economical for operators in the energy sector because it allows them to engage in client-based activities (allowing them to ‘make the difference’ with other operators in the open market), it opens up new markets and it could allow a better utilisation of local energy potentials. Representatives of environmental NGO’s see this configuration as a goal that must be pursued to a maximum degree in its own right, therefore justifying an urgent need to phase out nuclear power as the archetype of centralised production, or generally, justifying higher transition risks. They believe it is possible to turn the creed of market economics against complex large technical systems such as nuclear power, the reasoning being based on the prohibitively high initial investment costs and the lack of flexibility in answering the more quickly fluctuating electricity demand in a liberalised market. From a different perspective (voiced by other participants), such configuration was seen as complementary to the more traditional centralised production (including renewable energy sources). In any case, the argument is that the ongoing liberalisation of the energy

254

Chapter 4

markets creates a climate of uncertainty. The future structure of energy systems is then perceived to be very clouded. One would have to wait for some years until the energy markets have stabilised in order to enable a more clear vision on the future. Liberalisation is generally welcomed as an essential precondition for sustainable development, the argument being based on a faith in the functioning of free markets (e.g. eliminating economic inefficiencies) where costs reflect also the ecological and social consequences of development. Of course, while this rhetoric picture is often used as a general truism, in particular cases new economic instruments (e.g. the introduction of an energy/CO2 tax) are opposed for as long as a ‘level playing field’ on a European (global?) level cannot be guaranteed. On the other hand, there are some common worries (mainly expressed by representatives of labour unions and environmental NGO’s) about the opening of energy markets. They emphasise that the system of energy provision should remain socially equitable: energy, being an essential good in modern society, should remain accessible to everyone at a reasonable price, even when the price-capping system (in the protected market) is abolished. Care should be taken not to replace the equitable sharing of power under the old rules with a ‘big dogs eat first’ principle. Utilities404 should be submitted to public responsibilities. The attention for vulnerable social classes is perceived to be at danger in an open market, and should therefore receive appropriate attention. Large electricity consumers should not be put at an advantage at the cost of smaller consumers (i.e. the problem of cross-subsidies). Governments should create a climate for investment in the Belgian energy sector, especially when nuclear power stations are effectively closed down: the danger is real that Belgium will then become increasingly dependent on imported electricity. One interviewee added one more danger of liberalisation to this list: the danger of a ‘new medievalism’, meaning that gigantic energy or electricity companies might bind the small consumers to their products (energy services) through clever marketing, in fact rendering the idea of a ‘free consumer choice’ of producer a very deceptive rhetoric instrument405. 4.2.1.2 Structure and size of the Belgian economy The structure and size of the Belgian economy is of course a determining factor in the future level of energy demand406. We focus here on the role of energy-intensive industrial activity. Most participants pointed out that the siting of energy-intensive industrial sectors on the Belgian territory has been the result of an active policy over the past decades (e.g. one participant mentioned the expansion of the Antwerp harbour). Belgium occupies a 404

Who use the term here in a generic sense. In principle we should distinguish between generators, suppliers, transmission grid operators and distribution grid operators. 405 During the interview, one of the participants handed over a leaflet entitled “De strijd om de klant in de ‘vrije’ markt : Neo-liberalisme of neo-feodaliteit ? (‘The battle for the customer in the ‘free’ market: Neoliberalism or neo-medievalism ?’, our translation)” with quotes taken from various sources, underlining the dangers of the formation cartels, ‘binding’ of clients, etc. 406 For results of group interviews with Belgian stakeholders (i.e. representatives from industry, service sector, residential buildings, intermediaries, transportation, consumer groups and NGO’s) going more specifically into the question of energy demand management, see Fraunhofer Institute (2003, pp. 197-220).


255

central geographical position in Europe and is therefore extremely well fit as a supply centre for raw materials for other European markets. From this point of view, this important strategic advantage has to be preserved in the future, because of economic but also ecological reasons (i.e. the argument of ‘delocalisation’: if the production site would be moved to other countries with less stringent environmental standards, the global consequences would be worse). This point of view reflects the commonplace that, especially for these energy-intensive industrial sectors, policy should undertake action to ensure that energy costs are kept sufficiently low (e.g. by not unilaterally introducing too many new taxes), in order to be able to compete in global markets. For these energyintensive sectors (e.g. chemicals, steel, paper, cement, etc.) it is in their own best interest to reduce the costs of energy use; therefore, it was quite generally acknowledged that energy consumption in these industries is already at a very competitive level and cannot decrease significantly. Emission reduction targets should never endanger economic growth. However, one participant questioned the concentration of industrial activity. While he accepted that integrated industrial sites offer the possibility of efficient energy consumption (e.g. through integration of high and low temperature industrial processes), the increasing need for transportation could easily offset this advantage. In any case, he said, a global assessment should be made, not only from the point of view of one production activity. Such assessment should be based on a life cycle approach for particular products (i.e. a calculation of the energy consumed in the complete production cycle – i.e. from the exploitation of raw materials to the waste products – including transport requirements between different steps in the cycle). For now, the question remains open. Among the interviewees there was a large consensus that more efficiency gains could be attained in the less energy-intensive economical sectors. Transportation was considered to be a very difficult target sector for curbing greenhouse gas emissions through policy intervention, because of the still growing volume of road transport which for the time being largely surpasses any efficiency gains that will be made. Also, structural changes (e.g. switching from road to rail transport) were considered to be very hard to reach, especially on short notice. 4.2.2

Technology

4.2.2.1 Energy ‘production’ technology From one point of view (mainly represented by representatives of employers’ organisations and electricity utilities), the potential of technological options should be judged on their inherent technical and economical characteristics (e.g. the potential for wind energy is limited in Belgium because of a lack of available space, difficulty of integration of fluctuating power sources in the European network, etc.). From this perspective, technology in itself can never be ‘evil’ or ‘good’, it all depends on how technology is used by society (in fact rejecting any notion about the political character of technology – cf. infra). Great faith is placed in technological developments as the key to sustainable development: references were made to the hydrogen economy, the future potential of

256

Chapter 4

photovoltaic energy, and even (at the far horizon of expectation) fusion power. In general, from this point of view it is only acceptable to limit the use of certain technologies if: • its application threatens public health or safety – based on the best available expert risk assessments; • its use threatens to exhaust resources (thus endangering economic growth); • it degrades the quality of the environment – again, based on the best available expert assessment methods; • it introduces very significant social or psychological stresses in society. From this perspective, nuclear power can only be rejected based on the last criterion: evidently, democratic rules must be obeyed, and the decision to phase out nuclear power shows that there exists enough popular support for this decision. If society does not want to use nuclear technology, then so be it: “…the consumer is always right…”. Other interviewees showed faith in new technology, but were somewhat more reluctant towards a reliance on radical innovations and breakthroughs in science and technology. The opinion was often expressed that technological development is a gradual, almost evolutionary process that must be guided by government intervention (norms, standards, etc.) preferably. Therefore, from this point of view, it is not advised to give up known principles and solutions without having strong indications that alternatives are already feasible. It is therefore considered to be wise to investigate a broad technology portfolio in the face of uncertainty. For now, this means that investments should be guided towards the development of renewable energy sources (e.g. wind energy, solar energy), new technologies (fuel cells) and more attention should be given to demand side management (a historically underestimated option). This also includes a continued participation in international nuclear research programmes. A third point of view (mainly defended by representatives of environmental NGO’s and some ecology-oriented scientists) holds that generic technological options should be guided by internationally agreed ethical principles. Technology cannot be thought of as existing as some isolated sphere of human activity; on the contrary, technology is a law imposing activity, in the sense that it guides and limits our possibilities of human interaction. This observation implies a particular interpretation of the precautionary principle for the appraisal of technological options. Since we cannot fully appreciate the complex consequences of the introduction of any technology in society, it should be fault-tolerant to a maximal degree. Therefore, the decision to implement a technology should be as reversible as possible. It is clear that under this perspective, the decision to build nuclear power plants should never have been taken in the first place. One participant described this decision as “…a blank cheque : for years, money has been invested in this technology, and even now, the exact costs of high-level waste management and decommissioning of existing reactors are not yet fully known…”. Furthermore, technological ‘lock-in’ should be avoided. A ‘lock-in’ happens when one technology shapes its surroundings in order to mirror the perfect conditions for its continued use. For example, nuclear power generation


257

is especially effective when it is operated in base load; therefore, energy demand should be high enough at all time. According to this perspective, another aspect of this ‘lock-in’ phenomenon is that a significant share of renewable energy in the electricity generation mix is made impossible: nuclear power plants have an insufficient capacity to follow the rapidly changing load of the electricity network. From this point of view, nuclear power tends to ‘crowd out’ other options. 4.2.3

Governance

4.2.3.1 EU policy / international agreements A first perspective (mainly defended by employers’ organisations and utilities) strongly believes in company dynamics and less in management by government. This perspective is also less interested in the strictly Belgian context: European and even global economic trends have to be taken into account. The uncertainties associated with foreign developments must be taken into account, e.g. the withdrawal of the US. from the Kyoto protocol is perceived as a major threat to national industry. Interviewees stressed that they have nothing against policy intervention on environmental grounds (they can even induce a competitive advantage), but these must be set on a European or global level in order to protect domestic industries against disadvantages in international competition. Since energy is the motor of the economy, EU policy should principally strive to ensure free competition and security of energy supply. Other actors (e.g. labour unions) more strongly insisted on internationally binding agreements. Integration of development objectives should be carried out at the highest level possible. The Kyoto protocol (and eventual follow-up protocols) is considered to be the most important driving force for the transition to a sustainable energy system. A preference for clearly formulated objectives and standards (typically end-effect indicators concerning the environment and human health) was also often expressed in the interviews. From this point of view, the economic sector should not be given too much freedom of movement; covenants should only be used where strict rules do not work. EU policy should principally strive to ensure free competition, security of energy supply and a limitation of greenhouse gas emissions. Representatives of environmental NGO’s added to this a more locally-oriented perspective, whilst of course also remaining concerned with the ‘outside world’. They generally strived for global agreements, placing a lot of importance on principles of equitable distribution, precaution, etc. The general aim seemed to be a global ‘limits to growth’ policy, e.g. by trying to introduce stringent greenhouse gas emission limits for each country. It is acknowledged that the only possible way towards this objective is to make reliable agreements amongst all industrial countries, even though interviewees expressed doubts that this can be achieved. One participant spoke of the necessity of a global taxation scheme for energy-intensive products. He also pointed out that in the field of renewable energy, a closer co-operation within the EU would be needed, as it is almost

258

Chapter 4

impossible for a densely-populated country like Belgium to fulfil its energy needs (even under the assumption of a restrained energy demand) with inland renewable energy ‘production’ capacity. A European trade system of green certificates could be envisaged under binding targets for renewable energy ‘production’. Security of energy supply would then result from an increased emphasis on energy conservation, energy efficiency measures and promotion of green energy. 4.2.3.2 Effectiveness of policy measures In the interviews, we focussed on what the participants saw as a realistic potential for policy measures to limits the level of energy demand. While this is of course a very difficult question as such, we were more interested here in the general heuristics and ‘mental models’ used. Globally, everyone agreed that the most important percentage reductions of energy demand could be achieved in the domestic and tertiary sectors and in some industrial activities. The following picture can be drawn. While a first perspective (expressed mainly by representatives of employers’ organisation and utilities) is in favour of at least a limitation on overconsumption, a great scepticism exists as to the possibilities to actually decrease the demand for energy. The rise in energy demand must be compared to other countries with a similar economic structure. “…No other industrialised country has so far been able to voluntarily realise economic growth without an increasing demand for energy; why should Belgium, with its very energy-intensive industry, be able to accomplish this...” was an oft-heard objection during the interviews. Furthermore, governments will most likely not be willing to impose energy conservation schemes or less popular taxes on their voters; therefore, the decision to use energy will be a private decision. This perspective generally prefers policies that are considered to not need a lot of political debate, although sometimes a certain level of extra funding is acceptable: targeted and general information schemes, voluntary agreements, education, labelling and energy audits all belong to the policy mix. Interviewees often express confidence in self-regulation and a belief in the power of the market, e.g. one participant stated that “…a quick rotation of capital in a very competitive market will ensure that older, less efficient installations will be replaced by new and highly efficient ones...” Effectiveness should be regarded from the point of view of business. Other opinions could however also be heard. A representative statement for this group would be that “...energy demand must decrease over a long period of time to ensure a smooth transition towards a sustainable energy system…”. The reasoning is that, inevitably, other protocols will follow up the Kyoto protocol with more stringent reduction demands on greenhouse gas emissions. While the Kyoto norms can perhaps still be reached with improved efficiency schemes or fuel switching (e.g. between coal and gas in the electricity generation sector), further reductions will only be conceivable through an effective limitation on the absolute level of energy demand. Therefore, more stringent government action has to be initiated now on top of the more ‘soft’ measures; examples given were energy efficiency standards for appliances, use of financing instruments


259

(income-neutral energy taxes with exemptions for energy-intensive industries, loans, subsidies for lower income classes to invest in clean energy), etc. From a third point of view (mainly defended by environmental NGO’s), energy policy should cover a much broader domain than the traditional ‘carrot and stick’ measures. From this perspective it is asserted that in fact, energy policy decisions are made in a lot of different sectors of society, e.g. in housing policies, spatial planning arrangements, and principally, strategies of utilities or manufacturers of appliances, etc407. While this adds considerably to the complexity of the picture, it also opens up new possibilities for action. Reasoning from integrated social and technological concepts such as the ‘low-energy building’, this perspective tries to steer the market actors by mandatory energy efficiency standards (e.g. isolation standards for new buildings, etc.), prohibitions of (publicity for) electric thermal storage systems, differentiated energy taxes, consideration of energy demand in new spatial planning schemes, etc. Spatial planning is an especially important theme, as it determines our way of life and ultimately, our way of energy consumption for decades to come. Such questions of infrastructure were considered to be insufficiently addressed in the past. 4.2.3.3 Nuclear power: an ‘inherently undemocratic technology’? This question, formulated as an assertion by two of our participants, seems to occupy a central position in the reasoning of representatives from environmental NGO’s408. Of course, the argument is not new: in the ‘70s, at the height of the energy crisis, when fundamental choices had to be made about the future options for energy supply in the US, a heated debate took place between what was called by Lovins (1977, p. 6) – considered by some as the pope of green energy thought – “…two energy paths that are distinguished by their antithetical social implications…”. It is no surprise that this argument has survived until today, as it involves deep-rooted preferences about how society should be organised – preferences that are not easily shaken by empirical or practical observations. But, in our view, this line of reasoning has possibly gained a renewed vigour through the contemporary sustainability debate, as we will try to illustrate. Firstly, the technological potential for decentralised energy ‘production’ that Lovins already in 1977 claimed to be the cheapest in direct economic evaluation (which was not the case at that time) seems to be more promising (especially when external costs are taken into account)409; secondly, new advances in communication and information technology improve the chances of a more decentralised organisation of technological systems; and thirdly, the political component of the sustainability discourse (see Agenda 21 – Section 3)

407

One participant gave the historical example of the promotion of electrical thermal storage systems to promote the use of cheap nightly base load power, in order to better integrate the large nuclear capacity in the electricity system. 408 For a good summary of this argumentation, see also Knapen (2002). 409 “…Surprisingly, a heroic decision (meaning a decision that is based on social desirability) does not seem to be necessary in this case, because the energy system that seems socially more attractive is also cheaper and easier…” (Lovins 1977, p. 7).

260

Chapter 4

in fact advances this ‘democracy from the base’, ‘participation of stakeholders in decision making’ or ‘empowerment of important societal actors’. This last observation seems particularly important, since the ‘question of democracy’ is now recognised and legitimised at the level of global policy making (albeit on a principal basis), and this at least offers increased opportunities for actors to phrase their democratic concerns without concealing them as hidden agendas in a policy discussion that seem to be about efficiency, productivity, economic growth, etc. Upon further analysis, the argument of nuclear power being an undemocratic technology seems to be subdivided into four separate lines of reasoning: • Firstly, there is the observation that nuclear power plants, being very capital intensive, require a sufficiently large form of human organisation to be run effectively. The opening of the energy markets compounds this danger (from this perspective): energy companies might evolve, via mergers and acquisitions, into technological institutions of truly gigantic proportions – a ‘state within the state’. The amount of influence and power these institutions hold over governments tends to increase proportionally to their size; • Secondly, the very nature of nuclear power generation implies a control from a very limited number of power centres, i.e. nuclear activities fall under a specific legislative regime, controlled at the national level, involving a very limited number of organisations (e.g. for nuclear safety: the ‘federal agency for nuclear control’ (FANC), the authorised inspection organisation (AVN), the operator’s health physics department and, in the case of major modifications, the ‘scientific council’ (composed of representatives of relevant administrations and experts)). The question is whether such arrangements allow enough checks and balances, in the vein of traditional democratic principles – a question, which, as some interviewees observed, is difficult to answer, precisely because of the often confidential nature of the interaction between these organisations; • Thirdly, some interviewees pointed out the various ways that these large socio-technical organisations exercise power to control the social and political influences that are supposed to control them. One participant even spoke of “…the necessity to control thought patterns in the population…”, meaning that since the stakes for energy companies are so high, and given the fact that even a minor fault or oversight can cause a controversy, these companies have to resort to e.g. psychologically sophisticated communication techniques or plain silence to influence the public opinion; • Fourthly, there is the argument that the ‘undemocratic nature’ of nuclear power somehow spills over into society at large. One participant stated, not without some sense of the hyperbole, that “…nuclear power is typical of regimes such as China, the Soviet Union, etc. …”. If one looks beyond the rhetoric however, a crucial concern can be discerned, namely that central governments should not be too strongly involved in the choice and steering of technological options, as happened in the ‘70s, when the decision to build the current nuclear power plants was made. From this perspective it is clear that nuclear power plants can only operate in a very stable environment (i.e. by necessity involving government intervention to guarantee this stability): hence the ‘closed’ licensing procedures, the system of ‘price-caps’ in a protected market, and ultimately,


261

the need to guarantee a high level of electricity demand for a sufficiently long period of time410. Comparisons with other countries which have opened up their electricity markets learn that there is an element of truth to the above analysis. Roberts (1990) points out that in the UK, the institutional changes proposed by liberalisation both fragmented and alienated political support for the nuclear industry. Based on a historical analysis, she concludes that a ‘fully developed corporatist policy sector’, its authority unchallenged by other systems of representation, will usually be a necessary condition for a nuclear programme411. On a similar note, Weyer (2005, p. 20) points out the (historical) ‘structural affinity’ between strongly hierarchical political and large-scale technico-industrial systems – an affinity which, according to Weyer, has become ‘precarious’ in view of the general decline of public trust in technology and politics. Other interviewees shared some of these concerns, but were more focussed on the way nuclear power is managed by society nowadays. From this point of view, there are some fundamental equity problems: • The insurance obligation for nuclear power operators is limited – in fact, this is the equivalent of a subsidy to nuclear power; • Even if the provisions for decommissioning and radioactive waste management were sufficiently large, the government would still have to ensure that these provisions would actually be available at the time when they are needed. These funds should be controlled independently; • Electricity producers have enjoyed the benefits of a quick depreciation of the nuclear power plants (originally, a 20-year lifetime was estimated). The consumers have paid for this quick depreciation. So, from the equity point of view, these ‘stranded benefits’ (benefits in a protected market system) would have to be used for the ‘common good’, i.e. to cover ‘stranded costs’ (costs that are incurred because of decisions made in the past in a protected market system), or to finance a demand side strategy. These ‘stranded costs’ should not be paid for by society at large while the electricity producers only enjoy the benefits. According to one participant: “…If the nuclear sector wants to contribute to sustainable development, these issues have to be dealt with first…”. 4.2.3.4 The phase-out decision Representatives of employers’ organisations and utilities did not favour this decision at this time. To be sure, the decision to close down nuclear power plants after 40 years of operation will not influence current investment decisions in the energy supply sector, nor will safety practices be altered – these will be monitored according to the same stringent

410 For a historical discussion of Belgian energy policy, see Verbruggen (1986), Laes et al. (2004c) and Barbé (2005). 411 Historical research on the interplay between the success factors for nuclear power programmes and political institutions or culture mostly focusses on country-studies, while comparative research remains underdeveloped (to our knowledge).

262

Chapter 4

quality criteria (stressed by the representative of the major generator). However, from this perspective, the phase out decision does have an indirect effect: young people will not be stimulated to take up training in nuclear engineering, putting at risk the knowledge base for the future. What will happen when in a distant future, nuclear power is again considered to be a valuable option? Furthermore, large industrial consumers of electricity fear that this decision undermines the stability of the electricity prices that they have enjoyed for decades and the security of energy supply might be put at danger; in particular, fear exists that energy use will become very much dependent on imported natural gas412. Furthermore, the compatibility of the nuclear phase out and the (post-)Kyoto targets is questioned. All these observations contribute to the overall judgment that the phase-out decision is “…purely symbolical…”. Representatives of environmental NGO’s of course welcomed this decision, although some criticism could be aimed at the arbitrary nature of the 40-year lifetime. In any case, sooner or later nuclear power is bound to be cast aside as a valuable option for sustainable energy provision. A clear decision about the operating lifetime then has the advantage of sending a sufficiently strong message to possible investors in the energy supply sector to favour alternative forms of energy ‘production’. Moreover, the Belgian decision is not a stand-alone case; in fact, Belgium joins a group of European countries (Germany, Sweden, Austria, Italy, etc.) that have either decided to phase out or have refused to adopt nuclear power production in the first place. One might classify this point of view as a ‘technologyforcing’ argument: nuclear power is perceived as the embodiment of a number of factors (cf. supra) that tend to restrict the possibilities for renewable and decentralised energy options. That way, the innovation potential and creativity of the industry will be maximally drawn upon. Nuclear power will, according to this perspective, be phased out anyway: the technology is simply not compatible with liberalised market conditions. Therefore, it is better to strongly commit to alternative sources now. Other participants (mainly representatives of labour unions) express somewhat ambiguous opinions. While they share with the environmental NGO’s some common concerns about nuclear power as a socio-technical system, they are also suspicious of any decision involving abrupt changes. There is no immediate need to phase out nuclear power; rather, this perspective likes to rely on government to steer nuclear power and the electricity market in general into a more equitable direction. In the words of one of the participants: “…In the future, the feasibility of the phase out will depend on the social environment at the time: if the phase-out decision entails big losses for important social actors, then the debate will no doubt be polarised, and a lot of time and money will be lost. The phase out must be supported by society at large, and this already supposes that an open debate has been staged – for the time being, there are too many ‘grey areas’…”.

412

Belgium of course is also dependent on imported uranium for its nuclear power plants. However, these reserves are geographically spread (USA, Australia, Canada, Europe,...) and the fuel costs represent only a small part of the operational costs of a nuclear power plant.


263

4.2.3.5 The role of government On this topic, we are interested in the following question: what is, for the different actors, a good institutional arrangement – i.e. an arrangement that, in their view, furthers the cause of sustainable energy? In general, the following perspectives could be discerned in the interview results. A first group of respondents (mainly representatives of electricity generators, employers’ organisations and some scientists) is mostly interested in achieving accountable decisions, based on ‘objective’ data, thus combining an internal focus (a small group should decide) combined with an emphasis on stability once the political decision has been taken. Thus, they showed a clear preference for expert-based governance schemes (the possible shortcomings of which we have analysed in chapter 2 – Section 3.1). Correspondingly, it was hoped that the conclusions of the AMPERE report would be adhered to when discussing the future of nuclear power in parliament. The fact that politicians do not take up the responsibility for their decisions (since the implications are only felt in a far future) is perceived as a major problem. Hence, the proposal of law to phase out nuclear power was often referred to by this group as a ‘purely political’ (i.e. emotional, irrational) decision. This group did not care very much for public participation in decisions concerning fundamental energy choices for the future: “…technology hardly ever is at stake in the elections, people will generally not be motivated to participate in policy exercises…”. A different group (mainly representatives of labour unions) was more interested in achieving legitimate decisions, i.e. enough time should be reserved for the decision makers to collect the advice of the different social partners. It was not clear whether also the public should be included in the decision: of course it is useful to know public opinion, but there are some practical reservations (large investment of time and money, perhaps with limited results). The general suggestion was that it should suffice to know the opinion of the ‘traditional’ social partners, after all, they were considered to be representative of a large part of the population. This group believes in an open debate about the advantages and the costs of different alternative scenarios for the future413. The general sentiment was that there is no immediate need to phase out nuclear power; rather, government should reserve for itself the possibility to steer nuclear power and the electricity market in general into a more environmentally and socially sustainable direction (inspired by the perceived dangers of the ongoing liberalisation of electricity markets). Government should also provide for itself the means to intervene in the organisation of the electricity market. One interviewee pointed out the danger of institutional weakness: “…Because the electricity market soon will become liberalised, the government can no longer refuse to grant a license to run a power plant based on the argument that there is not enough electricity demand; and, at the same time, equipment plans become merely indicative, so government has bereft itself of all possibilities to influence the shaping of the electricity sector…”.

413 Questions considered to be relevant were e.g. ‘What will be the costs of high-level waste management if we continue with nuclear energy?’, or, ‘What will be the costs of decommissioning?’, and ‘Can these costs be met in a context of increasing competition in liberalised energy markets?’.

264

Chapter 4

There was also a plea for a better integration of the regional and federal policies; again, this can be achieved through consensus on important goals for policy (e.g. the national climate action plan). The appeal to a greater involvement of stakeholders should of course be evaluated carefully. It seems that this perspectives favours some form of governance through pacification (e.g. a renewal of the ‘old’ consultation mechanisms under protected market conditions), the possible shortcomings of which have been analysed in chapter 2 (Section 3.3). Although stakeholders may pursue their own specific interests, their activities can only be tolerated as long as they are perceived as being compatible with the general interests and values of society. By these remarks, we already anticipate a third perspective to the problem. This third group (mainly representatives of environmental NGO’s and some ecology and humanities-oriented scientists) holds that central government should decide on the basis of (strictly interpreted) internationally agreed upon principles, in order to identify desirable options for the future. These principles include transparency, strict liability, the precautionary principle, the polluter pays, etc414. Substantial savings in energy use in industrialised countries such as Belgium is seen as a moral obligation towards developing countries. Within the limits of these principles, they strongly support the involvement of citizens on a more local scale (cf. infra). 4.2.3.6 The science-policy interface As explained in section 3.1, the AMPERE report presents itself as a synthesis of existing (mainly technical and economical) knowledge relevant for future electricity policy. A first group of participants (particularly representatives of employers’ organisations and some scientists) implicitly or explicitly endorsed the methodology followed by the AMPERE commission415. Conversely, this methodology was heavily criticised by environmentalist groups. While several participants noted a general lack of attention for the demand side of the energy equation or for an integrated view on energy issues, only one of these groups (an environmentalist NGO) effectively attempted to engage in the cognitive debate by ordering a critical and systematic review of the AMPERE report by the Wuppertal Institute

414

The reliance on nuclear power is considered to be an infraction of many of these principles, e.g. liability for the catastrophic accident consequences is limited, generation of HLW is considered to be ethically unacceptable towards future generations, the nuclear power sector is perceived as being little transparent, ... 415 The participant showed his or her general agreement with the results and recommendations of the AMPERE commission, either without having made the effort to critically review the methodology, or after a thorough review. Some participants handed over leaflets with official statements, showing their familiarity with the argumentation as developed by the AMPERE commission.


265

(Thomas et al. 2001)416. By doing so, an attempt was made to transfer the political debate even to the most fundamental level of peer-reviewed ‘objective’ science, thereby illustrating Beck’s (1992) analysis of sub-political action. Table 11 summarises the methodological differences between both approaches.

AMPERE Commission

1st step

2nd step

3rd step

Wuppertal Institute

Collect and compare existing projections of electricity demand.

Conduct a demand-side bottom-up sectoral analysis of energy demand for different enduses.

Conduct a sensitivity analysis with regard to CO2-emissions.

Derive different scenarios for a range of demand-side energy efficiency potentials, taking into account barriers that could prevent the full implementation of the identified potential, and policy measures to address these barriers.

Analyse how the projection of electricity demand can be met with possible options for electricity generation.

Estimate supply side potential.

Focus on social cost (technical + external) and plausible potential.

Analyse share of potential that can be realised through different policy instruments. Calculate total primary energy consumption, emissions and costs for different scenarios.

Recommendations on future technologies, taking into account policy orientations (e.g. nuclear phase-out, promotion of renewables)

Table 11. Methodological differences in the approaches used by the AMPERE Commission and the Wuppertal Institute

The authors of the Wuppertal review maintain that in order to generate a reliable factual basis for informed decisions on the future energy policy in Belgium, a “…policy and discourse oriented integrated demand and supply side bottom-up scenario analysis…” (p. 5) should be developed. In this definition, every word is revealing of a certain perspective on the utility of scientific knowledge for decision making. Compared to the AMPERE methodology, this approach puts an emphasis on the modelling of concrete policy measures that can help to quantify the magnitude of the need for action (e.g. CO2 emission goals for each sector) or identify specific areas were more action would be needed. While the AMPERE approach mainly concentrates on an educational discourse about technological options for power generation, the methodology advocated by the Wuppertal Institute aims

416 In this paragraph, we only present a brief and schematic account of this methodology in order to highlight the different accents; for a complete representation we refer to the original publication. We wish to point out however that the Wuppertal report based its conclusions on partial information as the main report of the AMPERE commission is not available in English; only the ‘Executive Summary & Conclusions’, ‘Recommendations’ as well as only part of the ‘Synthesis Report’ has been translated. It is therefore possible that the authors of this report have been unable to grasp the full contents of the AMPERE reasoning on some points.

266

Chapter 4

at facilitating a discourse on the level of political planning by reserving a large role for an expert/stakeholder dialogue417. Regarding the demand-side bottom-up aspect of the analysis, the international peer review group commissioned to review the AMPERE results (Bourdeau et al. 2000) also notes a certain lack of attention to issues of demand-side management in the AMPERE report (in spite of the fact that an implementation of an active DSM policy is one of the main recommendations of the report)418. They make essentially the same observations as the Wuppertal Institute419. However, these observations are immediately qualified by adding that the criticism is more directed at the mandate of the AMPERE commission than at the way it fulfilled that mandate (p. 4) which originally asked to concentrate mainly on the supply side. Furthermore, the restriction to electricity instead of energy in general is a good example. The ‘official’ peer review group thus comes to a more generous overall appraisal of the AMPERE endeavour. If anything, the above discussion shows that there is no agreement on the so-called ‘objectivity’ of science, and that this disagreement is not mainly the result of a poor understanding of science, but rather of different criteria to assess the usefulness of scientific methods within a policy context. In making certain methodological choices, in stressing some aspects of the problem (e.g. technological supply options) while downplaying others (e.g. demand-side measures), researchers make inherently political and value-oriented choices. The methodological confusion heightens the danger of a strategic use of scientific information by social actors, i.e. results will only be used when they are in accordance with the interests of the user. In the Belgian debate on the future of nuclear power, we noted several instances of this phenomenon. The review by the Wuppertal

417

This educational vocation most clearly comes to the fore when the AMPERE commission defines the problem with nuclear energy as the public’s irrational fear and lack of understanding, and suggests that more information campaigns might be the solution (Executive summary, pp. 80-81). 418 AMPERE declares not to have studied whether it was economically more interesting to invest in demand compared to supply (in power plants) to reach the Kyoto objective (Synthesis report, p. 41). The peer review group observes that “...this is indeed the key question, and therefore the interest of Section C (concerning future power demand and DSM) for the decision process remains limited…” (Bourdeau et al., p. 14). Furthermore, the realistic potential for demand-side management estimated by AMPERE (8 %) is considered to be “…very subjective…” (p.15). 419 Bourdeau et al.(2000, p.16) “…The exploration of future power generation systems should be placed in the perspective of the whole Belgian energy system, notably by examining contrasted scenarios corresponding to different scales of end-use energy efficiency policies and programs. This would lead to the highlighting of scenarios in which the whole of the future energy system – production and consumption – would have been geared to respond to the constraints of energy supply security and the global environment…”, and, “…Using in particular the examples of DSM policies and programs developed in other countries and quoted in some length by the AMPERE report, detailed alternative scenarios of future electricity demand in Belgium could be established. This would provide a precise evaluation of the potentials, both technical and economic, for the various sectors of social and economic activities and for the different electricity uses…”(pp. 16-17).


267

Institute (ordered by an environmental NGO) explicitly states that (Thomas et al. 2001, p. 5)420. ... because of severe methodological problems, it is not possible to justify the recommendations for the future electricity generation options given by the AMPERE commission based on the commission’s own results. The reason is that both the methods used to derive these results, and the insufficiency of the results themselves, do not allow the recommendations to be justified…

Conversely, groups whose views largely concur with the AMPERE recommendations have no interest in a critical examination of the methodology. As discussed before, even the government’s explanatory memorandum to the phase-out law shows a very selective reading of the AMPERE report. In a way, it can be argued that the peer review process has to some extent identified gaps in the existing knowledge base, and thus fulfilled its role in a satisfactory way according to the precautionary principle. Nonetheless, the AMPERE report is considered by some as an incomplete exercise: closer and more frequent communication between the research level and the political level might have revealed shortcomings in the mandate of the commission (most notably in the restriction of the assessment to the electricity sector) and the methods used (most notably in the assessment of the demand side). 4.2.3.7 Citizen participation in decision making Citizen participation in (technological) decision making is often advanced as a panacea to the problems of public trust in democratic institutions. It turned out that our respondents showed a rather more pessimistic view. For a first group (mainly environmental NGO’s and some ecology-oriented scientists), citizen participation in decision- making is one of the cornerstones of sustainable development. They see the amount of support for a given policy as the determining factor in the success of that policy. One participant added that “…this means that every policy measure should enjoy the active and conscious support of at least fifty percent of the population…”. The suitability of the representative democratic model as a reference framework to fundamentally alter patterns of development was sometimes implicitly questioned. From this perspective, ecological problems are fundamentally about economic or technological system dynamics that have evolved separately from the day-to-day life

420

Correspondingly, a press release by the same environmental NGO was entitled: “Rapport Commissie AMPERE maat voor niets!” (‘Report of the AMPERE commission completely useless’, our translation). Glorieux (2002, p. 13) – a member of the green party – is of the same opinion: “…the AMPERE commission has not sufficiently taken into account the supplementary tasks set out by state secretary Deleuze, in agreement with the federal policy agreement, namely that options for demand side management should be studied and recommended in order to realise the nuclear phase out whilst respecting the Kyoto commitments. The independent evaluation done by two independent peer review groups also points out the singular focus on the supply side and the inadequate methodological approach of the AMPERE Commission. As a consequence, the final report of the AMPERE Commission does not constitute a valuable foundation for the future choices in energy policy…”. D’haeseleer (2003, p. 71) – member of the AMPERE Commission – retorts that the review by the Wuppertal Institute was “…very incomplete and very much ideologically colored…”.

268

Chapter 4

world of ordinary citizens. In principle then, this diagnosis leads to only one solution: a ‘dialogical form of democracy’ (see e.g. Holemans 1999). Public spaces must be created where debates concerning the ‘common good’ can be initiated without having to rely on technocratic decision making as the default option. This group of interviewees realised that this broadly defined strategy will be very difficult to realise, not in the least part because people nowadays are often not interested in societal questions and are often not willing to take up responsibilities. “…The link between the act (e.g. consumption leading to domestic waste) and the consequences (e.g. burning of domestic waste in incinerators) is not recognised; all responsibility is delegated to the governing bodies or society at large…”. Still according to this point of view, a decisive factor will be the scale of technological systems. Small-scale technologies will be an enabling factor for this ‘local and basic democracy’. One participant gave the example of Denmark, where a large percentage of the taxes are levied by local authorities; thus, the spending of tax money will be much more transparent and interested citizens will enjoy a much larger degree of control. He referred to “…the current practice of financing projects from an enormous government budget so that the total costs to society remain obscure…”. Similarly, small-scale energy ‘production’ technologies (solar panels, wind energy, etc.), owned by the consumers themselves, or local communities, will make energy use and its effects more transparent and accessible for debate at a local level. Such participation is perceived to be difficult to achieve in a national and strategic discussion, the nuclear power question being an example par excellence. It turns out that involving citizens in an abstract, strategic problem concerning the whole of society is very difficult, because an important element for participation is clarity concerning what is at stake. Overall, people are inclined to become involved in decision-making issues only when they think that the issue is in their immediate interest. While some participants agreed with public participation on issues of national importance in principle, a lot of objections to public participation in general were also phrased (by other participants than the first group): • Public participation is no guarantee for the best solution to a problem: short term solutions could be emphasised; • Public participation is time consuming and very costly to implement; • Decision making at the strategic level requires knowledge, a certain expert view. Only experts, government officials and interest groups have the knowledge, time and energy to engage in this kind of decision making; • Public participation can be misused as an instrument to quiet down social conflict without really envisaging solutions; • Public spaces for discussion will only be monopolised again by interest groups; • People may well act as consumers in a participatory process and not in the interest of the ‘common good’. Furthermore, future generations can never participate, a principal stance by a higher authority is needed to take their interests into account; • Another important aspect to take into account is the cultural environment. In Belgium, there is no culture of direct public involvement (e.g. in the environmental impact


269

assessment procedure, no public involvement is foreseen in the early scoping phase). A public debate is too often conceived of as “...the presentation of a plan (on a website, in town halls) for the public to react to on short notice…”. We must not forget however, that our enquiry was directed at the stakeholders (in a large sense) in the debate. These results thus only indicate that citizen participation to decision making will most likely not contribute to a greater level of trust in the political decision and between stakeholders from the viewpoint of the stakeholders themselves. Nevertheless, many of the above objections are shared by the scientific trust-related literature (EC 1999b; OECD 2002). 4.2.3.8 Role of new consultation channels (e.g. FRDO) In this section, we are interested in what the participants saw as the potential of ‘new’ advisory councils to reflect on and influence policy initiatives for sustainable energy421. A number of common observations can be made: • None of the participants saw the FRDO as the primary channel for their views. The CREG is seen as the most important council to influence in a modest way the structure of the future electricity sector, although some participants expressed their concern on the ‘purely legal’ discourse that is practised by the council. It seems that European directives leave not much room for interpretation or for debate on the ‘common good’; • Other channels mentioned where: direct lobbying, information sessions for political parties, brochures, and specific campaigns to influence public opinion or policy makers, etc.; • All the participants welcomed the discussions in the FRDO as an opportunity to freely exchange and discuss ideas. But some participants questioned the representativeness of other groups. And, as somebody remarked: “…The closer this council will be to the centre of decision making, the more dysfunctional it will become because of the conflicts of interest…”; • When energy questions are debated by the FRDO, they often concern the more technical aspects of energy policy. Even then, a consensus is hardly ever reached. Energy is a much contested topic; • The FRDO was also widely seen as being too entangled in short-term policy questions422. Since nobody can really oppose the elaboration of a long-term policy vision as a basis for an informed debate, one nagging question remains: why have these long-term scenarios not been developed for the Belgian context? It turns out that obstacles were

421

I.e. councils that are specifically appointed to advise on themes of general importance to society, in view of sustainable development, involving all relevant societal groups, as defined in Agenda 21 – as opposed to more traditional social deliberation in thematic councils. 422 At the time when the interviews were organised. Over the last few years, already more long-term or strategically oriented recommendations have been issued by the FRDO, e.g. on a national strategy for sustainable development (27 May 2005); on long-term strategies for combatting climate change beyond 2012 (26 November 2004); and on the vertical integration of sustainable development in multi-level governance (18 December 2003).

270

Chapter 4

often very opaque and had both cultural and institutional roots. One participant stated that “…policy makers mostly direct questions at councils concerning the short term…” and that “…these councils often lack time and means to think about long-term developments…”. Several participants noticed the lack of structures within which the parties involved can discuss and consult one another on the long-term strategic choices for society. Furthermore, “…the government’s interest in long-term visions is often purely rhetorical…” and there is a pervading sense of pragmatic conservatism, in fact arguing that “…Belgium is too small to effectively realise a long-term vision, we have to undergo European or even global tendencies…” and that “…Belgium is framed in a corset of geopolitical conditions…”. The circle is thus closed: long-term perspectives are not developed because this is made irrelevant by ‘the surrounding environment’, and ‘the environment’ cannot be influenced because long-term visions are not developed; • Some participants argued that society should not rely too much on councils but should rather encourage more direct forms of democracy.

4.2.3.9 Role of environmental NGO’s Some interviewees expressed a rather negative appreciation of ‘green’ NGO’s: they are experts at communication rather than experts in the technicalities of the debate and they further their interests through a discourse about symbols rather than facts. The nearly impossible reconciliation of the (post-)Kyoto commitments and the nuclear phase out in this perspective forms a good illustration: the environmentalist discourse is said to play on the public fear of catastrophes, rendering any rational discussion impossible. Other interviewees however encouraged environmental NGO’s to develop their point of view, in fact, several participants pointed out that without the pressure of environmentalist movements, there would not have been any discussion about sustainable development at all. One participant added that “…Often, these groups have proven to be right by history…”. From a democratic point of view, it is good to represent as many voices as possible. However, some interviewees expressed reservations about the representativeness of these NGO’s. Several participants also pointed out that over the years, representatives of green NGO’s have become experts in specific policy domains. Representatives of the environmentalist NGO’s pointed out the essential role (or pressure groups in general) in the agenda setting of society. While they welcome any opportunity to participate in different councils (e.g. the FRDO, the CREG, etc.), public hearings, advisory boards, etc. their principal and most effective aim should still be to influence the public and the political opinion (through ‘visible’ actions) in order to gain a real leverage on policy making. However, to this, one must add that membership of environmental NGO’s is lower in Belgium compared to other countries (e.g. Germany). Nuclear issues are followed up by only a few people.


271

4.2.3.10 Role of ‘business’ Representatives of employers’ organisations and utilities saw an important role for business in managing the transition towards sustainable development, reasoning along the lines of ‘sustainable entrepreneurship’. A lot of thought and energy is invested in this subject; some participants handed over information brochures intended for a broad audience. Sustainable development is discussed at the highest levels of corporate management. Selfregulation is considered to be very important: in anticipation of legally enforceable standards, the first mover gains the advantage423. The core concepts for industry seem to be424: ethical entrepreneurship, corporate citizenship425, life cycle analysis426, and responsible care427. All of them could lead to the so-called win-win situations. Through these concepts, an attempt is made to ensure good relationships with all of the interested parties, that is all those who have something ‘at stake’ in relation to the functioning and performance of the organisation. Employees, shareholders, other investors, suppliers, customers, neighbours (who may suffer from pollution) are amongst the interest groups normally identified. Electricity generators operating nuclear power plants are no exception: from this perspective, there is no reason why the rules of sustainable entrepreneurship could not be applied. Other interviewees however stressed the role of government in ensuring that the advantages of technological achievements, in casu energy, would be spread equally amongst the population. These people were much more attentive for some of the ‘hidden’ variables of the model of ‘sustainable entrepreneurship’: for instance, little is said about differences in power between different actors, or about actors that are not represented at all (future generations!). The stakeholder model tends to emphasise local and short-term issues. 4.2.3.11 Role of experts Some interviewees (cf. supra) fully endorsed any attempt to ‘objectify’ the reflection on energy options for the future based on a pluridisciplinary scientific analysis by experts

423An

article in the Financieel Economische Tijd (May 30, 2002), entitled “Wereldwijd steeds meer duurzaamheidsjaarverslagen (‘More and more sustainability reports worldwide’, our translation)” states that companies increasingly pay attention to their responsibility towards society (employees, pensioners, but also human rights, child labour, etc.) as a result of NGO activity, new stock indexes (e.g. the Dow Jones Sustainability Index), ethical or sustainable investment funds, and communication on a global scale. The article also observes a tendency for governments to increasingly withdraw from the direct control of company activities, while these activities seem to be guided now by the scrutiny of the public opinion. 424 Derived from leaflets handed over during the interview sessions. 425 Defining entrepreneurship as an “integral part of society, whereby people have access to and control over the means to make informed choices in order to contribute to a more human, sympathetic and just society” (Fedichem 2000). 426 Defined as “taking into account the whole production process, from raw material to the ultimate disposal, to evaluate the environmental impact” (Fedichem 2000). 427 Defined as “…a worldwide voluntary initiative whereby companies commit themselves to continuously improve performances in the field of safety, the environment and human health…” (Fedichem 2000). See also Espejo et al. (1998) for a discussion of corporate social responsibility and its role in sustainable development.

272

Chapter 4

from different universities, as was performed by the AMPERE commission. Furthermore, it was recommended that the conclusions of such an analysis should be distributed as widely as possible – e.g. to different decision makers (including political parties, trade unions, European Commission, etc.), the media, NGO’s, or even through inclusion in relevant educational initiatives (e.g. at universities) – the reasoning being that this ‘diffusion’ will automatically enhance the impact of the conclusions reached by the expert. From the second perspective (mainly represented by labour unions and some scientists), the role of experts is seen as somewhat more problematic. This group also likes to rely on a sound scientific basis for knowing effects and side effects of energy options. From this perspective, the interface between science and decision making is considered to be crucial: “…a good functioning of expert groups is essential for the democratic functioning of a country…”. Nevertheless, we collected the following propositions for a better organisation of scientific expertise in society: • Transparency of expertise should be clearly defined: experts should clearly highlight agreement and differences of facts and values, and should encourage relevant societal actors to pass judgement on at least the values involved in the issue. Important values include employment targets, social acceptability of nuclear energy, and protection of small consumers in an open energy market and energy independence; • Clear rules about expert consensus positions should also be defined: can expert commissions formulate minority points alongside the comments most experts agree upon? If a consensus is desirable, how should this be achieved – e.g. one of the participants gave the example of the ‘International Panel on Climate Change’ (IPCC) procedures, where draft texts are sent to all scientists in order to ensure that there is a real consensus amongst them; • Experts should also try to popularise their knowledge on important societal issues such as the role of nuclear power in a sustainable energy perspective, e.g. by the large-scale distribution of a clear summary of their findings. This was also considered to be a weak point in the procedure followed by AMPERE428. From a third perspective (mainly represented by environmental NGO’s), the role of experts, and certainly experts who enter a public debate where social polarisation prevails, is considered to be problematic. It may be useful to outline the possible reasons we collected during the interviews. First of all, some interviewees were very sceptical about the claimed objectivity of scientific expertise: scientists are either perceived to be beholden, directly or indirectly, to government or industry interests; or, which is more frequently the case, they are alleged to adhere to scientific paradigms, focussing on narrow technical problems and promoting the belief that scientific knowledge is separate from its

428

The commission organised only one open hearing for stakeholders (24 June 2000, Brussels).


273

applications429. Often, experts have gained their degree and their experience within one particular field of expertise (e.g. nuclear engineering) and have developed a certain ‘professional pride’, which has become firmly entrenched in their value system and which makes it harder to respond to criticisms from other perspectives. This fact makes it easier to predict which position a particular expert will take, or which information he or she will include in a scientific argument, especially in a polarised debate. Appointing experts in a scientific commission thus becomes a very precarious undertaking. To be clear, adherents of this perspective add that experts do not conceive their role in these terms; it is not the integrity of experts that is questioned. It seems that experts sincerely believe in the causes they support (“…all persons believe in their own virtue…”). Neither would this group argue that science should play no role in policy making. Rather, at the core of the arguments lies a dissatisfaction with the way scientific expertise has been brought in the policy process up till now. Of course, if the personal affiliation of an expert is taken as a sign of possible bias, it becomes very hard to find e.g. a trustworthy nuclear expert (since most will be employed by the nuclear sector or have some previous professional experience in that field). For the moment, we merely indicate the problem; seeking possible solutions will be postponed till chapter 5. 4.2.4

Socio-cultural reflections

4.2.4.1 Behavioural patterns Most interviewees agreed that changes in energy consumption behaviour can only be realised by targeting individuals as consumers430. This view seems to be inspired by a form of pessimism regarding voluntaristic actions by people to save energy in their daily activities. This observation is paralleled then by a search for key policy ‘buttons’ which can be pressed to steer peoples’ choices in a particular direction. While some interviewees expressed a preference for ‘soft’ instruments – e.g. only trying to influence consumers through information and publicity – others stressed the need for ‘correcting’ consumer behaviour, but still based on the rationality of individuals defined as consumers (e.g. through taxes, incentives, etc.). In short, one of the areas in which policy is expected to

429

Although now no longer relevant for investment decisions in the energy sector (because of the liberalisation of energy markets), it is instructive to take a look at the last national equipment plan for the electricity and gas sector 1995-2005, which was drafted by the BCEO. This organisation united both producers (Electrabel/SPE) and the distribution sector (80% Electrabel/municipalities and 20% pure intermunicipalities or distribution companies). Formally, it was then submitted to the government which can make some comments. So in fact the estimation of the realistic potential to reduce the electricity demand is coming from the electricity sector itself. This ‘realistic’ estimate was used by the AMPERE commission (Synthesis report, pp. 49-50). However, as stated before, the empirical (cognitive) question of a ‘reasonably’ attainable reduction in electricity demand is not our main point of contention. The point is that the procedure used to arrive at this figure does not provide policy-relevant knowledge on the necessary and priority measures to be implemented to attain this level of reduction of electricity demand. The demand-side of the energy equation is ‘bracketed’, and hence, any potential for policy learning in that area is foregone. 430 I.e. an individual that makes his or her choices based on primarily economic criteria (price and quality of the product).

274

Chapter 4

exert some influence, is in encouraging consumers to opt for ‘good’ lifestyles rather than ‘bad’ lifestyles. If consumer demand can be galvanised in these ways, the assumption is that other features, including actions of electricity and gas suppliers, manufacturers, and equipment suppliers will fall into place. Broadly speaking, consumers can exert their influence in two arenas: the market and politics. Consumer influence in the market makes use of ‘consumer power’ – i.e. a large group of consumers make the same environmentally friendly choices at the same time, which gradually eliminates the environmentally unfriendly products and creates a demand for a better alternative. In time, some interviewees maintained, consumer pressure will almost certainly lead to low-carbon (and generally less polluting) solutions to the energy problem: fuel cells, solar panels, etc. Some even were of the opinion that this will happen regardless of whether the Kyoto protocol will come into effect or not. Other participants (mainly representatives of environmental NGO’s), whilst sharing the same kind of pessimism regarding voluntary actions – e.g. one representative of an environmental NGO pointed out the difficulties of mobilising ‘civic passions’ on a longterm and abstract issue such as global warming – nevertheless showed less willingness to merely resign themselves to this state of affairs. As explained under section 4.2.3.7 (on ‘citizen participation in decision making’), the problem is compounded by the fact that the central state apparatus is seen to be responsible (at least in part) for creating this situation in the first place. This perspective is in a certain sense very ambiguous. It stresses the opposition between the social rationality of participation in a community of energy producers and consumers (the ‘virtual power plant’ – an interconnection of many individual small-scale electricity users and producers at the community level, supported by larger, nation-wide grids for emergency back-up) and energy policy guided by the state or pure market approaches. The socially rational energy community is most of all an ideal picture that must be aspired to in a far-off future, because at present, the necessary conditions for the realisation of this ideal picture are not considered to be fulfilled. This observation then produces the peculiar point of view that in order to close the gap between the present and the future, strong elements of the policy as rule are preferred. In order to promote participation in the future, participation in the present should be limited. Thus, governments need to exclude nuclear power based on strong ethical grounds, and rational use of energy and renewable energy must be promoted by making full use of a range of policy options open to policy makers under conditions of liberalised markets (e.g. efficiency standards for appliances, public R&D spending, energy taxation, internalisation of external costs, etc.). By combining a nuclear phase out with international obligations to lower greenhouse gas emissions (or generally, tightening the limits of ‘environmental space’), the only conceivable policy option lies in substantially improving the rational use of energy (through a combination of efficiency and sufficiency measures), together with a promotion of renewable energy sources.


275

4.2.4.2 Consumption patterns From one point of view (mainly represented by employers’ organisations and utilities), present and historic trends in energy consumption are indicative of what to expect for the future. The very high growth rates in energy consumption experienced in the past are generally not considered to be desirable (or realistic) any more, but, in any case, the demand for energy should almost be regarded as a given or a ‘law of nature’. To support this point of view, one interviewee pointed at a pile of studies on demand side management made over the years, and contrasted this ‘academic’ approach with the very limited impact on actual consumer behaviour in the ‘real world’. As levels of affluence continue to rise, people will also demand an increasing number of energy services, e.g. air-conditioning, new appliances, etc. Quality of service and price will remain determining factors in consumer choice. Policy makers will have difficulty to take any unpopular measures. From this perspective, energy demand can almost exclusively be limited through an improved end-use efficiency. Other interviewees on the other hand felt that government should more strongly intervene to manage the demand side. “…In the past, few incentives were given to limit overconsumption, and one could not expect that electricity producers would take up this responsibility as it was not in their best interest…”. However, it is still believed that a focus on the production side will achieve a much better result. A third perspective however (mainly represented by environmentalist NGO’s and ecology-oriented scientists) maintains that the current level of energy demand should not be taken as a yardstick. Current levels of demand are clouded by the distinction between individual and collective consumption. In the historic system of (quasi-)collective consumption, the link between the decision-making act and the benefits and the inconveniences were uncoupled. Politicians and powerful interest groups decide on the electricity production technology, consumers enjoy the benefits and taxpayers (or society at large) suffer the inconveniences. From this point of view, this systematic uncoupling is considered to be highly detrimental. The so-called ‘natural’ need for energy is in reality the result of a complex interaction between economic, technical, cultural and political choices. Hence, one should strive for an integrated approach targeted at all these levels at the same time.

4.3

Discussion

Arguably, energy is one of the most important links between the economy, social development and the environment. Of course, fuel and electricity are only secondary goods in the economy and their consumption is derived from a more fundamental demand for energy services. This simple fact was widely acknowledged in the interviews; however, the full implications with respect to sustainable energy governance were often found difficult to apprehend. Interviewees put the different pieces of the policy ‘puzzle’ together in different ways, or they lacked or omitted certain pieces for completing the

276

Chapter 4

puzzle. Moreover, it was not even evident that all participants were in fact making the same puzzle: • Generally, process-related questions carried the greatest weight in the responses. A lot of respondents pointed at the lack of attention for the long-term implications of policymaking, institutional barriers (e.g. political short-sightedness) or more specifically, organisational constraints (lack of time, lack of people, and lack of budget) to the development of visions on the long-term future. Very often, participants were preoccupied by near term problems431. This observation is particularly important since the presumed role of nuclear power in a future sustainable energy system begs for longterm visions; • As a result, a second level of discussion (substantial questions concerning e.g. technological development or expected levels of energy demand) was generally underrepresented in the responses. Participants generally felt that they did not have enough information to develop their vision and expressed uncertainty about technological developments, or even that it is not very useful to think about the long-term future in a national context, since Belgium increasingly will have to ‘undergo’ decisions made on a European or even global level. Thus, future oriented visions were difficult to derive from the interview results. Rather, participants focussed on contemporary problems; • While the interview logic was set up according to a certain conceptual scheme providing us with a logical and causal structure, it became clear in the course of the interviews that participants often identified transversal links between these hierarchically ordered levels. For example, a lot of participants brought up the issue that a particular configuration of the production park (high levels of centralisation) also reflected on the powers of electricity generators to influence the democratic debate (i.e. integration of politics in technology) or that certain scientific research paths were not chosen because of a reigning scientific paradigm (i.e. integration of power in knowledge). Nevertheless, for purposes of clarity, we made an attempt to reconstruct the different arguments used in the interviews into some coherent and consistent argumentation schemes. These argumentation schemes thus differ from each other in the assessment of different aspects of the sustainable energy policy question and in the resulting will to change the course of development. They are meant simply as frameworks for analysis and thus, essentially, as ‘ideal’ reconstructions. This implies that although participants will certainly recognise parts of their reasoning, they are not to be identified with the vision of a particular societal actor. We tried to construct the most robust argumentation scheme possible for each scheme. This analytic approach is only meant to guide the process of reflection, by drawing particular attention to some aspects of the problem and by systematically positioning the collective choices between different options against each other. The arguments revealed in the interviews have been conveniently summarised in a summary table at the end of this chapter. We have labelled the three perspectives the

431

‘Near term’ should be understood relative to the slow dynamics of energy systems. compliance with the Kyoto targets (2008 – 2012) is considered to be a near term challenge.

For instance,


277

‘management’, ‘controllist’ and ‘reformist’ perspective. These perspectives have been reconstructed in their structural dimensions, i.e. their communicated images of self and others (with respect to relevant actors), valid forms of communication, main problem focus, and main principal references. Boltanski and Thévenot’s commonwealth model (cf. chapter 1) have served as an aid in identifying them. The manager for a large part frames his arguments within the confines of the industrial and market commonwealth. He sees economic growth and technological advance as the most important component of sustainable development, to the extent that actions that might seriously endanger possibilities of growth or competitiveness in general must be discouraged. Furthermore, economic growth will most likely be driven by higher demands for energy and electricity. The manager is quite content for the market for electric power to stand as a surrogate for societal consent. To be sure, he is of course worried about safety and health issues; however, these are considered to be part of a technical design. It is the government’s responsibility to ensure standards and norms, based on ‘objective’ scientific rationality. Hence, sustainable development is at risk when the necessary long-term stability is undermined by a lack of (respect for) expert knowledge as an indisputable basis for the legitimacy of state action. Governments should set up a stable framework; business will then take up its responsibilities through ‘sustainable entrepreneurship’, ensuring relationships based on trust and consent with concerned parties (labour unions, stockholders, employees, local residents, etc.). There is no reason why electricity generators owning nuclear power plants could not be part of this. The controllist seems to be caught in a paradox. His position on the role of nuclear power in sustainable development was perhaps best phrased by one participant : “…As long as there is no real commitment to the development of a vision on long-term alternatives for nuclear power, a phase-out scenario is nonsense. But, if society does not want to consider the phase out of nuclear energy, the motivation to think about alternatives will also be very weak…”. The controllist is not so much interested in the ‘pro or contra’ discussion about nuclear power; rather, attention should be given to the institutional embedding of this technology is society. Fear exists that in the future, nuclear power will be ‘inevitable’ if one wants to respect post-Kyoto commitments and still foster economic growth. Rather, acceptance (or rejection) should be based on a democratic debate with the representatives of concerned parties, under conditions of full transparency. For now, these conditions have not been fulfilled, too much is left in the dark: costs of decommissioning, costs of high-level waste management, the real costs of the business-as-usual scenario, etc. – all ‘great unknowns’. In other words, the controllist is mainly concerned with the maintenance of the democratic system of ‘checks and balances’. Thus, more attention is given to the necessary framing of ‘industrial’ and ‘market’ values within the logic of the ‘civic commonwealth’. The controllist prefers a real balance between economic, social and environmental development – for now, economic growth is too strongly favoured. This perspective deplores the risk that in a liberalised market, the government’s power of intervention could be limited.

278

Chapter 4

The reformist sees the evolution of the Belgian electricity sector as an ongoing social process in which scientific knowledge, technological innovation (or the lack thereof in renewable energy technologies) and corporate profit reinforce each other in deeply entrenched patterns, patterns that, according to this perspective, bear the unmistakable stamp of political and economic power. In terms of Boltanski and Thévenot’s commonwealth model, people and objects are artificially kept in a state of permanent ‘misery’: perfectly valid technical options (e.g. renewable energy options) are underdeveloped and ‘rational’ behaviour (e.g. energy saving) discouraged (i.e. ‘misery’ in the industrial commonwealth), the true costs of energy use are being concealed (i.e. ‘misery’ in the market commonwealth), and people are kept in a state of political apathy (i.e. ‘misery’ in the civic commonwealth). For the reformist, nuclear power is not merely the symbol of this social order; it is a true embodiment of that order. The concerns are broad and directed at ethical and socio-cultural levels for which even regulatory environments might not be suited. Moreover, this perspective challenges and stretches the limits of the established commonwealths towards long-term and global ethical considerations432. The reformist’s explicit agenda calls for a new social order that would make the current distribution of resources more equitable. Resources must be understood in the broadest sense: not only in a physical (e.g. distribution of health and environmental risks) or monetary sense (e.g. distribution of benefits from nuclear power generation), but also culturally, involving democratic and governance issues. Consent for a technological or development option must be based on explicitly revealed preference in a dialogical form of democracy. Small-scale participatory institutions are regarded with more trust than central government. The reformist also feels that, as a result of this socio-technological nexus centred on nuclear power, his perspective on sustainable energy has not been addressed sufficiently and calls for a new research agenda: there is no culture of long-term reflection, there are no sufficient scientific data to perform a bottom-up analysis of electricity demand, energy issues in general are not high on the political agenda, etc.

5 Summary and conclusions In this chapter, we have made a reconstruction of the policy development cycle in the case of the Belgian decision to gradually phase out nuclear power. We have described the policy development process in terms of (social) positions, interactions, dynamics, and their political-cultural background. Our main conclusion for this chapter has been that the justification given for the phase-out decision was based on an attempt to recast the policy problem in a well-structured technical mould. This was evident from its selfproclaimed reliance on expert opinion, limitations on the possibilities for ethical debate, treatment of the policy question within the mandate of existing bureaucratic organisations, etc. A detailed analysis showed however that this ‘technical’ treatment could only be achieved by leaving some ‘white spots’ and/or ambiguities in the justifications given. This

432

The problems of extending Boltanski and Thévenot’s commonwealth model to include the interests of future generations and ‘non-human others’ have already been discussed in chapter 1.


279

finding was further strenghtened by the empirical part of this chapter (i.e. interviews with members of the FRDO), which shows that sustainable energy (and the role of nuclear power therein) is an essentially contested concept, and furthermore that there were virtually no ‘connecting’ or ‘translating’ links between the divergent concept and problem framings. This finding suggested that other possible views on the role of nuclear power in a sustainable development perspective existed which had to be actively ‘suppressed’ or ‘blurred’ (in order to proceed ‘as if’ there was a consensus). The interviews revealed mainly three different types of perspective. The manager for a large part frames his arguments within the confines of the industrial and market commonwealth. He sees economic growth and technological advance as the most important component of sustainable development, to the extent that actions that might seriously endanger possibilities of growth or competitiveness in general must be discouraged. The controllist on the other hand believes more attention should be given to the institutional embedding of technology in society. The reformist perspective sees the evolution of the Belgian electricity sector as an ongoing social process in which scientific knowledge, technological innovation (or the lack thereof in renewable energy technologies) and corporate profit reinforce each other in deeply entrenched patterns, patterns that, according to this perspective, bear the unmistakable stamp of political and economic power. As argued before, initiating a transition towards a more sustainable development model implies at least some form of common problem acknowledgement (i.e. a common understanding of the problem situation) and, a fortiori, a common understanding of what should be done. Moreover, such understanding should be relatively stable over time. Both requirements imply some measure of social learning and communication between the principal actors involved. Summarising the general picture however, it becomes apparent that conflict rather than mutual exchange was the dominant dynamic in the debate surrounding the phase-out decision. Exclusive relations between the different perspectives were caused by competing rationalities on the one hand and the governance framework on the other. Our analysis revealed that social learning was mainly hindered by the following issues: • Different methodological approaches (bottom-up vs. top-down analysis of energy system); • Lack of data (to perform the bottom-up analysis)433; • Different perceptions of relevant time scales (or how to link short-term issues with long-term issues); • Different framing of the problem (electricity vs. energy system); • Institutional barriers (e.g. to develop the needed long-term vision);

433

This was not so much revealed during the interviews, but see Fraunhofer Institute (2003) for an in-depth discussion on the availability of energy demand indicators.

280

Chapter 4

• Lack of communication (between political decision makers and scientists, between scientists and stakeholders); • Strategic use of scientific assessments by different stakeholders, or • Insufficient knowledge of scientific assessments. We believe the problem can indeed be qualified as ‘unstructured’ and could benefit from a much larger degree of social learning in the distinct phases of the policy development cycle. It is our contention that a profound structural change such as the nuclear phase out would require a sustained leadership defending clear-cut goals and procedures for the production of scientific knowledge and the interaction process with stakeholders with respect to sustainable development. Only under these conditions can a sufficient level of trust be built up between the involved players for a concerted action. If not, the discussion runs the risk of getting bogged down in endlessly repetitious arguing, as appears to be the case for the time being.

Summary Table Categories Economy

Crucial problem dimensions / Criteria for assessing development Opening of electricity market

‘Management’ perspective

‘Controllist’ perspective

‘Reformist’ perspective

Protected market does not allow the advantages of competition to come into play.

Protected market has : - no transparent cost structure (problem of cross-subsidies, stranded costs and benefits for nuclear power) - no ‘level playing field’.

Protected market has : - no transparent cost structure (problem of cross-subsidies, stranded costs and benefits for nuclear power) - no potential for structural change towards decentralised systems - no ‘level playing field’.

Opening of electricity market creates a climate of uncertainty; one has to wait for stabilisation of the markets in order to establish a clear vision of the future. A broad portfolio of options should be kept open.

Opening of electricity market could : - endanger social equity - limit steering capacity of government - increase import dependence (problematic for investment in national production capacity) - increase danger of bankruptcy; funds for decommissioning of nuclear power plants and management of radioactive waste could not be available when needed.

Opening of electricity market could : - endanger social equity - limit steering capacity of government - increase import dependence (problematic for investment in national production capacity) - increase danger of bankruptcy; funds for decommissioning of nuclear power plants and management of radioactive waste could not be available when needed. - lead to a ‘new medievalism’. Opening of electricity markets will shift focus from provision of energy as a commodity to the provision of energy services. Means of production should preferably be owned by local communities or ‘small’ producers.

Structure and size of Belgian economy

Belgium should continue to take advantage of its geographical position to attract energyintensive industries; these industries pose no problem as long as they live up to international benchmarks.

Belgium should continue to take advantage of its geographical position to attract energyintensive industries; these industries pose no problem as long as they live up to international benchmarks.

In the long run, demand for energy-intensive goods should diminish (e.g. by global taxation schemes based on the ‘energy content’ of products, recycling, re-use, etc.)

282

Chapter 4

Technology

Governance

Energy ‘production’ technology

EU policy / International Agreements

Technological development is a key to sustainable development! A competitive environment stimulates technology development. No technology should be excluded a priori.

Technological development is gradual and must be guided by government (norms, standards).

Limits can only be imposed if: - application threatens public health or safety - its use threatens to exhaust resources - its use degrades quality of the environment - its use introduces significant social or psychological stresses in society.

Limits can only be imposed if: - application threatens public health or safety - its use threatens to exhaust resources - its use degrades quality of the environment - its use introduces significant social or psychological stresses in society.

Energy policy will increasingly be developed at a European and global level.

Global and European agreements as a framework for national action : - binding GHG reduction targets - binding renewable energy targets - energy efficiency directives - internalise external costs.

Global and European agreements as a framework for national action :

Security of energy supply and ensuring free competition are central responsibilities.

Security of energy supply, ensuring free competition and limitation of GHG emissions are central responsibilities.

Security of supply will result from energy conservation and renewable energy policies.

Possibilities to decrease demand for energy are limited.

Logic of post-Kyoto commitments makes a slow and gradual reduction of energy consumption a necessity.

Logic of Kyoto commitments makes a reduction of energy consumption a necessity.

Energy provision drives the economy. Standards and norms acceptable as long as energy policy never unduly restricts company dynamics.

Effectiveness of policy

Technological development must be guided by ethical principles, most notably the precautionary principle : - the decision to adopt a particular technology should be reversible to a large degree - technological lock-in should be avoided - technology should be fault-tolerant to a maximal degree - technology should be compatible with a large share of renewable energy in the production mix. Society must take a holistic view to energy ‘production’ & consumption: how much primary energy is really converted into useful energy?

- binding GHG reduction targets - binding renewable energy targets - closer co-operation for green energy, e.g. European trade system of green certificates - energy efficiency directives - internalise external costs - European consensus on unacceptability of nuclear power (e.g. redirecting R&D funds away from nuclear research)

A reconstruction of the policy development cycle in the case of the Belgian phase-out decision

Strong preference for voluntary schemes (e.g. information, education, voluntary agreements, labelling, energy audits,...).

Stringent government intervention is warranted (energy efficiency standards, use of financing instruments,...).

Stringent government intervention is warranted, not only in ‘traditional’ energy policy domains (tariffs, regulation of supply, diffusion of energy efficient equipment, providing information) but also in domains with a large impact on energy use, such as spatial planning, housing policies, or actively shaping consumer culture.

Nuclear power inherently undemocratic

Technology is not judged on its ‘political merits’

Society and politics should reassert its democratic control over nuclear power operators (ensuring responsibility, liability and full transparency).

- nuclear power plants can only be run by large organisations - controlled by a limited number of power centres - influence over society (reversal of power relationship) - government intervention has to ensure stable conditions.

Nuclear phase-out decision

Decision will not influence investment decisions now; is merely ‘symbolical’.

Phase-out decision has to be the result of an open debate.

Phase-out decision sends strong signal and encourages creativity in development of energy alternatives.

Decision should be informed by a large debate and enjoy support of different social partners.

Decision should take into account interest of future generations and the environment on an equal footing and should be enforceable.

Decision has impact on human capital: young people will not be encouraged to take up careers in nuclear science. Role of government

Decisions should be made by a small group (e.g. central government), based on objective data and ensuring stability.

Science-policy interface

Based on trends (demographics, economic growth, energy prices,...) and assessment of potentials of different technologies to meet energy demand.

Policy and discourse oriented integrated demand and supply side bottom-up scenario.

284

Chapter 4

Social-cultural

Citizen participation

Strong reservations.

Strong reservations, preference for traditional social consultation.

Dialogical form of democracy is a cornerstone for sustainable development.

Role of ‘green’ NGO’s

Part of problem: irrational.

Part of democratic problem definition: inclusion of different opinions.

Part of solution: advocating interests of future generations and nature.

Role of ‘business’

Sustainable entrepreneurship.

More attention to command and control measures.

More attention to command and control measures.

Role of experts

Essential to objectify the discussion.

Necessary ingredient in societal debate, under conditions of transparency, better communication between experts and decision-makers, clear rules about expert consensus.

Problematic in polarised debates.

Behaviour

Focus on individuals (or individual entities such as firms) as consumers to determine energy demand.

Focus on individuals (or individual entities such as firms) as consumers to determine energy demand.

Focus on entities (firms, utilities, infrastructures, decision-making mechanisms...) that influence energy demand.

Consumption patterns

Increasing levels of affluence will inevitably increase demand for energy services.

Increasing levels of affluence will inevitably increase demand for energy services. Governments should intervene.

Historic trends cannot be taken as a yardstick. The ‘need’ for energy is a cultural construct, resulting from an interaction between economical, political and technical choices.

CHAPTER 5 A

PROPOSAL FOR A NEW GOVERNANCE STRUCTURE AS A SUPPORT

FOR SUSTAINABLE ENERGY POLICY Whereas the previous chapters were mainly written from a critical-analytical and reconstructive perspective, our primary concern in this chapter is to set out a constructive proposal for a governance structure more adapted to the complex challenges of sustainable energy provision. One of the main challenges identified so far has been to offer recommendations for improving the quality (in terms to be specified) of scientific advice used in developing policy measures under conditions of profound uncertainty (not only in terms of scientific evidence, but also in terms of moral or ethical uncertainties leading to the inability to base policy decisions on universally applicable principles). As previously shown, such conditions of profound uncertainty are often not recognised for what they are. Instead, political, moral, ethical and/or expert discourses alike often inappropriately transform uncertainty into certainty, with the added dangers of a (later) loss of (political, moral, ethical or scientific) authority. Thus, paradoxically enough, what seems to be needed is a sort of ‘immutable mobile’ governance structure – i.e. a governance structure that is capable of both providing a measure of stability, whilst staying flexible enough in order to adapt to possibly ‘surprising’ circumstances. In this chapter, we argue that the problems discussed call for a conceptual rethinking of political planning in terms of an iterative collective learning experience firmly tied to socio-material practices in the public sphere. The suggestions made for advancing such collective learning experience are set out in the form of fundamental requirements for sustainable energy governance, a model process securing those requirements, some supplementary notes going into the some of the details of quality assurance in the model process, and finally possible pitfalls and difficulties will be discussed. The fundamental requirements (Section 2) set out the general guiding principles for sustainable energy governance. These principles are meant to provide flexible and context-specific guidance: they may be of variable importance in different contexts, can be in conflict with other principles, and they allow discretion for decision-makers to balance them and be guided by those they find to be most important. In section 3, we develop a model process and discuss in detail how the fundamental requirements should be applied to the different steps of the policy development cycle. We argue in favour of harnessing the potential for institutional innovation contained in the precautionary principle, which is increasingly accepted as a principle for decision making in legal text on a growing number of domains (Section 3.1.1). In particular, the precautionary principle creates an opening towards a more open problem framing and traditionally under-represented perspectives. New ‘actants’ can be invited to the ‘negotiation table’, even if it is still uncertain what they have to ‘say’. In order to give ‘voice’ to these actants, a set of sustainable energy indicators should be developed

286

Chapter 5

progressively, drawing upon existing initiatives and bottom-up elicitation of matters of concern (Section 3.1.2). Next, strategic priorities can be decided upon by establishing a hierarchy between these ‘actants’, based on a ‘broad’ cost-benefit analysis. For this task, we propose to draw upon the tradition of scenario building (Section 3.2.1). In particular, we argue that the major advantage offered by such ‘broad’ cost-benefit analysis lies in the fact that it can bypass the difficulties engendered by a reference to particular scenarios as representing some form of objectivity (e.g. based on trend extrapolations, econometric modelling, etc.) and instead openly discuss the normative framings embedded in these scenarios. Furthermore, in order to support a collective learning experience, practical policy measures derived from strategic priorities should also uphold certain qualities, most notably reversibility (i.e. avoiding lock-in effects), flexibility (i.e. the ability to enlist a policy measure for different purposes) and diversity (i.e. adopting a number of disparate options in parallel) (Section 3.3). Policy measures should also be subject to a periodic review (Section 3.4). This review should concern the intended results of the policy measures themselves, as well as an evaluation of changes in the knowledge base and the rules of interaction. The quality of deliberations between such heterogenous perspectives will crucially depend on the creation of a ‘platform’ for collective learning, based on shared rules of interaction and clearly defined roles (Sections 4 and 5). This platform would have to serve as a consultative body with an output towards the formal democratic institutions; but still, should be located at some ‘distance’ from the deliberations going on in these institutions. In this chapter, we argue in favour of creating a new ‘energy agency’ that would act as such a ‘learning platform’. We also give some attention to the question of how the governance model and the new energy agency might be embedded in the context of Belgian energy policy as an elaboration of a precautionary approach (Sections 6 and 7).

1 Introduction Having followed us through the theoretical explorations in chapters 1 and 2, and the reconstructions of combined scientific/political practices in chapters 3 and 4, readers might be left with the somewhat forlorn conclusion that the practical realisation of a (more) sustainable energy system quite simply represents too monumental a challenge for political planning. Basically, the problem is that sustainable development – whatever else this means in detail – is per definitionem (in view of its future-oriented and global outreach) a long-term challenge to society. This means that policy measures for sustainable development have to include some normative aspects (or far-ranging expectations), which, in view of ensuring stability, should be broadly shared over a long enough period of time among the different actors involved. The question is of course whether political planning is capable of generating this consensus. Although perhaps offering a tempting prospect, reverting to a model of comprehensive political planning (sometimes referred to as a ‘blueprint’ approach) in our view would

A proposal for a new governance structure as a support for sustainable energy policy

287

seem to be an unwise and even blatantly unrealistic move. Unwise, because energy policy planning in the past – predicated as it was on finding and realising an ‘optimal’ blueprint electricity generating capacity for meeting a certain forecast of electricity demand (under a diverse range of ‘exogenous factors’ such as GDP growth, prices on international energy markets, etc.) – has revealed its shortcomings: it tends to lead to a succession of ‘monocultures’ (coal, then nuclear) and overcapacity (especially in the ‘80s) on the production side due to an insufficient attention to uncertainties and demand-side options (Laes et al. 2004c). Unrealistic, because the ongoing liberalisation of energy markets subjects the choice of generating technologies to the logic of short-term market economics, with governments becoming increasingly bereft of the planning power and instruments available earlier in protected markets. Combined with the broader dynamics of contemporary pluralistic ‘risk societies’ this evolution likely renders the task of marrying the (for the moment at least) ‘quasi-competitive’ energy industry to the wider sustainability demands altogether more difficult 434 . We do not want to enter here into the debate whether planning techniques are likely to be pressed by political elites into the service of dubious ends (as some would argue was the case for the Belgian nuclear power programme); the point is that even laudable and well-intentioned ends can result in disappointment because of the indeterminacy of social processes 435 . In view of these difficulties, this chapter does not seek to offer ‘a solution’ to the problem of sustainable energy use and provision. Rather, our aim is to contribute to a better practice of policy development by formulating practically feasible rules for policy design, taking into account the context-dependent and multi-faceted character of the energy challenge. In a way, our position covers the middle ground between so-called ‘rationalistic’ and ‘evolutionary’ perspectives on strategic planning (van der Heijden 1997). Rationalistic approaches basically assume that there is one ‘optimal’ answer to a policy problem, that people think and act rationally, and that implementation logically follows from the discovery of the ‘optimal’ answer. Internalising the external costs of development (cf. Chapter 3) is one example of such a rationalistic approach. Evolutionary approaches see strategic planning as an ‘emergent’ result which relies heavily on filtering out unsuccessful attempts in the past. The ‘incrementalist’ approach advocated by Lindblom (cf. infra) is an example of such an evolutionary perspective. At the bottom of our proposed governance model lies the particular theoretical conception of communicative (or discursive) rationality we developed in the first few chapters of this dissertation. We briefly recall here Habermas’s basic idea (cf. Chapter 2 – Section 3.4) of a reliance on an informally organised public sphere, ranging from private associations to the mass media located in ‘civil society’, to assume the central responsibility for identifying and interpreting problematic (‘unsustainable’) situations. We basically agreed with Habermas in tying our hopes to a ‘vibrant public sphere’ (that is, in a somewhat modified form, cf. infra) as a measure of quality assurance for public policy 434

The difficulties of (often much needed) long-term (technology) policy planning facing the often short-ranged aspects of modern pluralistic societies (e.g. election periods, public acceptance, etc.) are discussed in Grunwald (2000). 435 And often have done so in the past – for examples, see Grunwald (2000).

288

Chapter 5

making. The basic task of the public sphere then becomes one of laying siege to the formally organised political sphere by encircling it with ‘good arguments’, without however attempting to overthrow or replace it. However, we also raised some major objections against the Habermasian conceptual framework. Chief among these was that it remained firmly routed in traditional epistemological notions of knowledge (embedded in fundamental pragmatic-transcendental presuppositions of communication) and the pure distinctions between ethical, moral, truth-seeking and pragmatic discourses this entails. The point is not so much that this is wrong from a theoretical point of view, but rather that the search for these purified forms of communication tends to divert the attention from the practices, processes and contexts in which knowledge is generated and deployed and, as a result, informs the deployment of science and its embodiment by policy and institutions in less than optimal and sometimes even counterproductive ways 436 . As an alternative, we have embraced (a particular version of) the constructivist point of view advocated by Latour (and others). Constructivism takes issue with discourse theory on a number of issues (discussed in Chapter 2 – Section 3.4), but perhaps the most important among these is the emphasis on the practical engagement between ‘people’ and ‘things’ (called ‘actants’ by Latour) in combined socio-material domains (constituted by theories, skills, organisations, materials, etc.). Securing democracy then not only becomes a matter of involving all those with a legitimate interest in the issue at stake (as proposed by proponents of ‘participatory democracy’), but crucially also facilitating processes in which all relevant ‘actants’ (taken from an ‘enlarged’ public sphere – if you wish – made up of socio-material practices) are accounted for. Latour has developed the first general outlines of such approach in his “Politics of Nature” (Latour 2004a) but, due to its mostly theoretical (and somewhet utopian) focus, fails to provide more practically-oriented guidelines. It is precisely this gap we will attempt to fill in in the present chapter (albeit with a focus on energy policy). Thus, besides taking a cue from Latour’s core conceptual apparatus (see e.g. Latour 2003b, 2004a) – elaborated and commented upon in chapters 1 and 2, the model process proposed in this chapter is further enriched by ideas gathered from: • Our case-studies on the practical application of scientific advice to energy policy problems (cf. Chapters 3 and 4); • A sample of personal experience and views of representatives of public interest groups, scientific institutes, advisory councils and other stakeholders collected in the interviews (cf. Chapter 4);

436

To be sure, Habermas does propose a solution for mediating between different spheres and discourses in his discussion of the law functioning as a ‘bridge’ between the public and the political sphere, but in view of the largely metaphoric character of his discussion of the interaction between both spheres, it remains unclear what specific proposals for mediating between them might arise.


289

• The structure and functioning of Belgian energy policy in a historical (Leroy 1979; Verbruggen 1986; Laes et al. 2004c) and contemporary (Verbruggen and Vanderstappen 2003; London Economics 2004) perspective; • The empirical literature on advisory science for regulation (Jasanoff 1987, 1990; OXERA 2000) and diplomatory science (Ishii 2002) – the difference between both being that ‘regulatory’ science implicitly models policy-advisory science at the domestic level (and thus is concerned with furthering the task of policy development in a context where statutory agencies are ultimately able to enforce compliance with policy measures, with accountability being guaranteed in theory by legislative measures), while ‘diplomatory’ science models policy-advisory science at the international level (and thus is concerned with finding ‘diplomatic’ solutions, meaning that they must be in conformity with the norm of sovereignty in international negotiations). Although our model is concerned with the domestic level, the ‘diplomatory’ science model still is of interest to us because it necessarily seeks for conciliatory solutions to the tensions arising at the science-policy interface, and thus is more demanding in terms of trust building among the actors involved (besides indirect ‘normative’ pressures, no actor can be ‘forced’ to accept a solution; instead, acceptance is based on self-binding commitments) 437 ; • The literature on policy planning (van de Graaf and Hoppe 1996; Meadowcroft 1997), and in particular, incrementalism 438 . There is a wide acceptance that the incrementalist model of policy making captures something of the spirit of decision making in contemporary pluralistic democracies (Meadowcroft 1997) 439 . Whether it is the best 437

Climate change negotiations under the ‘United Nations framework convention on climate change’ (UNFCCC) and the European ‘long-range transboundary air pollution’ regime (LRTAP) often figure as prominent examples of ‘diplomatory science’ (Ishii 2002, p. 262). Diplomacy also plays a crucial role in Latour’s “Politics of Nature” (Latour 2004a, pp. 209-220), as one of the crucial tasks in the slow composition of the collective. 438 Originally developed by Charles Lindblom in the ‘60s, incrementalism has become a well-known theory in policy sciences. Lindblom has offered a consistent critique of comprehensive approaches to political planning or policy analysis (see e.g. Lindblom (1965); for a discussion of Lindblom’s core ideas, see e.g. van de Graaf and Hoppe (1996, pp. 239-243) and Meadowcroft (1997, pp. 434-439)). Two concepts figure prominently in Lindblom’s argument, namely ‘incrementalism’ and ‘partisan mutual adjustment’. Incrementalism as a method of policy analysis seeks to reduce the complexity of real-life policy-making context through: a) limitation of the analysis to a few somewhat familiar policy alternatives; b) an intertwining of analysis of policy goals and other values with the empirical aspects of the problems; c) a greater analytical preoccupation with ills to be remedied than positive goals to be sought; d) a sequence of trials, errors, and revised trials; e) analysis that explores only some, not all, of the important possible consequences of a considered alternative; and f) fragmentation of analytical work to many (partisan) participants in policy making. Lindblom argued that all policy analysis should rest on the premise that complex social problems cannot be analysed exhaustively; therefore, policy makers should shun the promises of a ‘comprehensive’ form of policy planning, for by seeking to approximate this ideal, they would most likely fall into more severe failures than those who ‘knowingly and openly muddle with some skill’. As a political strategy, ‘incrementalism’ recognises that in democratic systems changes can only occur through taking ‘small steps’. Lindblom’s other central concept, that of ‘partisan mutual adjustment’, is in his view typical of contemporary pluralist democracies. Here a wide range of participants and interests contribute to outcomes, and policies are, according to Lindblom, perhaps best described as ‘happening’ rather than as the result of some form of conscious choice. At the core, ‘partisan mutual adjustment’ represents the political equivalent of the ‘invisible hand’ theory in economics: somehow, the ‘political mechanism’ – the coordination achieved through interaction, conflict and accommodation ‘on the ground’ – assures the rationality brought to bear on the decisions. 439 In this regard, it is instructive to notice that one of Lindblom’s criteria for a ‘small step’ is a policy measure whose impacts remain rather insignificant within the next five years – the Belgian phase-out decision thus clearly falls into this category.

290

Chapter 5

analytical strategy to follow in the face of the long-term challenge of sustainability remains to be seen. Nevertheless, particular elements taken from that strategy – such as the emphasis on ‘trial and error’ and the ‘fragmentation of analytical work among many agencies’ – seem to be promising in the context of sustainability problems 440 . As explained, the complexity of these problems implicates a plethora of loosely connected and fragmented experiences in which groups or individuals achieve partial insights into complex (social, physical, etc.) reality – features which render them particularly amenable to management through flexible networks. However, it is of course not sufficient to just multiply new forms of bonding; there must also be some questioning about how this bonding is achieved and what can be learnt from it 441 . On this issue, we agree with Grunwald (2000, p. 129), who proposes a model of ‘directed incrementalism’ – i.e. a model that allows long-term orientations to be maintained though the method of proceeding is an incremental one, whilst also allowing short-range flexibility requirements to be taken into account – as an alternative to a ‘purely incrementalist’ planning procedure. Advocating a form of ‘directed incrementalism’ however immediately raises the question to what extent network interactions can and should be ‘directed’, or rather shaped, by government intervention – a question we will take up in section 3.4; • The literature on long-term planning in the field of technological (policy) choices (Grunwald 2000, 2004) and in particular ‘transition management’ (Rotmans et al. 2000; Rotmans 2004; Turkenburg 2004) 442 ; • The literature on the precautionary principle and its relation to strategic policy making (de Sadeleer 1999; Stirling 1999; Dratwa 2002; von Schomberg 2006). The ‘lessons learnt’ from these investigations are structured as follows. We first discuss the general requirements for the governance model, which are meant to be fundamental and as comprehensive as possible, and pervade the recommendations made (Section 2). Secondly, we set out our view of a model process, which, according to us, represents the most appropriate way to proceed in planning for sustainable energy policy (i.e. the model process is consistent with the general requirements). The process addresses all stages of the policy development cycle, including the identification and structuring of policy

440

Cognitive psychology has by now amply proven that people in general (including decision makers) lack the ability of ‘correctly’ dealing with uncertainty on a conceptual level (i.e. according to the fundamentals of statistics) – the picture of man arising from this research rather is one of a trial-and-error learner, relying predominantly on habit or simple deterministic rules (for an overview of relevant research insights, see Slovic et al. 2000). Thus, it seems realistic to depart from such ‘trial-and-error’ learning behaviour as a ‘realistic baseline’ in order to produce more detailed guidance on more ‘productive’ forms of trial-and-error (cf. Section 2). 441 Contra Beck’s (1992) (in our view) overly optimistic and uncritical assessment of the emancipatory potential of sub-political actions (cf. Chapter 1). 442 Transitions are defined as broad collective changes comprised of technological and behavioural changes, guided by institutional innovations creating an overarching framework. Transitions commonly take some decades to be completed. Transition management then is a collective learning process facilitated by government aimed at realising desirable transitions while averting undesirable ones (Rotmans 2004). An example of the latter would be the avoidance of choices dictated by short-term interests, and contrary to longterm desirable goals.


291

problems; definition, comparison and selection of options; political decisions; initiation of policy instruments and measures; and the matters of concern to be taken up in the next iteration of the collective learning process (Section 3). Next, we discuss quality assurance (Section 4), before going into the contribution of different ‘roles’ (e.g. experts, stakeholders, ethicists, etc.) to the tasks set out in the model process (Section 5). The last two sections of this chapter deal with some more practically-oriented questions, i.e. how such a (demanding) procedure might be embedded in actual Belgian energy policy-making practice (Section 6), and what the possible ‘pitfalls’ or difficulties facing our proposal might be (Section 7), before concluding (Section 8). Although the explicit aim of the present chapter is to provide a context-dependent framework to consider the design of policy as well as the contribution of the appropriate instruments and institutions in the field of sustainable energy policy (and hence, in no way we are claiming ‘universal’ validity), we have nevertheless tried to set apart the more ‘general’ model process from observations which particularly relate to the context of Belgian energy policy. Hence, we think the model might still prove to be useful (perhaps with some minor modification), provided it is applied to decisions which a) are in part the responsibility of government; b) are in part dependent on science, which however is uncertain (either because theoretical insights are uncertain or controversial, data are missing, etc.); c) involve potential and ‘significant’ risks to the public or the environment which are however highly controversial (and even the ‘desired outcomes’ might raise conflict). The present chapter thus sets out the model in quite general terms; however, the next chapters (Chapter 6 and 7) explore (in a tentative way) its practical consequences in the context of Belgian energy policy planning.

2 Requirements Before setting out the model process itself, we thought it would be useful to distil from the previous chapters a collection of central ideas or requirements (which up till now might be found lying somewhat scattered), expressing the underlying philosophy applicable to the model process. Some of these requirements may sound rather self-evident or even trivial, but we have nevertheless chosen to take them up here in order to provide a complete picture of complementary requirements which, taken together, should provide guidance on planning for sustainable development. 1. Sound science: this may sound obvious, since no-one could seriously advocate the rejection of the significant advantages offered by a reliance on ‘sound science’ – i.e. the production of reliable theories with explanatory and/or predictive power through the use of logic, reasoned arguments, evidence, etc. – also in political contexts. However, scientific advice for political planning often calls for information that is less firmly established or is characterised by significant uncertainties, and, as already elaborated, the norms, rules and actors and/or institutions involved in ‘science for policy’ differ

292

Chapter 5

profoundly from those involved in ‘research science’ 443 (cf. the ‘trough of uncertainty’ argument developed in chapter 2) 444 . Thus, our emphasis on sound science as a requirement for the model process seems relevant for at least two fundamental reasons. The first is that, in order to obtain scientific legitimacy, participation of prominent scientists to the process is particularly important. Nevertheless, it is always a rather difficult task for scientists to participate in political arenas, for instance because the danger exists that the involved scientists would become ‘contaminated’ by acquiring a ‘politicised image’ and thus do harm to their scientific career. The second is that acceptance of the precautionary principle as a guideline for political action (as we propose) has often been attacked on the premise that it is supposed to be in conflict with the requirements of ‘sound science’ 445 . The challenge becomes then to uphold ‘sound science’ (and secure the participation of prominent scientists) in the model process by a careful management of the science/policy boundary. Indeed, we (as others have done before us – e.g. de Sadeleer 1999; Stirling 1999a; EC 2000; Dratwa 2002; Lierman 2004) argue that embracing a precautionary approach in no way equals embracing scientific relativism. However, it does mean that we have to be able to go beyond what Grin (2000, p. 16) has called the ‘Cartesian anxiety’, i.e. the fear that we are lost without a fixed Archimedean point to guide our decisions. Thus, ‘sound science’ should be reconceptualised in terms of discovering contextually situated ‘wise decisions’ rather than universally valid theoretical claims. 2. Managing the science/policy boundary 446 : in view of our previous discussions, it is clear that there is no such thing as absolutely objective and value-free science. This should not be seen as a ‘fault’ of scientists (who ‘deliberately’ mix up science with politics), but rather as a result of the particular arrangements in which they operate 447 . Science should rather be seen as having a dual role of one of the means to legitimate the reconfiguration of the collective, as well as one of the ‘teachers’ on how this reconfiguration should be made. Thus, managing the science/policy boundary cannot be

443

We are well aware that there can be no sharp boundary between ‘disinterested’ science and ‘science with political implications’; however, we use the term ‘research science’ to refer to science that is (relatively) unrelated to specific political decisions. 444 Another example would be dealing with resource constraints (e.g. time, budgets). In research science, such constraints might prompt researchers to look for pragmatic solutions in line with disciplinary requirements; however, in science for policy pragmatic solutions have to include the interpretation and presentation of scientific knowledge that is useful and significant for political negotiations (in addition to the fact that time is usually ‘on the side’ of the research scientist, since conclusions will not be generally accepted until most of the members of the relevant scientific community are convinced). 445 Some argue that without clear scientific evidence of risk, regulation may reflect arbitrary or ill-informed fears, risk perceptions or plain misconceptions, or illegitimate motivations such as trade protectionism. 446 Again, we are aware that speaking of a boundary between ‘science’ and ‘policy’ might seem to send a conflicting message with all of our previous arguments. It would be better to speak of a boundary between the power to take into account particular actants (in the words of Latour – i.e. concerning the task of ‘problem structuring’) and the power to put them into order (i.e. concerning the tasks of ‘institutional support’ and ‘reasonable choice’). However, the contribution of science will be most accentuated in the first task while the political work will carry its greatest fruits in the second, so we chose to keep the traditional (if somewhat inaccurate) ‘science-policy’ separation. 447 E.g. in our discussion on (the use of) external cost calculations, we have shown that these are only definable by having institutional mechanisms in place for defining them (e.g. in the case of dismantling funds for nuclear power plants).


293

a question of separating once and for all scientific from political ‘products’, but rather of carefully managing (e.g. through a clear definition of different tasks for participants in the discussion) the continuous negotiation process between both worlds. This does not mean that we have to revert to a (rather impotent) ‘purely’ procedural solution, as each discussion also needs to lead to a (provisional) decision or closure. Such closure should however remain accountable to both ‘scientists’ and ‘politicians’. 3. Overriding duty to the ‘good common world’: though one of the central tenets of our argumentation so far has been that there is no such thing as ‘neutral’ science or ethics (or politics) having an undisputed access to ‘the common good’, we nevertheless maintain that our model process should strive (as far as possible) for a consensual definition of the ‘good common world’ (based on both the right ‘facts’ and ‘values’ held in common). An overriding duty to the ‘good common world’ means at least avoiding that knowledge is used in the process which is perceived to embody particular ideological or unevenly distributed interest biases. Science should not be seen as just another ‘stakeholder’ amongst others (Grunwald 2004). As there are no absolute criteria to judge a commitment to the ‘good common world’ (or, in our case, sustainable development), this rather has to be judged on a common interpretation of the substance of the scientific advice and the processes governing its production and use (i.e. reflexive knowledge). Still, we have to accept that the ‘good common world’ cannot be established once and for all, and thus the governance process has to be capable of ‘closure’ through the ever provisional expulsion of certain ‘actants’ (i.e. through the creation of ‘adversaries’ – cf. Chapter 2). 4. Collective learning: the uncertainty and incompleteness of knowledge and the provisionality of evaluations make a complete implementation of sustainability, in the sense of a ‘blueprint’ or detailed planning, simply impossible. It is ex ante not stringently decidable whether and to what extent a political measure, a new technological innovation (or existing technologies) or a new institutional arrangement will contribute to sustainable development. Hence, planning for sustainability cannot be a question of simply constituting a one-off ad hoc advisory mechanism (no matter how carefully composed or planned), but rather calls for continuous assessment and learning. Governance for sustainability remains – because of a fundamental provisionality of knowledge and evaluation – bound to the metaphor of an experiment. In our view, this is one of the crucial contributions of Latour (2004a): instead of the incrementalist ‘trialand-error’, Latour proposes the extension of the model of scientific experimentation – i.e. well-considered arrangements of technical apparatus, methodical procedures and observational measures – for answering specific questions. The difference between both approaches lies mainly in the fact that experiments are developed through careful planning, even though the results cannot be predicted with certainty. It is furthermore required for sustainability planning that the results of these learning processes have some consequences for future practice; hence, sufficient leeway must be given for the integration of new knowledge. 5. A multi-tiered, decentralised, multi-actor process: from the observation that planning for sustainability can only take the form of a collective experimentation, one should not get the wrong idea that there is only one ‘collectivity’ or one central agent conducting

294

Chapter 5

the experiment 448 . On the contrary, there can be no definite answer to the number of collectivities engaged in experimenting and learning 449 . However, in the concluding section of chapter 2 we still argued, in view of the unique functional capacities of the state apparatus in assuming an orienting role and/or coordinating the nexus between international commitments and national performance, in favour of a ‘central body’ located near the political system, which would collect the most important lessons learnt. The ‘central body’ we envisage, whilst not having a privileged access to the ‘common good’, would fulfil an essential role of focusing towards institutionalised decision making in legislative bodies. This focusing has two aspects. Firstly, it should help in bringing together all relevant perspectives, arguments and learning experiences. Secondly, it should be able to give a more detailed consideration to arguments than can for instance be expected from the individual decision maker who is often enmeshed in different pursuits at the same time. These requirements in turn impose some technical conditions on the composition of the ‘central body’, so as to ensure an adequate representation of viewpoints. Furthermore, there must be some guarantees that there will be a real exchange of arguments, so that the weaker ones are screened out and the better ones survive. This condition in turn points at the importance of a design of procedural mechanisms that are able to compensate for the distorting effects of selfinterest and power. 6. Representing all relevant perspectives: our focus on a ‘central body’ should not render us blind however to the necessity of engaging with everyday life world experiences. This engagement is important because the focus on ‘representational’ understanding at the higher strategic level will tend to obscure how this understanding might connect with everyday practices and behaviours 450 . We might then come into a position to effect sustainable energy development by directly engaging with matters of ‘lifestyles’, practices and behaviours, rather than, as different ‘dominant cooperative schemes’ tend to do, by way of changes to technology or economic incentive structures (in the case of energy policy, promoting more ‘rational’ forms of energy use could be an excellent

448

For the sake of simplicity, one can introduce some general distinctions between learning at a ‘macro-level’ (of society as a whole), a ‘meso-level’ (of industrial sectors) and a ‘micro-level’ (of individual technologies). 449 The context of energy policy is increasingly being shaped by complex interactions between people, political mechanisms, technologies etc. on multiple levels (global, regional, national, local, etc.): one might think of complex associations made up of greenhouse gases, the IPCC, the WTO, EU directives, individual consumers, etc. Furthermore, these associations are increasingly mediated by ‘instruments’ – e.g. databases, model calculations, scientific colloquia, media interventions, consumer actions, etc. – which contribute to their form and significance in the political sphere. In fact, the political will to pursue a particular policy may trigger so far unconnected institutional resources to come together under the umbrella of that particular policy in previously unsuspected ways. Indeed, the variety of possible organisational forms – that is, of possible unfoldings vis-àvis a wide range of catalysts (e.g. policy measures, innovations, unsuspected events, etc.) – can be very large indeed. 450 Policy relevant knowledge often results from a process of ‘giving and asking reasons’ (especially when developed according to the rules of the discourse ethic) and thus tends to privilege theoretical knowledge over other, more practical and contextual, forms of knowledge. Even when ‘public participation’ is part of the policy-making process, this is often conceived in the form of a retrieval of representations held by ‘laypeople’ and the transmission (by symbolic means) of information to responsible policy makers.


295

example – cf. Section 3.4.1) 451 . ‘Representational politics’ should be complemented as far as possible by ‘politics of presence’. This, however, requires a much more careful attention to participatory practices than witnessed by a certain propensity in literature on the subject to conceive of participatory procedures as a ‘bolt on’ input to conventional processes of governance (see e.g. Renn 1998). Optimising public involvement requires a creative interplay between the typically highly contextualised understandings of laypeople and the decontextualised expert understandings. In section 3.4 we will briefly indicate some ways in which the ‘politics of presence’, with an emphasis on life-world practices, might be facilitated. 7. Quality control: facilitating learning and striving for consensual acceptance of results inevitably implies that the model process must in any case be accountable to the collective engaged in the learning experience, hence calling for a mechanism of quality control. Quality control generally deals with commonly accepted norms of fairness, competence and transparency. For instance, when using computer models to define potentially politically-sensitive parameters (e.g. costs of reducing greenhouse gas emissions, impacts on GDP, etc.), scientific advice cannot rely on non-transparent ‘black-box’ models, because it cannot be accountable to the actors involved in the discussion. Quality control will be dealt with in section 4. 8. Uncertainty management: in facing complex long-term questions such as demands for more sustainability, uncertainty is simply irreducible; hence, it should form an integral part of problem-solving strategies. As our investigations have shown, resolving uncertain issues is not simply a matter of finding the right data in order to fill in knowledge gaps as the relevant data might be contested between scientists, or simply not be available on principal grounds. Furthermore, our case studies show that various elements of uncertainty are a source of political conflicts, effortlessly deployed by participants in the conflict to attain their political agenda; with closure often reached not on the basis of the ‘better argument’ but rather after adversarial debates (with little learning on both sides) and the wilful imposition of a ‘solution’ by the most powerful actor. Although we can of course not guarantee that actors involved in the sustainability debate will refrain from having recourse to such adversarial proceedings in the future, reliance on adversarial procedures for bringing about closure of the debates would make a reasoned agreement impossible from the outset. Instead, uncertainties should be dealt with in a candid and honest way. However, scientific advisers themselves may not be the best judges of uncertainty in their own advice, although they can certainly make a major contribution towards its assessment. Resisting the temptation to close off scientific issues prematurely should thus be achieved through careful procedural design (cf. Section 4.3).

451

Experiences with public involvement in the field of radioactive waste management have already demonstrated public participants’ great capacity to broaden the frame of reference beyond technical issues in order to explore social, ethical, environmental and historical perspectives. For the Belgian experience, see Bombaerts (2004).

296

Chapter 5

3 A model process The process developed and described here can in a way be seen as describing ‘good practice in governance for sustainability’, in agreement with the requirements set out in the preceding section. Relying on a model process brings a number of advantages (Webler 1994; OXERA 2000): a) to some extent, adopting constraints (‘rules’) for deliberations is desirable in view of efficiency considerations (rather than designing ad hoc interventions each time, people like to rely on ‘proven’ solutions), and, furthermore, once a process has been installed it is easier to follow it; b) a codified process is easier to audit and review (hence enabling public accountability and legitimacy, providing of course that the ‘rules’ are generally agreed upon); c) a process reduces dependence on the judgment of individuals (trust in a justified process sequence might be easier to achieve, as it is based on the implicit ‘social contract’ of modern democratic societies (Grunwald 2000)); and d) if the same process is used over time (perhaps with minor modifications), it assures consistency of decisions. The model process we propose is built around the functions of the policy development cycle (used in chapter 2 and 4); they are set out here in a logical sequential order. However, as mentioned earlier, this does not imply that in reality these steps will be taken one after the other, as there will be considerable overlaps, feedbacks, anticipatory moves, etc. In all of these steps, the precautionary principle will function as a guiding principle, as we will show how it could possibly be mobilised for implementing our model process in the first place, for the choice of a particular level of protection (defining the boundary between ‘sustainable’ and ‘unsustainable’), the choice for particular measures implemented as part of the process, and the assessment of scientific uncertainties in the knowledge underlying the decisions. We are not claiming that we have solved all problems of sustainability governance with our model process, indeed, work has only just begun and the model process itself cannot escape from the requirement of collective learning. But is does have the advantage of at least setting out an ‘experimental protocol’, so that e.g. after learning in practice, a justification can be given for omitting or adapting parts of the process and/or for the adoption of an alternative (the general requirements set out above should not be violated however). Chapters 6 and 7 will be concerned with such (preliminary) experimentation; while in the last sections of this chapter, we will also explore the possibilities of practical implementation in the Belgian policy context (Section 6) and some reasonably imaginable ‘pitfalls’ facing our proposal (Section 7).


3.1

297

Selection and structuring of policy problems

As always, when proposing something new, one is confronted with the problem of bootstrapping. We are faced here with the difficult question of how it can be possible for rather deep-seated ideas and (antagonistic) patterns of interactions to change – and change quite rapidly, if one is to believe the appeals of urgency often voiced in the context of sustainable development 452 . Latour (2004a, p. 61) offers no further guidance than merely asserting that actors engaged in the debate will be willing to surrender their ‘metaphysics’ if we move from a ‘warlike’ version of public life to a ‘civil’ one 453 . However, as Wynne (1995) points out, collective learning is in itself related to normative issues – and this normativity has to be justified again by recurring to ‘underlying cultural rationality standards’. In the following sections, we will defend invocation of the precautionary principle as one of the serious candidates for fulfilling the role of such ‘cultural rationality standard’ of the future. 3.1.1

Invoking the precautionary principle

So how should we kick start the model process then? At this point, it seems useful to remind our readers of our reflections on what construction of a ‘sustainability commonwealth’ (in the words of Boltanski and Thévenot) should look like. In chapter 1, we explained that in order to begin constructing this new commonwealth, the ‘state of smallness’ would be to know exactly and irrevocably what a thing ‘does’ (what its function is); while the ‘state of grandeur’ accrues to leaving open the question of the ordering of means and ends. The contours of this new commonwealth imply a suspension of the certainties of the other commonwealths; it induces a moment of ‘perplexity’ – no negative form of scepticism is involved, rather it designates the starting point for a collective process of building associations between the ‘actants’ involved in a situation. In our view, this requirement coincides largely with the general philosophy behind the so-called precautionary principle. 3.1.1.1 The precautionary principle in institutional practice The precautionary principle has generated an enormous body of literature over the last decade from the standpoint of lawyers, environmentalists, economists, and ethicists; but for our present purposes, it is not necessary to add yet another interpretation to this growing mound of commentaries, since we are not concerned here with an investigation of the

452

Cf. the comments made on the difficulties of establishing a new ‘commonwealth’ made in chapter 1. Surrendering ‘metaphysical’ positions also implies accepting the role of emotions in technology assessment (Grunwald 2000) – at least as a starting point (and not the cornerstone) for building the ‘good common world’. Feelings of anger, anxiety, confusion, etc. towards particular policy options should not be dismissed as ‘irrational’, ‘motivated by self-interest’, ‘an expression of cultural rigidity’, etc. (to name but a few epithets which seem to pop up now and then in the nuclear power controversy). Politicians, experts, stakeholders, etc. might often have very strong (personal) commitments towards particular options, so it is not only ‘abstract’ technological solutions that are being negotiated in policy planning, but also the identities of those involved in the planning process. But the point is that emotions per se do not replace arguments – there might be very good or very bad reasons behind them. Hence the importance of procedural measures (as we propose) to check emotional responses with respect to their generalisability as a basis for taking legitimate decisions. 453

298

Chapter 5

implementation of the precautionary principle in particular regulations or national laws, but rather with the more general question of invoking precaution in a particular subject matter (i.e. energy policy) before any more precise regulation or law is available. Obviously, at this level, the decision to invoke precaution is a political one – no international institution is going to ‘force’ the Belgian government to build in precaution in its energy policy measures. But this does not imply that the decision is ‘arbitrary’ or solely dependent on the initiative or goodwill of (individual) political decision makers: in the first place, such decisions are (and should be!) informed by scientific deliberations (as well as various other ‘matters of concern’ waiting to be taken into account by the political system – cf. infra) 454 ; secondly, such decisions can take recourse to a (gradually developing) general understanding of the precautionary principle at the international or EU level which provides a rationale for action. This embedment in established legal doctrine also implies that, as a legal principle, precaution cannot forego the requirements of reasoned and consistent rule making, thus rendering it less vulnerable to particularistic political aspirations (van den Daele 2000). It is important from the outset to recognise the character of a principle (in legal terms). The traditional concept of a legal principle is that it provides an argument in a particular direction, but does not determine a specific outcome. Principles provide flexible and context-specific guidance: they may be of variable importance in different contexts, can be in conflict with other principles, and they allow discretion for decision-makers to balance them and be guided by those they find to be most important. Unless a specific formulation requires it, therefore, the precautionary principle will not determine a specific outcome or decision, and in particular will not necessitate one particular decision that would guarantee ‘total’ protection (Cooney 2004). Generally speaking (i.e. this seems to be a ‘minimal consensus’ on the principle), precaution has emerged as a broad principle weighing in favour of environmental and/or health protection in the case of uncertainty 455 . The core of the principle can be understood as countering the presumption in favour of (economic) development 456 . Where there is uncertainty concerning the impacts of an activity, rather than assuming human economic activities will proceed until and unless there is clear evidence that they are harmful, the precautionary principle supports action to anticipate and avert environmental harm in advance of, or without, a clear demonstration that such action is necessary. Precaution thus shifts the balance in decision making toward ‘prudent foresight’, in favour of monitoring, preventing or mitigating uncertain potential threats.

454

The climate change case has convincingly shown that scientific ‘matters of concern’ can be translated quite effectively into political ‘matters of concern’. 455 von Schomberg (2006) rightly notes that the precautionary rationale is particularly new for environmental questions, since in the case of threats to human health most governments in constitutional democracies already undertake early preventive actions (even in the absence of a formal reference to the precautionary principle), whereby economic benefits not easily outweigh human health concerns. 456 ‘Development’ is to be understood broadly here – not in the specific sense of raising standards of living in Third World countries. Van den Daele (2000) argues that modern liberal democracies are structurally biased in favour of technological development for three reasons: a) the institutionalisation of ‘objective’ science that can easily be converted into technology; b) the reliance on the capitalist economy with associated imperatives of innovation; c) the institutionalisation of a system of rights that legitimises the production of and access to new technologies. Reasons must be given for restricting technological development, not for introducing it!


299

Probably the most widely cited version of a generally applicable principle of precautionary action in the environmental context is Principle 15 of the ‘Rio Declaration on Environment and Development’ (UNCED 1992, p. 10). Principle 15 reads: …In order to protect the environment the Precautionary Approach shall be widely applied by States according to their capabilities. Where there are threats of serious or irreversible damage, lack of full scientific certainty shall not be used as a reason for postponing cost-effective measures to prevent environmental degradation…

Principle 15 establishes a rationale for action (e.g. by lowering the threshold level for action) and specifies particular circumstances for which the possible invocation of the precautionary principle is reserved. Points to note include: • Threats: precaution becomes relevant where there are threats of harm to the environment (or human health). There is little guidance in this formulation as to what level of evidence or suggestion or indication of a threat is required, other than that the damages involved should be ‘serious’ or ‘irreversible’ (cf. infra). The European Commission sees science-based risk assessments as an essential first step for applying precaution and estimating the threats involved in a particular policy area (EC 2000 – cf. infra); • Lack of full scientific certainty not a reason for postponing action: ongoing scientific controversy, disagreements or the lack of full scientific evidence should not count as a reason to postpone government intervention; • Cost-effective measures: the measures applied should be cost-effective. This implies some assessment of the costs and benefits of proposed measures, and some sort of proportionality between the costs of the measure adopted and the benefits to be gained (cf. infra). For the European Union, the precautionary principle provides an overarching framework that interconnects all community policies. The EU ‘Maastricht Treaty’ (Article 174) (1992) states that …Community policy on the environment shall aim at a high level of protection taking into account the diversity of the situations in the various regions of the Community. It shall be based on the precautionary principle …

Furthermore, the treaty specifies that the requirements of environmental policy (and thus also the precautionary principle) must be integrated into the definition and implementation of other community policies. In 2000, the EC published a communication on the precautionary principle, subsequently adopted by the European Parliament, which provides important guidelines for translation of the general principle into operational measures (EC 2000), particularly in order to ensure that the precautionary principle would not be abused

300

Chapter 5

as a protectionist measure in disguise, or to prevent anti-innovative impacts 457 . This communication sets out that implementation of the precautionary principle should be guided by the principles of proportionality, non-discrimination, consistency, examination of the costs and benefits of action and inaction, and examination of scientific developments 458 . Three consequences seem to arise which are relevant for our present discussion. The first is that the rationale for action embodied in the precautionary principle is by now reflected in the broad EU legal framework (in the ‘Maastricht Treaty’, the communication of the EC, but also for a part developed by case law – see e.g. Lierman 2004), thus taking away the need for particular political decisions to justify whether to apply the precautionary principle to a particular area or not. In any case, the immediate relevance of the precautionary principle for energy policy is clear: energy ‘production’ and use entails major threats with the potential for large, but possibly poorly understood, consequences (e.g. air pollution, catastrophic risks, etc.). Secondly, having the precautionary principle in place does not as such imply any particular standard-setting and thus, for instance, invoking the precautionary principle does not in itself justify the application of stricter environmental or health standards than is the case now. The EU treaty only specifies that the level of protection should be ‘high’, but community institutions still enjoy a broad discretion with the regard to the risk level deemed to be still acceptable to society. The EC guidelines merely specify that if the precautionary principle is invoked and applied to a particular case or policy area, this should be done in a consistent (i.e. the same standard and measures should be applied in similar cases) and non-discriminatory way (i.e. the same standard and measures should be applied in comparable situations) in relation to the chosen level of protection. The third point is however that, in the particular circumstances where the precautionary principle could be invoked (i.e. circumstances characterised by scientific uncertainty), it will simply not be possible to specify the level of protection in quantified terms (e.g. either because such quantification is part of the debate itself, or because of uncertainty whether the possible adverse effects actually pose a problem for the chosen level of protection). 457

With regard to the precautionary principle, fears are often being raised that scientific or technological options might be rejected because they are seen to be to ‘risky’ or ‘uncertain’ at the time the decision to invoke precaution is being made. The point is then that some potentially promising (on the longer term) scientific or technological developments might be excluded from consideration too early. We think such fears are unjustified in the sense that, if precaution would be exercised according to the guidelines of the model process set out in this chapter, political planning will not unequivocally act upon ‘fears’ expressed at a certain moment with regard to certain technologies. Listening to such (real or presumed) fears will in our model result in an increased sensitivity to the (possible or hypothetical) problems with the technology in question, but will not result in automatic rejection. We believe this to be the case because the model rests on careful contextual judgments (and practices of reason-giving) rather than decontextualised risk perceptions (e.g. ‘being for or against nuclear power’). 458 It is worthwhile to note here that with regard to costs and benefits of proposed actions, the commission explicitly states that this examination cannot be reduced to a pure economic weighing of costs and benefits. How the other recommendations will be taken into account in our model process will be made clear in the following sections.


301

Thus, we agree with von Schomberg (2006) that invoking the precautionary principle therefore implies a delicate interplay between the choice of standards of acceptability for certain risks and scientific assessments whether such standards would be violated without regulatory measures. This also implies that, while invoking the precautionary principle, one should accept to redefine the level of protection in view of new (scientific) insights. In section 3.3, we will elaborate further on the importance of such transformable standards for our model process. Ultimately, we believe that the kernel of the precautionary principle does not so much consist in shifting the burden of proof to the organisation (or a political planning mechanism) proposing a potentially harmful activity or policy (as is often claimed), but rather in shifting the notion of what counts as proof in combined scientific/political arenas (see also Dratwa 2002) 459 . It opens up the classic connection between knowledge and action by highlighting the inescapable issue of uncertainty. Furthermore, it adds to a ‘purely’ (research) scientific questioning that of uncertainty regarding the procedures which should be followed to substantiate or operationalise learning processes in collective (political, institutional, expert, etc.) controversies 460 – i.e. the ‘collective experimentation on acting while doubting’ that Latour (2004a) proposes as a solution to the dilemmas facing ‘risky’ choices in modern pluralistic societies. 3.1.1.2 Limitations of a principle of precaution However, mobilising the precautionary principle for pragmatic (i.e. its growing acceptance and progressive definition and implementation at the international, EU and national levels) and normative (i.e. promotion of collective experimentation and learning) reasons still remains somewhat unsatisfactory for the purposes of installing a new governance mechanism for sustainable development. The precautionary principle is often seen as an integral principle within sustainable development, based on an argument that by safeguarding against serious (and particularly irreversible) harm to the natural resource base that might jeopardise future generations’ capacity to provide for their own needs, it is closely linked to intergenerational equity, and thus part of the overarching concept or policy of sustainable development. But focusing in particular on the potential threats of 459

Whatever other benefits of a ‘reversal of the burden of proof’ are claimed, such rule from a legal perspective cannot meet the requirement of consistency (van den Daele 2000). The assumption that a technological innovation could entail yet unknown risks can always be made and is of course very hard (if not impossible) to refute. Empirical proof of a negative fact (i.e. that presumed risks do not exist) is logically impossible! Therefore, a complete reversal of proof would simply stop all technological innovation. 460 Von Schomberg (2006) discusses the case of EU regulations concerning the deliberate release of genetically modified organisms (GMO’s) as an actual regulatory implementation of such a collective learning approach (European Directive 2001/18). In this case, the precautionary principle was implemented in the form of a caseby-case and a step-by-step procedure. The case-by-case procedure facilitates a mandatory scientific evaluation for every single release of a GMO. The step-by-step procedure facilitates the progressive development of GMO’s by evaluating the environmental impacts of releases in decreasing steps of physical containment (e.g. from laboratory experiments. Such framework is thus able to bypass the inconclusive and unproductive deliberations on whether GMO’s in general are ‘safe or not’, by replacing such discussions by deliberation and scientific information on the safety of a particular GMO application, in particular circumstances, for a particular space of time, with particular safety measures, etc.

302

Chapter 5

technological developments might obscure other crucial dimensions of sustainable development, most notably the wider cultural or social impacts of technology. Based on text excerpts taken from Agenda 21 (UNCED 1992), Börlin (1994) identifies the following criteria for judging technological developments (besides being generally more environmentally friendly, less resource-intensive and fault-tolerant): contributing to technological capacity building in Third World countries (i.e. an articulation of the principle of ‘common but differentiated responsibilities’), better balancing of risk exposure (e.g. no exports of wastes) and reducing dominance by the North over common world heritage. More generally, a particular weakness in relying on the precautionary principle for setting into motion a collective learning practice lies in the over-emphasis of the contribution by natural sciences (e.g. biology, physics, etc.). Furthermore, most contributions on the precautionary principle (e.g. Stirling 1999; von Schomberg 2006) are at fault in emphasising too much the cognitive-representational criteria behind its rationale, for instance by insisting on characteristics such as the incommensurability of the risks in question, the latency period for potential harmful effects, potential for irreversible damages, etc 461 . Of course, (natural) scientific assessments of the uncertainties involved has its role to play as a critical resource for action, but the dimensions of (international) solidarity (e.g. concerning issues of unjust distributions of risks and/or benefits in the international community) and institutional management (e.g. preventive measures, liability regimes, etc.) of proposed technological development paths should not be overlooked. Agenda 21 does propose such an integrated vision on (technological) development, but on the downside however, from a strictly legal-political perspective, its status – whilst being signed by the majority of world states – is probably best described (compared to the precautionary principle) as ‘ultra-soft’ (i.e. there are no concrete examples of implementation of these principles in practical regulations – hence, no more detailed interpretation on their practical interpretation is at hand). Nevertheless, notwithstanding possible substantial differences in interpretation of these ‘ultra-soft’ principles, nationstates will usually try to avoid overt incompatibilities between their national laws and international laws and conventions, especially where environmental protection is concerned (Bérubé and Villeneuve 2002). 3.1.2

Structuring problems for sustainability

Invoking the precautionary principle thus seems to be one of the best opportunities for introducing Latour’s requirement of ‘perplexity’ – i.e. an investigation into the best way of detecting propositions that are candidates for constituting the ‘good common world’, without consigning them too rapidly to one or another fixed essence (e.g. a ‘scientific fact’,

461

The contradictions and difficulties of such approach are made abundantly clear in reading von Schomberg (2006), who insists on the ‘political-normative’ dimension of invoking the precautionary principle for a particular policy field, while insisting on ‘scientific assessments’ for defining the ‘state of affairs’ and type of uncertainties involved.


303

a ‘public perception’, etc.) – before considering further policy steps 462 . But of course, perplexity can never become a permanent fixture (we have to doubt, yet still act), nor will it ever be absolute (we have no reason to doubt everything at once)463 . We have to make sure that the selection of propositions to take into account show at least the possibility of contributing to the dynamics of sustainable (energy) development. As we have repeatedly argued in the previous chapters, the stage of structuring problems and propositions for sustainability is of crucial importance, since it is one of the determining factors for the following steps to be followed (e.g. selection of options, evaluation of policy measures, monitoring of the relevant trends, etc.). In particular, ‘productive’ problem structuring will need to avoid certain pitfalls – we refer here to the arguments already developed in chapter 1 (Section 2.3), which we will briefly reiterate here for the sake of clarity. 3.1.2.1 Productive problem structuring Firstly, for a problem (or problematic situation) to be taken into account, there has to be some indication that it is really amenable to policy intervention – i.e. there has to be at least some plausible hypothesis linking the problem to certain causes which have to fall under the competences of political decision making (in a broad sense, e.g. one should not a priori downplay the importance of ‘symbolic’ policy measures – cf. infra) 464 . Secondly, at the other extreme, one should not limit political attention too quickly to those problems which are already ‘structured’ in a technical sense – e.g. a clear description in terms of quantitative indicators exists, causal chains are firmly established, possible policy measures are clearly defined, etc. Thirdly, one should also not assume too quickly that a vision exists which is capable of providing a normative background for the experienced problems – i.e. transforming the problematic situation into a ‘bridgeable gap’ by setting ‘targets’ or ‘benchmarks’ 465 . As explained earlier, reverting too rapidly to a vision exposes policy makers to the risk of committing type I, II, or III errors (cf. Chapter 2). Rather, the list of ‘actants’ to be taken into account should be determined by thinking back and forth between problem scanning and vision assessment (cf. infra), with ultimate policy goals being set only at a later stage of the decision process (cf. Section 3.3).

462

Of course, so far we have only set out the more formal normative/juridical resources that can be drawn upon for introducing ‘perplexity’. Section 3.5 provides some more practical arrangements for detecting matters of concern. 463 Of some of the actants we know reasonably well what they ‘do’ (e.g. ‘a wind farm emits virtually no greenhouse gases over its entire lifetime’); other influences or links between actants are less clear (e.g. ‘liberalisation of the electricity market will lead to a new kind of medievalism’) – we refer here to examples taken from our interviews (cf. Chapter 4). 464 For the difference between ‘unfounded suspicion’ and ‘reasonable cause for concern’ we refer to section 3.3. 465 Regarding this requirement, we firmly disagree with the co-design method proposed by the Belgian ‘Centre for sustainable development’ (CDO – University of Ghent) (briefly explained in Boulanger et al. (2003, pp. 123-139)). The co-design method consists of four steps, the first of which is called ‘concept and vision forming’, and is supposed to result in a consensual ‘vision text’ consisting of policy intentions regarding sustainable development for all relevant policy sectors. Such vision is supposed to be derived from the Brundtland report (WCED 1987). The strangest feature of this approach (variants of which can be found in Valentin and Spangenberg (1999) and Grunwald (2004)) is that it assumes that a ‘collective world’ already exists before the laborious work of construction has taken place! We suspect that this assumption can only lead to a very superfluous form of ‘rhetorical closure’...

304

Chapter 5

So what is the right road to follow then? According to us, the answer lies in reserving a much more modest role for visions and technical problem factoring at the outset of the process. Defining the relevant ‘actants’ to be taken into account is a logical first step relating to all further decisions, prior to actual ‘measurement’, and without assuming we have one vision capable of integrating all actants into a coherent whole 466 . Of course, these ‘actants’ will have to ‘say’ something relevant to sustainability, but this relevance should not be interpreted in a too restrictive sense to begin with. In the following section, we argue in favour of a connection with the literature on finding adequate indicators for sustainable development, as in sociological research indicators can generally be regarded as something which renders abstract concepts (such as sustainable development) able to ‘speak’ more clearly 467 . 3.1.2.2 Sustainable energy indicators Indicator sets for sustainable development (in general) are suitable for meeting the following purposes (Boulanger et al. 2003, p. 124): a) allowing measurement and learning at a strategic level; b) showing evolutions of ‘factors and actors’ (equivalent to Latour’s ‘actants’) relevant for the context in which a particular political or administrative organ is active; c) giving support for strategic policy of this political/administrative organ by giving an input in the strategic policy planning cycle and the societal debate at large 468 . Thus, they are crucial enabling elements for facilitating the ‘collective learning experience’ we envisage with our model process, in the sense that they allow the development of a protocol for revealing whether improvement towards more sustainability is actually achieved. Developing indicators generally proceeds along the following lines 469 : • Review of long-term (and immediate) objectives and activities: Diagnoses and appraisals of the present situation or of observed developments from the point of view of sustainability should not orient themselves solely on singular or sectoral criteria (according to the principle of integration), but have to take all relevant evaluative criteria in the various dimensions of sustainability into consideration (Grunwald 2004). Inevitably, normative considerations come into play. Hence, as a correlative, the question immediately poses itself as to what extent these criteria can be formulated by

466

This is another way of saying that the reasoning of actors involved in the process of building the ‘sustainability commonwealth’ will tend to proliferate across the different commonwealths of Boltanski and Thévenot’s model. 467 According to Blalock (1971), quoted in Boulanger et al. (2003, p. 8): “…Sociological theorists often use concepts that are formulated at a rather high level of abstraction. These are quite different from the variables that are the stock-in-trade of empirical sociologists. In attempting to bridge the gap between the two levels, we have tended to refer to the latter kinds of variables as indicators of the former concepts…”. 468 Guidance on the construction of general indicators for sustainable development can for instance be found in UNCSD (1996), Bossel (1999), and Boulanger et al. (2003); more information on specific indicators for monitoring sustainable energy production and use can be found in Boonekamp (2002) and IAEA et al. (2005). 469 The following paragraph is essentially an annotated version of the ‘Participatory Assessment, Monitoring and Evaluation’ (PAME) method proposed in Slocum (2003, pp. 107-115). We merely set these out here in quite general terms; section 5 will provide more details on who will contribute to the different steps, whilst chapter 6 and 7 will make a concrete start with the construction of an indicator set for sustainable energy.


305

experts or to what extent they have to be legitimised through political processes 470 . Criteria for sustainable development will certainly have to take into account political commitments already made 471 . Moreover, a network analysis will have to reveal which actants are mobilised by the different parties in order to support their actions. Here, it might be useful to make use of scenario techniques (cf. Section 3.2) to think about longer term goals in a more systematic fashion; • Choosing or designing indicators: taking the actants that were identified in a first phase, direct and indirect indicators that are able to ‘speak faithfully’ for the actants need to be identified 472 . Such indicators can be ‘direct’ (i.e. pieces of information that expressly related to what is being measured – e.g. the global warming impact of energyconsuming activities is of course directly dependent on the amount of GHG emitted) or ‘indirect’ (i.e. pieces of information chosen to serve as an answer to questions that are difficult to ‘measure’ – e.g. the fraction of electricity cross-border exchanges as an indicator for the functioning of a free European electricity market); • Data collection: the next step involves for each indicator the identification of what information sources are already available, which sources to choose and how to obtain the necessary missing pieces of information. If information is not readily available, it must be decided which information gathering tool will be used. This also includes addressing the question who is most competent to ‘speak’ for particular relevant actants (e.g. in principle, the CREG for the degree of liberalisation of energy markets). Data collection deals mostly with the more technical aspect of quality of databases in function of what has to be ‘measured’. Pragmatic considerations (e.g. amount of time or resources needed for data collection) also can be of importance. Data collection is also influenced by the culture of the institute which is charged with the data collection (e.g. a specific task force, direct data collection for a specific problem, routine bureaucratic data collection, etc.); • Data analysis: analysis is concerned with understanding how the different partial indicators relate to the ‘whole’ (i.e. ensuring sustainable development, whatever that means in detail). In the process of analysis, one can also take note of similarities or overlaps between certain indicators (i.e. the indicators say more or less the same thing) or relate pieces of information to establish relationships between them. Analysis is also concerned with the purpose the indicator set is supposed to serve (e.g. a broad

470

E.g. through reliance on existing (technical) indicators of sustainable development, as proposed by the UNCSD (the UN commission on sustainable development), the OECD, EUROSTAT (the EU statistical office), etc. – for an overview, see Boulanger et al. (2003), and more specifically in the context of energy policy Boonekamp (2002). 471 E.g. as a signatory to Agenda 21, a commitment is expressed by the Belgian state to the so-called Rio principles – e.g. investigating the consequences of proposed development paths in an inter- and intragenerational perspective (in particular concerning impacts on relations with developing countries, e.g. international security, development aid, humanitarian aid, international environmental protection, etc.), institutional opportunities for participation, etc. (cf. Chapter 1 – Section 2.3). 472 ‘Indicators’ need not be numbers – maps, historical accounts, personal statements, etc. can all function as indicators.

306

Chapter 5

communication instrument vs. guidance for specific policy measures) as this will also determine the level of aggregation sought after 473 ; • Review and feedback: review then relates to the processes of quality control with respect to the particular indicator in question, with attention towards the particular interactions with the socio-political context (e.g. ‘extended’ peer review, external or internal peer review, independency of review, etc.) (cf. Section 4). It is useful to keep in mind that the problem structuring phase should be iterative and part of the larger planning process we propose. Inevitably, uncertainties in the knowledge base will present themselves which cannot simply be resolved through more research or through the comparison of existing research results. Extra care should be taken to ensure that the data collected answer the real questions being asked: all too often, a particular method of data collection is chosen because it is readily available and/or because certain groups might be very experienced with certain methods. Developing the knowledge base should rather become an inherent part of the policy process, certainly when the required knowledge is still in its infant phases (as can be expected in the first implementation of the model process). Interaction processes with the ‘external’ world then become crucial for the robustness of the findings, as these concern the framing of the problem, the choice of methods, the design of a strategy to gather the data, the review and interpretation of results, the function of the results in the policy arena, etc. Even under these conditions, there will be remaining uncertainties which are very difficult to verify through the use of experimental protocols (e.g. measuring the influence of advertising or educational packages on consumer behaviour). Such difficulties of experimental verification should not be used per se as a reason for postponing policy intervention (e.g. a prohibition on advertising for the promotion of wasteful behaviour). Conversely, focusing exclusively on such ‘symbolic’ interventions (i.e. policy interventions with consequences that are hard to predict to some extent) is likely to damage the credibility and legitimacy of the governance model. This observation is of particular relevance to policy measures aimed at interventions on the demand side of the energy equation – a long neglected area of policy intervention where significant and demonstrable progress can be made (Fraunhofer Institute 2003). All in all, the importance of having a database of indicators for sustainable energy provision lies in its function as a ‘boundary object’ between policy and science. The concept of a ‘boundary object’ was introduced in social studies of science to describe how members of different ‘social worlds’ 474 manage to cooperate successfully despite their very different viewpoints and interests (the different roles stemming from different ‘social worlds’ are described in section 5). Broadly speaking, a boundary object should be both

473

Examples of very aggregated indicators include the ‘Human Development Index’ (HDI) or the ‘Ecological Footprint’ discussed earlier. 474 The term ‘social world’ is defined as “…a group with shared commitments to the pursuit of a common task, who develop ideologies to define their work and who accumulate diverse resources needed to get the job done…” (Gieryn 1995, p. 412).


307

plastic enough to adapt to the needs and constraints as experienced by the different parties involved in negotiating the ‘sustainability commonwealth’, while still being robust enough to maintain a common identity. Boundary objects thus acquire different meanings in different social worlds, but their structure is still common enough to more than one world in order to make them recognisable – in other words, there are a means of translation (Ishii 2002). For instance, one important function of indicators for sustainable energy would be to protect scientists on one side from accusation of bias or illegitimacy (because the indicators are clearly situated as ‘official’ objects of advisory science, and hence no confusion with ‘pure’ research science is possible), while protecting policy makers on the other hand from accusations of allowing technocratic intrusions into their domain of competency 475 . This means that indicators have to fulfil conditions of both scientific and political legitimacy. Scientific legitimacy for sustainability indicators is generally predicated on (Boulanger et al. 2003): a) pertinence of the indicator (i.e. does the indicator or set of indicators adequately represent the full scope of the sustainability dimension(s) in question?); b) fidelity (i.e. is the measure chosen for a particular indicator a good representation of its content?); c) sensibility (i.e. is the measure chosen for a particular indicator sensible to small changes); and d) specificity (i.e. is the measure chosen for a particular indicator only weakly correlated to other measures ‘speaking’ for the indicator?). Political legitimacy depends rather on the degree to which the purposes articulated by the indicators are underpinned by socially accepted and/or acceptable norms. It is clear that such claims will always be essentially contestable – hence, the vital contribution of decision takers for achieving closure in this domain (cf. Section 5).

3.2

Definition, comparison and selection of options

The tasks outlined in the previous section correspond roughly to what Latour (2004a) has called the ‘power to take into account’ – i.e. ensuring that the number of ‘actants’ that could contribute to building the sustainability commonwealth is not arbitrarily shortcircuited. As has been argued before, for the more ‘common’ problem types – i.e. problems where the relevant ethical principles and scientific foundations are relatively well-known – the prevailing tools of problem-solving are generally sufficient (e.g. costbenefit analysis, application of clear-cut and generally accepted decision principles, etc.). Solutions can then be readily derived from an adequate and thorough problem description.

475

There are other ‘social worlds’ or ‘research cultures’ involved in our model process (cf. Section 5) to which boundary objects should be adapted. The example of politics and (natural) sciences is just given here for illustrative purposes.

308

Chapter 5

However, the more complex issue of sustainable energy provision requires more creative solutions in terms of (technological) innovations or the invention of new vocabularies which provide new meanings and open up new perspectives. The problem we have to take on in this next step of our model process is one of actually discovering and constructing a hierarchy for the demands voiced by the different actants identified in the problem structuring phase, without any guarantee that such hierarchy is simply evident from the application of ‘well-established’ facts and values (because the information is either uncertain, unreliable, inconsistent, or a combination of all of the above at once) 476 , 477 . For performing this task, we propose to revisit the research tradition of scenario building 478 . 3.2.1

Constructing scenarios

Although scenario-building as a systematic way of thinking about the future has already a long history of application in decision-making context, a renewed interest in response to the challenges presented by sustainable development is easily explicable in view of the subsidiary sustainability principles of integrating across policy issues, presenting a longterm view, and promoting participation (Raskin et al. 2002) 479 . In general terms, scenarios may be thought of as coherent and plausible stories, usually told in words and/or numbers, about the possible and/or desirable co-evolution of ‘the collectivity’ (commonly described in terms of the ‘social world’ and the ‘environment’). Scenario building corroborates very well the constructivist tradition we adhere to in this dissertation, as it corresponds to what Keulartz et al. (2004, p. 20) call the constructivist task of ‘dramatic rehearsal’ – i.e. the imagining of a plurality of possible futures and the way that leads to their realisation. ‘Dramatic’ should be understood in three senses: in a concern with the interaction of personalities, a concern with a plot (e.g. creative redescriptions, new narratives), and a concern for open-endedness 480 . Most important for our purposes is that scenarios – just as the list of actants and associated indicators – have to fulfil their role as ‘boundary objects’. 476

Whilst not precluding the possibility that for some of the actants no place can be found in the hierarchy – hence, these are declared ‘adversaries’ for the time being (cf. Chapter 2). 477 Cf. our discussion of sustainability as a ‘manifest image’, a ‘vision’ or a ‘policy target’ in chapter 1 – section 2.1. The task of finding a hierarchy of actants corresponds to the construction of a vision (or rather, ‘visions’ in plural) on sustainability. 478 We are reversing here the sequencing of the ‘hierarchisation’ and ‘scenarisation’ tasks proposed by Latour (2004a). Putting the requirement of establishing a hierarchy between the new ‘actants’ and the ones already established in institutions (e.g. laws, norms, scientific facts, etc.) before the one of developing a ‘scenarisation of the whole’ seems to us to be an unduly restrictive and conservative move. 479 The first use of scenario-based methods for decision making is generally ascribed to Kahn and Wiener, who in the late ‘60s explored the consequences of nuclear proliferation (Mieg 2002). Systems’ modelling is another antecedent to contemporary scenario analysis, with a controversial first effort at forecasting the economic pressures on the environment and resource constraints made by the ‘Club of Rome’ (Meadows et al. 1972). Another early strand of scenario work focused on envisioning desirable futures, particularly in the energy fields (building on the so-called ‘soft energy paths’ advocated by Lovins (1977) – see e.g. Robinson (1982)). After Brundtland report (WCED 1987) and the Rio Conference (UNCED 1992), scenario studies again took a prominent position in the debate as a means to integrate across themes such as climate change, water scarcity, public health, etc. (see e.g. the ‘integrated assessment’ approach advocated by Rotmans and de Vries (1997)). In particular, scenarios play an important role in the global climate change regime, as the ‘International Panel on Climate Change’ (IPCC) relies on increasingly more sophisticated greenhouse gas emissions scenario studies for its recommendations towards policy makers (Mieg 2002). 480 Hence, the power of Latour’s concept of an ‘actant’, extending the category of the ‘actor’ (and the associated notion of agency) also to ‘things’ or ‘non-human others’.


309

Scenarios are ‘good’ if they are relevant to the concerns of decision makers (i.e. they show possibilities for practical intervention and are politically legitimate), and if they are able to withstand scrutiny by scientists (i.e. they have to be based on an adequate analysis of the present situation and the range of possible futures implied by this present situation) 481 . However, literature on the subject provides no conclusive guidelines on how this goal can be realised: in view of the above characterisation of scenarios as contributing to a ‘dramatic rehearsal’, it will come as no surprise that scenario analysis is a still evolving field of scientific inquiry which has not yet been codified into a common set of definitions and procedures 482 . Nevertheless, the US ‘board on sustainable development’ (USNAS 1999) identifies the following broad (and most commonly applied) approaches to scenario building: qualitative consultation among experts in study panels (i.e. the so-called ‘Delphi’ approach); formal elicitation of expert judgment in forms such as probability distributions; creation of structured and internally consistent narratives or scenarios; various forms of ‘gaming’ exercises; formal extrapolation of past trends using statistical methods; and a wide variety of different kinds of causal modelling (see also Dammers 2000, Annex 2). In what follows (cf. Chapters 6 and 7), we will focus on the scenario method, though this method is flexible enough to also draw upon the other approaches. Within the confines of the scenario method, many relevant distinctions can still be made. For instance, the difference between quantitative (modelling) and qualitative (narrative) traditions of scenario building can be underscored (the former approach prevails in the field of energy). Earlier attempts at forecasting (prediction) have proven to be largely unsuccessful (particularly in the case of long-term scenarios) and are increasingly being abandoned by scenario builders – although there still appear to exist some expectations of correct prediction on the part of policy makers 483 . But for our present purposes, the most relevant distinction to be made is the one between primarily descriptive or exploratory scenarios – i.e. scenarios describing possible developments starting from what we know about current conditions and trends, and primarily normative, anticipatory or backcasting 481

Though this does not mean that they have to coincide with the current beliefs of decision makers. On the contrary, challenging existing assumptions – hence the crucial contribution of experts in humanities to our model process (cf. Section 5). 482 For methodological and theoretical considerations of scenario-building as a boundary practice caught between science and policy, see e.g. Dammers (2000), Blass (2003) and Harries (2003). Such evaluations of scenario-building methods usually either follow a theoretical route (e.g. using theories of organisational learning, policy analysis, etc.) or a case-study approach. A full experimental testing of theoretical insights remains elusive because controlled experiments examining either objective (e.g. correct prediction) or subjective (e.g. organisational learning) measures of performance are bound to miss the delicate real-world interplay between e.g. the methods used, the policy context, organisational factors, etc. Theory-building is also severely hampered by the fact that most accounts of scenario exercises do not occur in the open scientific literature, and hence are prone to the inherent biases of self reporting (Harries 2003). 483 For abundant examples of incorrect predictions in the energy policy field, see Smil (2000). Smil’s historical overview of scenario practices (and their use in political contexts) leaves him with the rather glum conclusion (p. 260): “…Having no illusions about the usefulness of my advise I will, nevertheless, offer the following conjoined axioms: no truly long-range forecast can be correct in all of its key aspects; most of these forecasts will be wrong in both quantitative and qualitative terms; some forecasts may get a few quantitaties approximately right, but they will miss the real qualities arising from subtly to profoundly altered wholes…”.

310

Chapter 5

scenarios – i.e. scenarios which are constructed to lead to a future that is afforded a specific subjective value by the scenario developers. Neither of these two types is ‘value free’, since both embody extra-scientific judgments, for example about ‘reasonable’ assumptions. However, they differ in terms of overall purpose. That is, the choice between exploratory and anticipatory approaches depends on the objectives of the scenario development exercise. Anticipatory scenarios represent organised attempts at evaluating the feasibility and consequences of trying to achieve certain desired outcomes (or avoid the risks of undesirable ones). Exploratory (or ‘what-if’ analysis), on the other hand, tries to articulate different plausible future outcomes, and explore their consequences. The emphasis is mostly on prioritising technological choices, the analysis is performed in a relatively closed process by technically or economically schooled experts, and the government (or administrative bodies) mostly assumes the role of client (they ‘place an order’ for the analysis). In the following sections, we will draw upon these distinctions for demonstrating the potential of scenario analysis as a flexible tool for addressing the core questions that have to be answered in our model process. We propose to split up the next task in the policy development cycle – that of finding, comparing and selecting policy options that further the cause of sustainability in the energy field – in a strategic choice for a particular policy framework, and a tactical choice of potential policy measures. The question of finding an adequate framework is concerned with setting out the principles which can or should be used for establishing a hierarchy between the actants claiming a right to participate in the constitution of the ‘sustainability commonwealth’, while the tactical level is concerned with finding the kind of policy measures which actually implement these principles in practice. In what follows, we will try to show how scenario analyses of the anticipatory kind could be destined to pave the way for an informed discussion at the strategic level, with participation of a range of stakeholders, representing a wide range of preferred orders for the ‘sustainability commonwealth’. In particular, their major strength lies in the fact that they can bypass the difficulties engendered by a reference to particular scenarios as representing some form of objectivity and instead openly discuss the normative framings embedded in these scenarios 484 . In doing so, the scenarios can serve as vehicles for reflexively representing a collectivity to itself. On the other hand, when discussing the tactical level (set within the confines of the broader framework), exploratory scenarios can be useful for discussing how different ‘plausible’ trends could work out on a short term, and then explore ways how these can be taken into account in the design of practical policy measures. Thus, both approaches in fact respond to different questions raised by the demands for more sustainable forms of energy governance, and should clearly be presented 484

In spite of the fact that foresight exercises clearly fall beyond the domain of ‘traditional science’ (since the results of foresight exercises cannot be tested empirically against ‘hard facts’), debates between competing policy advocacy groups are still largely framed in terms of the ‘truth content’ of such exercises (with each group having ‘their’ scenario at hand – see e.g. the debate on the validity of the scenario exercises in the context of the AMPERE report reconstructed in chapter 4). While taking refuge in ‘pure scientific objectivity’ of course enables policy makers or stakeholders to legitimise their positions, such debates all too often lead to endless (and fruitless) oppositions that stand in the way of the learning practice we envisage.


311

as such. In any case, it is important that both the strategic and the tactical questions are explicitly dealt with in our model process (a requirement which in itself is already quite challenging to ‘normal’ political practice, where the strategic level is often not explicitly discussed), whilst also recognising that scenario exercises cannot provide every answer. Their role can only be complementary to a daunting amount of non-scenario work required for building the ‘sustainability commonwealth’ (cf. Section 3.4). 3.2.2

Definition, comparison and selection of strategic options

In the previous chapters, it has been repeatedly argued that having the sustainability principle as a ‘constitutional’ (in a figurative sense) principle does not fully exhaust its political and practical importance. Rather, depending on the area in which it will be applied, it will result in quite different types of policies with a range of possible measures, which are all in need of a proper justification. At the strategic level, we are concerned with setting out the broad lines – i.e. the choice of selection rules to be applied, the choice of boundary conditions, priority setting, etc. – of what the ‘sustainability commonwealth’ should actually look like in the practical case under investigation (i.e. energy policy). If we choose not to leave such strategic choices over to the outcomes of political bargaining or place our faith in the ‘invisible’ hand of incrementalism, broad ‘scenarisations’ and foresight exercises are a crucial support for the wider deliberations we propose. With hindsight, it becomes clear that we have already encountered several scenarisations in the course of our investigations. In chapter 1 (Section 3.2), we discussed (and criticised) Beck’s preferred scenario of ‘differential politics’ in answer to the ambiguities of contemporary risk society as a covert way of introducing a thoroughly normative stance under the cloak of objectivity. The ‘three-capital model’ (Chapter 1 – Section 3.5) can be seen as a way of casting the choices implied by a strive for a more sustainable development into the scenario of a manager deciding about the right allocation of different forms of capital. Chapter 3 provided us with other examples of scenarisations for conceiving strategies for sustainability – e.g. with either a principle of political expedience serving as the integrating element (in Klinke and Renn’s proposal for a risk classification and management scheme discussed in section 2.2) or an economically efficient allocation of resources, including health and the environment (in the theory of external costs discussed throughout the entire chapter). Now, whatever other criteria are used to defend the value of one scenario over another, it is clear that ‘validity’ or ‘truth content’ – in view of the generally rather abstract level of discussion, where whole discourses, political systems or broad technological development can be put up for revision – takes a backdrop to other criteria (which we will discuss later on in this section) 485 . The real world can of course never be confused with a scenario! As argued in chapter 1, such scenarios should be used as heuristics for the testing of

485

Of course, scenarios should not be unduly unrealistic, but this condition will generally be realised because political actors or stakeholders in the debate have an interest (that is, in order to be convincing to other parties) in presenting ‘their’ favoured scenario in a realistic mode. However, the point is that in general, we have no ‘objective’ criteria (e.g. in terms of a ‘meta-scenario’) for judging the truth content of scenarios.

312

Chapter 5

hypotheses or the detection of ‘matters of concern’, and not as explanations. However, in no way is this statement meant to imply that engaging in scenario building is an innocuous activity. On the contrary, scenarios can (and should) also help in attenuating the consequences of hegemonic scientific/political practices of the ‘business-as-usual’ scenario. For instance, many scenarios developed by environmentalist ‘think tanks’ (e.g. LTI-Research Group 1998; Fischedick et al. 2002) have based their view of sustainable development on a long-term trajectory leading to the fulfilment of stringent equity and environmental impact parameters 486 , 487 . Such normative scenarios, outlining what should happen rather than what is likely to happen, may be useful providing, of course, that they always serve a probing or heuristic function, rather than being used in adversarial debates by ‘true believers’ (as seems to be the case for the time being). Irrespective of the personal estimations one might have about the feasibility or even desirability of such ambitious goals, it is clear that they will be unlikely to become a political reality while the policyrelevant knowledge upon which its implementation would rest is incapable of envisaging the possibility for it. As already put forward in actor-network theory (and experimentally shown in many cases – for an overview, see e.g. Rip et al. 1995), a new technological system will only succeed when it is able to attract a whole universe: a network of sociomaterial relationships has to be put together, persuaded and enlisted; and having a coherent and/or convincing vision or scenario at hand is often of vital importance in such processes (Chermack and van der Merwe 2003) 488 . Discounting such scenarios on the basis that they lack reality or involve value judgments thus acts to constrain the perspectives encountered in decision making and also conditions which knowledge-producing practices will be favoured. Thus, ‘due process’ becomes a matter not only of guaranteeing ‘political pluralism’ but also ‘epistemological pluralism’. If not their truth content, what else distinguishes a ‘good’ scenario exercise from a ‘bad’ one? Obvious other criteria include consistency (i.e. applying the same principle of integration to all the actants involved in the scenario) and coherence (i.e. one scenario should correspond to one proposed hierarchy of actants). But most importantly, for scenario exercises to be effective in the context of our proposed model process they should be able to integrate across the wide scope of actants and (stakeholder) perspectives clamouring for a place in the ‘sustainability commonwealth’; as a correlative, they should provide enough contrasting evaluations of the situation at hand (and also challenge common assumptions); and they should support the collective learning experience we envisage. Now, over the previous chapters we have attempted to demonstrate from a more

486

Such long-term ‘backcasting’ scenarios have not yet been developed for the Belgian context. E.g. an 80% reduction in greenhouse gas emissions in industrial countries by the year 2050 is commonly advanced on the basis of an ‘acceptable’ climate change impact combined with equal global per capita emissions entitlements. 488 In the context of energy systems, this can be achieved by situating the time horizon for anticipatory scenario building in a distant future (e.g. 2050 is commonly used in the context of energy infrastructures), investigating options for the future in which most of the current equipment will have been replaced becomes possible, leaving the possibility of politically influencing the choice of this new equipment. Hence such scenarios deal with systemic rather than marginal changes. 487


313

theoretically-inspired perspective that Boltanski and Thévenot’s ‘commonwealth model’ is an excellent candidate for meeting these requirements. Chapters 6 and 7 will be concerned with a further empirical investigation of this statement in the concrete context of Belgian energy policy; therefore, in the remainder of this section, we will limit ourselves to some broad considerations and first indications of how the model could be used to support collective learning 489 . Chermack and van der Merwe (2003) – in a broad review of theories of individual, organisational, cultural-historical and evolutionary perspectives to learning – propose two interesting concepts for understanding the contribution of scenario exercises. The first of these is the ‘zone of proximal development’. Briefly, the ‘zone of proximal development’ is the space between the tasks an individual can accomplish on his/her own, and that which he/she can accomplish with some guidance. Chermack and van der Merwe (2003, p. 449) also refer to this zone as “…the place where the client’s newly acquired, but as yet disorganised concepts meet the logic of experienced learning…”. The other concept, ‘scaffolding’, extends the perspective on learning towards mutual dialogue. ‘Scaffolding’ is then meant to indicate the process whereby the facilitator of a scenario exercise provides a ‘scaffold’ (or framework) that allows the participants in the exercise to relate new experiences or concepts to existing ones – thus stretching the ‘zone of proximal development’ to its upper limit. As already stated, we believe that Boltanski and Thévenot’s commonwealths are excellent candidates for taking up the role of such ‘scenario scaffolds’ 490 . However, besides having the necessary ‘scaffolds’ at hand, even more importantly the success of a scenario exercise also crucially depends on having the right kind of ‘construction site’ or ‘building environment’ which encourages the participants engaged in

489

There is comparatively little literature on the reception of scenarios by policy makers. Mostly, the importance of institutional factors, especially the measure of pluralism or corporatism, is stressed. Dammers (2000) is one of the rare authors who has done systematic empirical research into other factors influencing learning on the policy maker’s side like the qualities of the scenarios (scope, contrast, and validity – defined by Dammers as the plausibility and explanatory value attributed by policy makers to the scenario in question), the degree of dissemination (canals, meeting information needs of policy makers, timing), conflicts among policy makers (moderate vs. high conflict), and exogenous factors (trends, events, etc.). The most important conclusions of his empirical research are: reframing (a change in the core values) has more chance of occurring when members of stakeholder groups participate in an ‘open forum’ (i.e. they participate on their own responsibility, not bound by organisations); exogenous factors which are perceived as a threat (e.g. oil crises) are much more stimulating towards reframing than events which are not perceived as a threat (e.g. falling oil prices); scenarios mainly play a role in reframing when policy makers want to defend a change in their core values and opinions post factum; scenarios mainly play a role in stabilising the relationships between groups who already more or less share the same opinions (by making formal reference to the same scenario studies); policy makers tend to rely on their own expertise and other information sources (rather than scenario studies) for selecting between ‘relevant’ information and ‘noise’; and finally, most policy makers did not find any ‘new messages’ in scenario studies, and they saw a lack of consideration for future changes in values or opinions as one of the major drawbacks of scenario studies. However, Dammers’s case studies are mainly (almost exclusively) based on short-term predictive decision-support scenarios. 490 In chapter 4, we have already given a first indication by using the ‘commonwealth model’ as a structuring ‘scaffold’ for the information collected in the interviews with members of the FRDO. Chapters 2 and 3 also included examples of the model’s outstanding flexibility and scope in accommodating both ‘factual’ and ‘normative’ statements and theories.

314

Chapter 5

the scenario exercise to openly acknowledge uncertainties, indeterminacy and the various choice dilemmas inevitably involved. The deliberative body mandated to draw up the scenarios should be encouraged to define itself as a collectivity whose members are committed to crafting solutions to shared problems. The best way to do so would, in our opinion, be to situate this body at a convenient ‘distance’ (in terms of the line of accountability) from direct legislative bodies (whose members often see it as a task to promote preformulated partisan agendas) 491 . Although it is of course unrealistic to suppose that ideological cleavages will not surface within the deliberative group, one should strive to stress the group’s commitment to the common purpose of sustainable development (albeit that this common purpose will be worked out in a pluralistic way). The framing of the scenario exercise might also be helpful in this regard. Defining the purpose of the scenarios as one of setting out the broad strategic context for the development of more concrete tactical measures perhaps serves to decrease direct political relevance, but conversely is likely to be more forthcoming towards participation by (generalist) experts (cf. Section 5) 492 . Thus, in order to introduce the specific mindset we are looking for (i.e. dedication to a common purpose, yet acknowledging a plurality of visions), one could envisage a certain number of small groups to devise, elaborate and defend a plausible implementation of (a selection of) the different commonwealths that would deal with the problem of sustainable energy production and consumption. Each of these groups should also receive support (in terms of resources and support by experts) to adequately develop ‘their’ scenario. However, in order not to fall back into an unproductive adversarial juxtaposition of scenarios, the groups should start from a common basis. This basis can be provided by a previous consensus on the list of actants to be taken into account (i.e. in the problem structuring phase) and the acceptance of the commonwealth model as a general framework for the scenario exercise. This acceptance could for instance start from making the participants in the scenario exercise familiar with the model, e.g. by giving examples of how the grammar of the commonwealths can be recognised in historical examples of political decision making (as we have tried to do in 491

Besides being desirable in view of the requirements for our model process, Fiske and Tetlock (1997) present an argument why such a deliberative body might also actually have chances of survival in the harsh political reality. Fiske and Tetlock observe that a commonly used mechanism for dealing with ‘taboo trade-offs’ in multiparty systems with regular rotation of power is to bury the taboo trade-offs in bureaucratic, regulatory or judicial enclaves ‘where the light of public scrutiny rarely extends’ (p. 292). This mechanism seems to be predicated on tacit transideological agreements between politicians who might reasonably expect to alternate between government and opposition. The hypothesis is then that, when sitting on the opposition bench, politicians will not have very strong incentives to generate anger in the public opinion about certain trade-offs, because shortly after, they could face the same trade-offs in a government role. Thus, there might be some incentive to ‘de-politicise’ the issue. 492 Policy makers often like to rely on scientific authority for advancing particularly preferred options. Therefore, scenarios containing tactical decision ‘joints’ (i.e. active variables in the scenario that can be controlled by political intervention, e.g. introduction of an energy tax, prohibition of certain activities, etc.) perfectly match the needs of decision makers, both cognitively and politically. However, for scientists the development of such tactical scenarios is often more problematic, since it is virtually impossible to integrate all the uncertainties involved in political decision making (e.g. the effects of certain policy measures) in a model. Of course, frictions might also develop when scientists try to prescribe certain political options which are not particularly preferred by the policy makers. Thus, as Mieg (2002, p. 73) correctly observes, a successful cooperation between both spheres depends on increasing the possible options for both sides.


315

Laes et al. 2004c). This process of familiarising ideally leads to the establishment of the commonwealth model (and the resulting scenarios) as boundary objects, legitimate both to the scientific and the political world, through the development of a collective commitment to the validity of each type of strategy and to the multiplicity of reasonable options. In the next step, the set of scenarios can be used to test potential policy measures against the conditions contained in the scenarios 493 . 3.2.3

Definition, comparison and selection of tactical options

Whereas the previous steps in the model process served to render the inevitably unstructured problems posed by demands for more sustainable forms of energy provision more tractable to policy makers, the present task is to actually define, compare and select policy measures against the background of the wider strategic considerations. In contrast to the previous phase, political considerations (e.g. of feasibility, vested interests, recognised political objectives of the present government, ensuring public support for policy measures, etc.) will carry a greater weight in the deliberations. Still, of course, one should not abandon any hopes of a reasoned judgment. Again, we believe Boltanski and Thévenot’s commonwealth model can be very helpful in this respect. As already explained in chapter 1 (Section 3.1.4), reaching an agreement on the ‘legitimate’ commonwealth involved in a situation can take a number of possible forms. One (ideal) case is of course that people generally agree that the policy problem can be solved within the confines of one commonwealth (e.g. market instruments in the ‘market commonwealth’, a reliance on technically proven and feasible options in the ‘industrial commonwealth’, etc.). Another favourable case would be that a tactical policy option under consideration actually performs well according to the orders of ‘grandeur’ of the commonwealths deemed to be most relevant. Such clarification within one commonwealth, or a synthesis between a number of commonwealths, can be stimulated by encouraging the critical reflection on why policy makers might reasonably be expected to assimilate a given problem to a certain commonwealth. Here it would be helpful to find at least some valid ‘pure’ option for each commonwealth (e.g. advertising ‘responsible’ energy consumption behaviour in the ‘opinion commonwealth’). But, while being desirable (and also helpful to stakeholders or decision takers who wish to explain or justify to their respective constituents the devised solutions), the possibilities for finding such consensus options will generally be limited. It is much more likely that compromises have to be struck, which might be informed by a critical discussion of the problems each ‘pure’ policy solution might face when reasoning from another commonwealth. Such criticisms would preferably in the first place be framed as questions (which might be resolved in compromises, or left open for later justification), but it is also possible that participants in the deliberation express decisive refutations when the perception is that deeply-held moral sensitivities are being violated. Also, clearly there would be vast potential for lobbying and

493

Again, rather than the simple linear process set out here, one can imagine in reality the design group going back and forth between strategic and tactical considerations until some kind of ‘reflective equilibrium’ is reached.

316

Chapter 5

proposals based on particular interests (rather than ‘the common good’). However, this should not count as an argument against our proposal, especially since such behaviour can be expected in any body engaging in public or private decision making. The difference in this case is that such ‘interest-based behaviour’ would be made more visible and transparent; and to the extent that particular groups would do no more than just defend these interests they would most likely discredit themselves and damage their credibility in a broader sense. Now, we believe it to be a fair assessment that the contemporary tactical move towards a competitive structure for the electricity supply industry rests on the belief that Pareto optimisation – i.e. a compromise between the ‘industrial’ and the ‘market’ commonwealth in the sense of an efficient allocation of different resources – can be achieved. On the other hand, the centrally-planned predecessor under protected market conditions rested (at least in theory) on the optimisation stimulus provided by energy forecasting (through the tenyear investment plans) and the pursuit of minimum long-run marginal costs 494 . Thus, although the earlier planned framework appears to provide a structural antithesis to today’s ‘free market’, the two systems nevertheless bear a resemblance in the sense that they both conceive the solution to energy policy to lie in the adoption of an ‘optimal’ strategy. These similarities, despite the shifts in orientation over time (e.g. ranging from an emphasis on limiting dependence on oil in the ‘70s to the greater importance of environmental goals in the ‘90s) also extend to the general character and scope of energy policy making, which has remained remarkably constant over the years. Manipulation of prices, regulation of supply, accelerating the diffusion of energy efficient equipment and providing information: these remain the staple ingredients of policy analysis and practice. This becomes for instance clear when taking a closer look at the two most recent energy policy support documents developed in the Belgian context – the AMPERE report (2000) discussed at some length in chapter 4 and the report by the Fraunhofer Institute (2003) on the possibilities for meeting the Kyoto targets through a dedicated energy demand policy – who, despite significant differences in overall focus adhere to the same analysis space opened up by an industrial/market commonwealth framing 495 . Summarising the approach and orientation of the report by the Fraunhofer Institute (2003) which analyses the factors behind energy consumption and discusses policy approaches for achieving rational energy use, the common thread in the argumentation is that changes will come about by focusing policies on the end use market for energy and on the ‘end user’ (households, industrial users, etc.). Economic instruments and information are considered to be key elements to facilitating the diffusion of ever more efficient technologies or ‘rational’ behaviour, while ‘actants’ from other commonwealths (e.g. considerations of political expedience) are commonly identified as ‘barriers’ standing in the way of realising the ‘optimal’ solution. However, such individualistic focus on choice and energy demand (with national energy 494

Moreover in both planning approaches the long-term ‘backcasting’ view advocated under section 3.2.2 has remained conspicuously absent. 495 The AMPERE report being largely dedicated to a discussion of the supply side of the energy equation while the Fraunhofer Institute provided an in-depth analysis of the demand side.


317

consumption represented as the aggregation of individual practices) tends to omit to important entry points for energy governance, namely: a) the way energy ‘needs’ or ‘consumption habits’ are not merely given but rather generated as a result of a complex interplay between institutional, material, cultural, technological, corporate etc. networks, and b) the possible contributions of actors in the ‘middle’ territory between the level of national government and individual energy users – e.g. corporate players, NGO’s, trade unions, etc. To be sure, such factors do not easily lend themselves to the quantitative modelling supportive of the search for ‘optimal’ policy measures (and hence, cannot lead to the ‘hard’ predictions often favoured by experts, stakeholders and politicians alike), but omitting them altogether has the effect of reinforcing, rather than potentially challenging, underlying theories of change (with a concomitant framing of research programmes). Section 3.4 will include some first reflections on how this can be avoided. In short, from the above discussion it becomes clear that we cannot rely on the requirement of ‘optimality’ alone for defining tactical options. In an abstract sense (i.e. relevant to a number of possible contexts), it has to be replaced by the requirement of ‘proportionality’ enshrined in the European Commission’s guidelines for the application of the precautionary principle (EC 2000). According to community law, the principle of proportionality requires that measures adopted by community institutions should not exceed the limits of what is appropriate and necessary in order to obtain the legitimate objectives pursued by the legislation in question, and where there is a choice between several appropriate measures, recourse must be had to the least onerous, and the disadvantages caused must not be disproportionate to the aims pursued. As von Schomberg (2006) shows through an investigation of case law, this somewhat complicated phrase comes down to the general line of thinking that a proportionate application of the precautionary principle involves the least onerous measure while still attaining the legitimate objective. And, more importantly for our purposes, in case law it is considered that cost-benefit analysis is only one particular expression of the principle of proportionality. Thus, the outcome of such a ‘proportionality analysis’ could favour another option over the (in terms of a cost-benefit analysis) ‘optimal’ one, depending on the decision rule adopted. Such decision rules could range from e.g. maximin criteria (i.e. maximising the benefits of the least advantaged over the range of possible options) over ‘only minimising costs/disadvantages’ to other decision rules such as avoiding certain health effects or ‘irreversible’ effects. Whatever position one takes in such a debate, the choice made here is of course crucial for the design of possible policy measures and therefore should be made explicit – a requirement that is not always met in everyday political reality, as we have shown in chapter 4 496 . Thus, this phase in the model process should result in the identification and characterisation, for each possible policy option, of a range of possible outcomes. In 496

Of course, policy makers can prejudge the results of a cost-benefit or proportionality analysis and spell out the decision rules and types of measures in advance. In such cases, the policy maker takes on a heavier burden of justification.

318

Chapter 5

addition, when presenting policy options for consideration to the political decision maker, the robustness of the analysis behind this choice should be made transparent. This might be achieved by characterising the policy options according to e.g. the degree of reversibility or any other precautionary argument that might be deemed to be relevant (cf. Section 3.3.2), the reliability of the key assumptions and evidence on which the recommendation is based (cf. Section 4), etc. Also, it is very likely that on these issues a consensus will not always be possible; therefore, it is advisable to include the possibility to also formulate dissenting points of view.

3.3

Political decisions

No matter which governance model is followed – be it technocratic, based on some aggregation rule, or guided by a broader deliberative framework (as we propose here) – in the end, decisions have to be made; and in a democratic society, such decisions need to be justified. It is generally accepted that decisions should be informed by the results of some systematic form of assessment, e.g. concerning the structuring of policy problems and the strategic and tactical options under consideration (cf. supra). The question is, however, to what extent can ‘closure’ of the debate in the form of a political decision be justified solely on the basis of assessment? The crucial point emerging repeatedly in previous chapters and which bears on this question concerns the intrinsic and unavoidable open-endedness of the assessment process. No matter how complete or rigorous the analysis, nor how inclusive or exhaustive the associated consultation of involved ‘actants’, the determinants of a final decision will in many important respects remain essentially open (e.g. concerning the relative importance accorded to various assessment criteria or different appraisals of the ‘evidence’ presented in support of particular policy options). This is not to say that ‘anything goes’ – with any decision being essentially as good as any other – but simply that a range of alternative assumptions might be defended as equally reasonable under different, and equally valid, contexts or perspectives. The ‘business’ of making real regulatory decisions over technologies such as energy options will, in the end (and at some political level), require the exercise of individual or collective judgment on the part of the decision makers themselves 497 . In a democratic society, such judgments are justified not simply by reference to the preceding assessment process, but also crucially in terms of the credibility, trust and mandate enjoyed by the institutions and individuals responsible for the resulting decisions (following the more formal channels of governance). The final responsibility for defining the ‘good common world’ lies with institutional legitimacy and political accountability. However, in view of not prematurely closing off possibilities of collective learning (central to our model process), the ever provisional nature of the solution should be stressed. At present, having progressed further in our discussion, we believe we are in a position to offer some more guidance on what this means in practice. In the following 497

This task cannot be left to experts, for the simple reason that pragmatic conditions which constitute ‘proof’ for advocating certain standards of performance are different in a scientific context than in a political one. From a ‘purely’ scientific point of view, the dangers of nuclear power, for example, cannot be refuted by pointing out that there have been no serious incidents in Belgium (or the Western world) for some years. However, this can be a perfectly valid argument in a political arena.


319

sections, we will argue in favour of setting transformable standards for the collective (Section 3.3.1) 498 , and considering strategic factors such as flexibility, reversibility and diversity for the policy measures deemed necessary to attain these standards (Section 3.3.2) 499 . 3.3.1

The choice of standards

Arriving at a closure of the debate necessarily implies defining the boundaries of the collective, i.e. setting clear standards (informed by prior consultation) for separating the ‘good’ from the ‘bad’ actants. From a precautionary point of view however, the choice for transformable standards (defining the provisional borders of the collective) seems a logical one, in view of the importance of a learning process under conditions of uncertainty. A transformable standard is a standard that sets out a certain target without however tracing in detail every possible step leading in that direction. In addition, the target itself can be moving. However, this statement should immediately be qualified: while precautionary action on a general level is not aimed at setting ‘absolute’ standards (e.g. categorical bans of certain technologies), it certainly does not exclude such measures in particular cases. Decision makers need to negotiate the choice between having a precise and definite normative standard (without which it becomes of course impossible to draw a valid conclusion on the acceptability of a certain technology or practice) while at the same time leaving open the possibility that, in light of the acquisition of new knowledge, the standard has to be redefined. We concur with von Schomberg (2006) that the best way of negotiating this trade-off consists in choosing widely accepted and uncontroversial norms or standards which however remain open for discussion (in a later iteration of the collective learning process) in their precise practical implications. Moreover, appealing to such uncontroversial (but changing) standards offers the further advantage of affording the objectives long-term political authority, since they can be construed to appeal to governments over several terms of office regardless of their political persuasion. Examples of such widely accepted but ‘open-ended’ standards include e.g. a ‘sufficient demonstration of safety’ 500 , ‘comparison with best practices’ (i.e. the benchmarking approach) or even a reference to the ‘natural situation’ 501 for defining acceptable impacts (i.e. after an actual occurrence of the impact in question, nature would still be allowed to return to its original situation). Over time, as the collective goes through a number of iterations (and knowledge is accumulated), such ‘transformable’ standards can be defined ever more stringently, up to a categorical ban on certain technologies or practices. Of course, there can be no general rule concerning the level of evidence deemed to be sufficient for such bans. However, here

498

Here, we generally agree with von Schomberg (2006). The importance of these considerations is particularly stressed (and explored in greater detail than we can afford here) by Stirling (1999). 500 With expectations of safety varying over time – e.g. the mandatory ten-yearly reviews of the operating permits for the Belgian nuclear power plants in view of the evolving safety standards at the international level can be considered as an example of a precautionary regulation. 501 Making reference to a ‘natural situation’ is still a ‘transformable’ standard, since our perception of nature changes over time as scientific knowledge grows. 499

320

Chapter 5

again, we can point out some widely shared cultural rationality standards also embedded in precautionary legislations and jurisprudence. Considerations of proportionality guided by the principles of a broad cost-benefit analysis were already mentioned in section 3.2.2. A criterion of consistency – i.e. equal application of rules in equal cases – is another essential element of the principle of law. Since no objective criteria exist to determine a level of acceptable risk, it seems an obvious approach to refer to what has already been accepted in similar cases in the past and to consider whether there are any compelling reasons to deviate from this model. The requirements of consistency and non-discrimination in the application of the precautionary principle (EC 2000) for instance imply that it is not enough that some minority opinion in science advances a risk hypothesis (without providing empirical indications or possible theoretical foundations) that the majority cannot refute on the basis of existing experimental or theoretical knowledge. Such ‘blanket arguments’ (e.g. ‘there may be some hidden risks which have escaped scientific consideration so far’) – whilst not refutable on a strictly scientific basis – would apply to all technology all the time, and hence would lead to a general ban on technological innovation. The discerning element between ‘unfounded suspicion’ and ‘reason for concern’ is the availability (in the latter case) of an empirical (causal) protocol that would enable testing the mechanisms through which the supposed risks would materialise. The same holds true for preventing ‘worst-case scenarios’ – another case where the precautionary principle could be invoked to justify a total ban on certain activities. In fact, catastrophic scenarios can easily be construed for most technologies (along the lines of the now widely popularised version of chaos theory that ‘a flap of the butterfly’s wings can cause a tornado in a far-off place’). Nuclear power of course provides a counterexample: here, causal mechanisms that would lead to a potentially catastrophic release of radionuclides – providing certain initial conditions are met (e.g. a breakdown of all cooling systems, a breach in the containment structure, etc.) – are relatively well-known. Thus, one can certainly argue that a nuclear phase-out policy could be part of a precautionary regulatory framework for preventing possible catastrophic accidents (if this is considered to be a political priority). However, by setting a clear (and furthermore arbitrary) time schedule for such policy (e.g. a 40-year lifetime for the nuclear reactors in Belgium, which means they will be taken out of operation over the period 2015-2025), and furthermore prejudging the results of any proportionality analysis (as we argue is the case in the Belgian phase-out decision, cf. Chapter 4), politicians and policy makers take a heavy burden in justifying such policy measures 502 . In particular, the possible later ‘relaxing’ of the standards (e.g. when economic consequences become apparent or risks in other areas – e.g. GHG emissions – increase) might be damaging to the level of trust invested in the political system as a whole.

502

According to the phase-out law, this standard is only ‘transformable’ in case of a ‘force majeure’ threat to energy security.


3.3.2

321

The choice of potential measures

In addition to the ‘traditional’ instruments of energy policy (manipulation of prices, regulation of supply, accelerating the diffusion of energy efficient equipment, providing information, etc.) generally appraised on the basis of their contribution to an ‘optimal’ economic solution to energy problems (cf. Section 3.2.3), a precautionary practice draws the attention to some other ‘desirable’ characteristics (besides of course contributing to sustainable development, made operational in (transformable) standards), most notably the degree of reversibility, flexibility and diversity displayed by potential measures. Irreversibility can easily be explained in terms of the constructivist jargon. The constructivist approach acknowledges that there are different ways to interpret a technological innovation and also stresses the dynamics of the technological innovation process, which (particularly in the actor-network variant) is related to transformations in the way ‘actants’ (technologies, legislation, social groups, etc.) are articulated. While the exact nature of these transformations – described by Callon (1991, 1995) in terms of a fourstaged process of problematisation, creation of interest, enrolment and mobilisation – are of less importance for our present purposes, it is clear that as a technology iteratively goes through (some of) these stages over time, the new ‘hybrid socio-material practices’ will be constrained to some extent by the existing practices in the previous phase. New practices (implying theories and versions of the social and material world that differ from those that existed before) are possible, but nevertheless, because of the backdrop of existing practice such differences tend to be limited and the technology (and the associated network of actants) becomes increasingly solid and obdurate. Callon (1991, 1995) shows that the concept of irreversibility can be analysed in terms of increases and decreases in the convergence of actor-networks. A network is convergent if agreement is reached on the description of its content and on the list of actants involved; irreversibility is produced as a result of increasing convergence over time. Furthermore, Callon maintains that potential changes in the articulation of the network crucially depend on the nature of the coordination rules, determining who negotiates with whom, within which framework conditions and in which sequence. Thus, membership in the network becomes one of the most potent means of having effect on its development and (if membership becomes fixed Conversely, over time) of contributing to relatively irreversible situations 503 . irreversibilities can be avoided by opening up membership of the network to new actants. Complete reversibility of a technological or policy commitment is of course impossible, as there will always be some form of ‘lock-in’ (e.g. the capital investment in the

503

This is arguably the case for nuclear power programmes in different countries, which tend to be supported by a strongly corporatist political culture (see e.g. Roberts (1990) for a discussion of the U.K. and Laes et al. (2004c) for the Belgian context).

322

Chapter 5

technology needs to pay off) 504 . However, one can always strive for some measure of flexibility when considering policy options. Seen from the point of view of precaution and leaving open possibilities for collective learning, it is evident that the degree to which a particular (technological) strategy may be adapted to avoid adverse consequences as they become apparent constitutes a factor worthy of consideration. Flexibility in general can be operationalised in a number of different ways – e.g. a technological option will be more flexible if it can serve different purposes (e.g. coal power plants can be co-fired with biomass), if it meets different strategic objectives, if it is not overly dependent on other supporting technologies, etc. However, with regard to energy policy, a higher priority for flexible solutions in itself seems to imply that greater attention should be given to demandside options, as the supply system (both generation and transmission) typically consists of large and capital-intensive plants and installations. Consequently, supply-side policy initiatives will likely bear their fruits in the medium to long term, whilst policy adjustments levelled at the demand side of the energy equation are more likely to have effect on a shorter notice. Moreover, because most demand-side technologies (perhaps with the exception of large industrial operations) are of a smaller order of magnitude – both financially and physically – than centralised generating technologies, it seems reasonable to conclude that policy interventions on the demand side have the added advantage of possibilities for iterative fine-tuning and flexibility. All else being equal, a good example of such ‘flexible’ solution would be to incrementally increase the costs of energy to enforce a reduction of energy consumption 505 . On the one hand, the process of incrementally and stepwise increasing energy costs (e.g. through taxes) allows learning processes (e.g. in terms of modifications in order to slow down or speed up the planned increase, to exempt certain categories of users, etc.); while on the other hand, the direction of the process makes clear that a long-term agenda and goal is being followed. Finally, from the perspective of adaptiveness (and here technology policy could take a cue from evolutionary theory), preserving some diversity in the options pursued can generally be seen as a good precautionary practice in view of the risks associated with possible ‘shocks’. Typical ‘shocks’ in the field of energy policy include price (e.g. sudden increases in oil prices), quantity shocks (e.g. sudden bottlenecks in production capacity) or technology shocks (e.g. the sudden discovery of a generic design fault in a type of reactor) (Helm 2002). Diversity (including a practical measurement) will be further discussed in chapter 6; for the time being, we simply appeal to the common wisdom of ‘not laying all eggs in one basket’: the more options available, the more information produced, or the more values to choose from – all of this adds to conditions favourable for collective debate and learning. 504

One may safely disregard the rather philosophical objection that, strictly speaking, no decision is reversible since historical development and natural evolution are ongoing and that therefore, the world can never be the same as it was before taking any decision. From a practical and legal point of view, it is sufficient to distinguish between consequences which are more or less reversible and those which would appear to be irreversible. For instance, pharmaceutical products can be recalled once they are released on the market, whereas an accidental release of radionuclides into the environment would be more irreversible. 505 Here, we do not pronounce any judgment on the possible side-effects (e.g. financial impacts) of such policy.


323

Of course, maintaining diversity, flexibility and reversibility of technological and policy options comes at a price. As a rule of thumb, one can assume that their importance increases the farther away the energy system is from a ‘sustainable’ state, and when there is no evidence of a significant change in the right direction. Of course, such judgments should in itself be subject of (potentially contrasting) evaluations, but arguably, most people would agree that energy provision in Belgium is far from a ‘sustainable optimum’. 3.3.3

Tactical considerations in political decision making

Of course, while it can be generally argued that there would be great merit in encouraging decision takers to consider transformable standards and strategic factors (and more generally, to make public the reasons for their decisions), decision takers have no legal duties to do so, except perhaps through their accountability to parliament 506 . OXERA (2000, p. 32) are of the opinion that this state of affairs is not unduly problematic: they maintain that, in addition to formal accountability, giving reasons is also generally advantageous for decision takers from a strategic point of view. This belief is grounded on the following perceived strategic advantages: 1. revealing the relation of scientific advice underlying the decision places the decision taker in a stronger position to defend the decision; 2. making uncertainties explicit can help to protect decision takers if the policy measures have to be changed subsequently (e.g. because new evidence has emerged); and 3. justifying a decision makes it more likely to be acceptable to a broader audience, especially when it is potentially contentious. However, we believe that this view might be somewhat overly optimistic, since it oversees the difficulty that requiring justifications from decision takers also might restrict their ‘freedom of movement’ (and hence, a more ‘strategic’ form of giving justifications might be the rule rather than the exception – cf. Chapter 4). At the other extreme, von Schomberg (2002, p. 232) – setting out the general requirements for the institutionalisation of discursive procedures – argues in favour of a number of general norms which, according to him, would have to be given the authority of administrative law: 1. an information law entailing the right to obtain, and the obligation to provide, information; 2. a selective suspension of the majority principle in the case of ‘irreversible’ decisions (e.g. approving only those risks which have a particular, limited time frame); and

506 OXERA (2000, p. 32) give an example of what such structured explanation for a decision might look like. This includes: a) an indication of the importance of the decision; b) a summary of the available options; c) a summary of the advice received on the subject; d) a summary of risks and benefits of each option; e) an overview of relevant policy fields and other commitments impinging on the decision; f) the decision criteria which were taken into account; and g) reasons for possibly rejecting certain types of advice (or making a choice between conflicting advice).

324

Chapter 5

3. the elimination of unfavourable possibilities for actions which could give rise to irreversible damage or imply irreversible decisions. While we agree with the first of these norms, we believe the second and third requirements sound particularly unrealistic, simply because the degree of ‘reversibility’ of political or Deriving an technological measures might be a contentious issue in itself 507 . uncontroversial measure of reversibility as a necessary basis for such rigid form of regulation thus seems to become a quite daunting task. Furthermore, we see no reason why ‘reversibility’ should be given such an absolute status as a criterion for decision taking, nor why an ‘irreversible’ decision cannot (and should not) be justified by potential advantages (no matter how substantial). In short, we believe the only realistic solution lies in having a judicial procedure at hand which obliges decision takers to produce an explicit justification for their decision which takes into account notions of ‘reversibility’ (and/or flexibility, diversity, etc.) without however prejudging the decision itself. Again, instruments of international law – and in particular, the Aarhus convention – might be harnessed for this purpose 508 .

3.4

Initiation of policy instruments and organisation

As expounded in the previous sections, for a purposive sustainability policy it is decisive that the inadequacy or general incompleteness of knowledge does not paralyse or hinder action, but rather that, in the interpretation and implementation of practical measures, a maximum range of opportunities for learning from ‘collective experiments’ is seized.

507

Keulartz et al. (2004, pp. 11-12) give a particularly enlightening account of how ‘irreversibility arguments’ are actually used by opponents of new technological developments (in this case, biotechnology and cloning) in order to stop any form of further debate. These authors show how irreversibility is often deployed in the form of a ‘slippery slope’ story-line, according to which taking a particular decision will inevitably and irreversibly lead to a whole series of steps, which in turn will lead into ‘moral abyss’. In the nuclear power debate, a direct link is sometimes made between the decision to build nuclear power plants and a general culture of wasteful electricity consumption behaviour (supported by a lack of political attention for promoting more responsible consumption) (see e.g. Glorieux 2002). The argument, though perhaps correct from a historical perspective, is of course faulty from a purely logical point of view. 508 The ‘Aarhus convention’ legally enshrines the minimal requirements for public access to information, public participation in decision making and access to justice in environmental matters at an international level. Although at present the Aarhus convention does not include the (demanding) requirements we would like to see included in a model process, it should be mentioned that it is constantly evolving (as a result of additional protocols, international jurisprudence, etc.) and can thus be expected to become ever more ‘stringent’. For example, the (mandatory) drafting of a ‘regulatory impact assessment’ for new proposed regulations (as an active measure for the promotion of public access to information and possibly also public participation) could offer an adequate opportunity for justifying government decisions. In the Flemish region, regulatory impact assessments have to be included since the 1st of January, 2005 as an annex to the policy note accompanying proposed regulations presented to the Flemish government. The Flemish regulatory impact assessment covers the following topics: 1) a description of the proposed regulation; 2) a description of the problematic situation addressed by the proposed regulation, including the goals of the regulation; 3) an identification of the most relevant options for realising this goal; 4) an analysis of the expected advantages and disadvantages of each option; 5) a record of the consultations which have taken place in order to prepare the proposed regulation; 6) a discussion of the recommended implementation strategy of the chosen option, including assessment and revision procedures; 7) a summary of the arguments leading to the adoption of that particular regulatory measure; 8) contact information for further questions about the impact assessment or proposed regulation.


325

Organising these learning processes consists (to remain within the metaphor of an experiment) in firstly ensuring as good a preparation of the experiment as possible (i.e. by carefully structuring the problem, modelling different scenarios, and choosing appropriate policy standards and measures – cf. supra), in supervising the careful execution of the experiment (i.e. monitoring), and finally in observing the results of the process, comparing them with goals pursued and, if necessary, investigating reasons for deviations (i.e. evaluation). The practical application of certain policy measures over a sufficient time span therefore offers an essential opportunity for gaining the knowledge which was not available in the preparatory phases of planning; therefore, monitoring and evaluation become crucial aspects of ‘collective learning’. 3.4.1

Monitoring

Our reasoning so far has been mainly concerned with governance and decision-making processes in the ‘political system’, with conversations taking place mainly between (officially recognised) stakeholders, experts in various disciplines and politicians. However, as both Beck (through his concept of ‘sub-politics’) and Latour (through his assertion that there can be no definite limit to the number of collectivities participating in the construction of the ‘good common world’) constantly remind us, there is a continuum of forces which exert political influence. One can think here of a number of ‘public’ networks – both ‘official’ and ‘unofficial’ (various government agencies, pressure groups, the general expression of ‘public opinion’ in the media, etc.), research institutes, funding institutions, the actions of energy consumers, and of course, business firms. The problem we have to address then is that the influence exerted by these other ‘collectivities’ in shaping the energy future might not go in the direction desired by ‘official’ governance structure. The challenge is then to have some form of ‘intentional’ shaping in place, providing guarantees that the (transformable) standards will be met. Now of course, such intentional shaping of the ‘good common world’ (and the monitoring of its borders) is influenced by the particular perspective or theory one adheres to for conceptualising the interactions occurring between different collectivities. In what follows, we will focus on the possibilities for ‘steering’ business firms and consumer preferences, which arguably represent the most important categories of actors for implementing desirable sustainable energy solutions in practice. Various authors (e.g. Coombs 1995; Rip 1999; Grunwald 2000) have pointed out the limitations of the traditional ‘interventionalist’ policy stances towards steering technological development, which tended to focus on the two main arenas of ‘promotion’ (either direct – i.e. government-dominated technology – or indirect – i.e. through research promotion, technology-support programmes) and ‘control’ (e.g. through regulations, safety

326

Chapter 5

standards, technology-relevant taxes, etc.) 509 . Put very briefly, these authors argue that these two arenas have been dominated in the past by different policy discourses and corresponding monitoring or research efforts, leading to fragmented (at best) or sometimes even conflicting (at worst) efforts. They describe the promotion arena as being concerned mainly by the quantitative aspects of technology development and its effects on economic growth (dealing essentially with externalities, distributional issues, and various perceived instances of market failure – e.g. in the area of basic research), with a predominant academic input by economic sciences. The control arena on the other hand is seen by these authors as being centred on pragmatic responses to particular kinds of problems which acquire a high profile at certain times (e.g. environmental pollution, workers’ safety, etc.), and much less influenced by the input of academic disciplines (with the exception perhaps of risk analysis). Furthermore, Coombs (1995) maintains that, despite theoretical advances in understanding technological innovation and diffusion of new technologies, these new intellectual frameworks have made little impact on actual policy practices 510 , 511 . In particular, Coombs (1995) argues for a radical re-thinking of the control arena, where actor-network theory has made us aware of the fact that ‘control standards’ are not functioning as unambiguous ‘constraints’ (separating ‘good’ and ‘bad’ ways of applying technology), but rather are made present (i.e. are articulated, interpreted, etc.) in concrete situations. For our purposes, the important point is not only that there is a parallel between the actor-network approach to analysing technological innovation (connected to general business strategies) and governance structures (as we have been using the theory), but also

509

Rip (1999) reminds us here of the so-called ‘control and entrenchment dilemma’ (originally coined by Collingridge (1980)). The dilemma arises when the introduction of new technologies in society also produces unforeseen effects (sometimes desirable and sometimes less so). Because most of these effects become manifest during and after introduction of a new technology into more general use, the possibilities for correcting them (if necessary) by adjusting the technology are limited. By the time the negative social impacts are recognised, technology is already firmly embedded in sectors, institutions, and practices. Thus, a dilemma between control and entrenchment. 510 This observation (written down in 1995) should in our view be qualified somewhat in view of present practice. Indeed, as von Schomberg (2002, p. 231) observes, administrations have become increasingly involved in bargaining and negotiation processes with stakeholders. However, many objections remain, as such involvement usually arises in the context of strategic action and under conditions of unequal bargaining power. Furthermore, by centring the mediation around administrative tasks the (possible) tension between ‘control’ and ‘promotion’ is not addressed. 511 Coombs is referring here to actor-network theory and evolutionary economics, which, in his view, show many similarities. Both in the evolutionary and constructive perspective, many actors are involved in making choices. For example, the choices made during the design process are not only influenced by the technical know-how of the designers but also by the chances of success of technologies to be developed as estimated by the planners, the managers, the sales departments, the investors etc. The properties of technology, its future impacts (including the distribution of risks and benefits) are shaped in interactions between the social actors. The shaping of technology, of its properties and impacts extends beyond the development stage into implementation, adoption and wider use. The composition of interacting social actors changes from one phase of the process to the other but each significant interaction leaves its imprints on technology and on the social environment. Technology and the social conditions co-evolve in the same movement, and assessments of various kinds occur all the time. 511 This observation should in our view be qualified somewhat. Indeed, as von Schomberg (2002, p. 231) observes, administrations have become increasingly involved in bargaining and negotiation processes with stakeholders. However, many objections remain, as such involvement usually arises in the context of strategic action and under conditions of unequal bargaining power. Furthermore, by centring the mediation around administrative tasks the (possible) tension between ‘control’ and ‘promotion’ is not addressed.


327

that business strategies in principle represent a site for the intervention of other ‘actants’, whose motivations and values (provided they are not fundamentally opposed to that of the business firm) can introduce alternate concerns. The challenge is to participate in those networks within the institution of the firm, but from a standpoint beyond that of the firm (which will subsequently ‘translate’ the various ‘constraints’ into business opportunities). Of course, what we are proposing here is not the planning of technological products or technology development as a whole (in the sense of a ‘planned economy’). We cannot (and should not) expect that the ‘collectivity’ of a business firm as a whole responds to the requirements and quality control criteria we are setting out for governance structures. What we are proposing is planning in the general sense of ‘an intended exertion of influence on future potentials for action’ (Grunwald 2000, p. 126). Here, Coombs’s (1995, p. 338) distinction between different levels where strategic discussions in a large firm take place can be useful: the ‘corporate level’ (concerned with survival, overall profitability, risk, finance, etc.), the ‘technology level’ (concerned with core technological assets which contribute to the overall corporate rationale), the ‘business level’ (concerned with the competitive strategy to be followed in a particular market) and the ‘product level’ (concerned with concrete product offerings in a particular market, the performance levels and characteristics of these products, the details of the articulation of the product with the changing patterns of demand, etc.). Following Coombs, we suggest that it would be most productive to try to connect the ‘technology’ and the ‘product’ level with our governance network 512 . Connecting with the strategic ‘technology’ level can be achieved by giving private firms with an interest in the general area of energy policy (e.g. electricity producers and suppliers, large industrial energy consumers, technology developers, etc. – cf. section 5.3) a seat in the governance structure we are elaborating 513 . Participation of private enterprise in this institution would guarantee that the business logic would be connected to the ongoing deliberations with other stakeholders, underpinned by independent research, on what energy futures would be generally desirable from the perspective of sustainable development. Policy decisions informed by these deliberations would then serve the purpose of ‘prestructuring’ the future by promoting or selecting individual or (preferably) bundles of options out of the existing diversity of possibilities, not from the point of view of business opportunities alone, but also incorporating concerns of relevant government agencies, consumer desires (expressed through their federations) and ‘the common good’. Moreover, given that our proposed model process explicitly internalises uncertainties in the knowledge base and ‘sub-optimality’ in the nature of energy policy (e.g. by selecting multiple visions, explicitly striving for flexibility and diversity, etc.), actions based on data

512

Coombs’s argument is based on the observation that engagement with the other levels is not feasible (it will be strongly resisted by companies because it touches upon matters that are sensitive and very close to competitive interests) nor desirable (interfering with the business level might actually reduce the range of technical possibilities and solutions offered by firms, thus reducing the general diversity of possible options). 513 Of course, much of the energy-relevant technology is not developed in Belgium. Nevertheless, these technological developments can be followed up in the deliberations.

328

Chapter 5

highly susceptible to revision should have been implemented with great restraint. Hence, the need for major revisions of the policy measures should be limited to those at the fringe, and not at the centre, of the evolving ‘sustainable energy commonwealth’. Deliberations at this ‘top’ level can furthermore be supplemented by deliberative forums at ‘intermediate’ levels – e.g. centred around technological developments considered to be crucial for sustainable development (e.g. organisations for the promotion of cogeneration or renewable energy technologies) – some of which might already exist (cf. Section 6). Connecting the wider concerns of sustainable energy governance with the ‘product’ level might be achieved through ‘strategic niche management’. The concept itself was introduced to highlight an important aspect of successful introduction of new technologies (Rip et al. 1995; Rip 1999). Strategic niche management is the orchestration of the development and introduction of new technologies through a series of experimental settings or (technological) niches, which are temporary screened off from the working of selection mechanisms. Such settings are formed around promising technological developments of which it is not known whether or not they will fulfil the expectations. Niches can be centred around particular technologies, but also on small-scale experiments with new lifestyles (e.g. car sharing), or experiments with new institutional arrangements (e.g. a limited emission trading regime involving only specific industrial sectors in selected countries). In these experimental settings different categories of actors are brought together: the designers of technology, the envisaged future users, sometimes governmental agencies and if possible other agents who engage in facilitating and modulating the interactions between the first two categories of actors. In the interactions, various actors learn about the technical side of the design, about the needs and requirements of the users, about the cultural and political acceptability of technology in development. The idea is that as designers of technology become aware of broader societal issues they are able to incorporate new aspects into the actual process of design and development. Through the use of strategic niche management, the contours of the institutional and technical context necessary for the new technology to function should become visible and first steps in the process of embedding technology in society can be made. More specifically for the context of energy policy, we argue that the different forms of user representation generated during innovative processes involving energy consuming equipment are of critical importance. In this case, strategic niche management should attempt to discover how it is possible, in the process of developing and marketing a technological system, to reconcile what may well turn out to be barely compatible or even conflicting user representations (Akrich 1995). For example, many opinion polls show that a majority is in favour of reducing energy consumption, which seems to be a ‘good idea’ from a number of possible points of view (on the grounds of economic considerations, environmental protection, geo-political considerations, etc.). However, it is unlikely that this ‘diffuse’ support for reducing the country’s energy demands will materialise in the form of actual energy-saving behaviour merely on the grounds of an appeal to people’s sense of responsibility. It seems that what is needed are systems which are able to superimpose user representations of personal comfort with careful consumption of energy (Vandenabeele 2005).


3.4.2

329

Evaluation

At regular intervals, results of ongoing monitoring activities should also be fed back into the deliberations going on at the governance level. The significance of this element of the model process cannot be overstated: contrary to contemporary and historical developments in energy policy, the success of the entire process is related to its explicit embrace of uncertainties and learning dynamics, with continual feedback needed to create a flexible and adaptable structure. This can for instance be achieved by the formalisation of a review procedure. In practice, we see three distinct roles for the review procedure: firstly, to perform an ‘implementation review’; secondly, to conduct an ‘evaluation of changes in the knowledge base’; and thirdly an ‘evaluation of the model process’ itself. The first of these roles, the ‘implementation review’, is the most straightforward one. This review should analyse the success or failure of the chosen framework of (transformable) standards and the derived policy measures. In case of success in meeting the standards, it can be considered to tighten them; conversely, in case of failure, the review procedure should analyse whether the standards remain feasible in light of the feedback received (e.g. in the case of unexpected developments or ‘unrealistic’ expectations), or rather, whether the failure is the result of non-compliance with agreedupon commitments by certain actors (including government). In the latter case, the review should be able to issue recommendations on how this situation could be rectified, or possibly even enforced. The ‘evaluation of changes in the knowledge base’ (e.g. the database of sustainable energy indicators, the methodologies for deriving them, the models used, etc.) is concerned with feedback on what can be learnt from the various collective ‘experiments’ (e.g. in strategic niche management, additional research commissioned, etc.) 514 . The policy initiatives that emerged from the previous iteration will inevitably have been dependent on the accuracy of the data gathered during the previous problem structuring phase, which, in view of the uncertainties involved and/or limitations on the actual breadth of the analysis (in terms of resources) that could actually be internalised, of necessity will remain open for revision. Moreover, as neither industry nor society as a whole remains static, it will be periodically necessary to review those elements of the knowledge base particularly susceptible to those changes. Furthermore, this part of the review should also check whether, in light of the changes to the knowledge base, the scale of these changes requires the re-evaluation of policy measures. In view of this ‘political’ importance of the review procedure, it might be useful to consider forms of ‘extended’ peer review (implying an engagement with a wider community of stakeholders) as advocated by Funtowicz and Ravetz (1993) in their plea for ‘post-normal science’ 515 .

514

Such periodic review is for instance already implemented in the ExternE-project (cf. Chapter 3). OXERA (2000, p. 33) list some disadvantages of ‘expert’ peer review: 1) there are numerous cases where falsification of evidence, errors in analysing data, poor experimental technique, etc. was not detected by expert peer review; 2) expert peer review has a tendency to reinforce the existing scientific consensus, and is therefore insensitive to ‘surprises’; 3) time devoted to peer review is usually short. 515

330

Chapter 5

The final aspect of the review concerns the deliberation process itself. It should deal with questions such as how this process is perceived by the participants, whether the ongoing deliberations are considered to uphold criteria of fairness and competence, whether new forms of participation should be considered or whether new ‘actants’ should be allowed entry into the debate, etc. In short, this final aspect is equivalent to a periodic quality assurance with regard to the process parameters.

3.5

Detection of matters of concern

Of course, not everything can be captured in intentional experiments and political planning: ‘surprises’ – e.g. accidents, unexpected technological innovations, legal proceedings, etc. – cannot be excluded. Some philosophers even consider such ‘surprising events’ (in the sense of an experience that cannot be captured within the existing moral vocabulary) vital for real ethical commitments (Badiou 2005). Furthermore, policy issues may emerge from many sources (and not just from carefully planned ‘collective experiments’), including parliamentary questions, the media, scientific debates (fought out in scientific journals or colloquia), and general public opinion. Thus, although we recognise the inevitable occurrence and even the value of ‘surprises’, we nevertheless believe that early detection is desirable, because it maximises the number of policy options available and the time available to analyse them. Therefore, systematic procedures should be employed to maximise the chances of detecting issues promptly. Achieving this would be an entire subject in its own right (see in particular the “Late Lessons from Early Warnings” report (Harremoës et al. 2002)); nevertheless we present some general ideas taken from that report: • It could be useful to organise an ‘early warning’ function in different public services, for instance with regard to policies that are being developed at the international or EU level. This would make it easier to evaluate the effect of the proposed policy measures at the Belgian level, to prepare an official position on these measures, and/or to work proactively for their realisation; • Another idea is to include a duty on all standing advisory committees and/or research councils (e.g. the public health council) to produce regular briefings on key issues, which review recent developments within their competence relevant to energy use and its impacts; • Such review should in particular identify certain types of impacts, which – even if no conclusive proof of their harm is available – nevertheless point at the possibility of such negative impacts emerging in the future (and thus might serve as ‘warning signs’). In particular impacts which show a high degree of irreversibility (e.g. persistent and/or bioaccumulable pollutants, in particular if they are highly dispersed) should be targeted; • Having regulatory bodies and public services in place which are independent of special political or economic interests; • Emphasising the value of independent academic research;


331

• Reducing barriers to interdisciplinary research on adverse impacts (e.g. impacts on human health vs. ecological impacts, chronic vs. acute impacts, epidemiology vs. molecular biology, etc.); • Dedicated research into areas of uncertainty which might reasonably be expected to be clarified further with a reasonable research effort; • Dedicated, long-term monitoring of supposed adverse impacts, including an appropriate dissemination and utilisation of the monitoring results; • Finally, the set of sustainable energy indicators might itself incorporate some indicators for detecting ‘surprises’, e.g. an unexpected increase/decrease of energy consumption might set off a search for clarifying factors. An attentive reader might at this point however object that we have not addressed the role of arguably the most important ‘agenda-setter’, i.e. the media. It is evident that at least some communication to the general public of the complexities involved in energy policy planning might greatly aid our model process to move forward. Here, we must humbly admit that we have no easy answers, especially since we tend to share Harremoës et al. (2002, p. 189) rather pessimistic assessment that …it is therefore ironic that the media and other professional communicators have been moving in the opposite direction, emphasising only ‘positive knowledge’ and the ‘clear and simple message’ in a soundbite culture. This tends to exclude the communication of ignorance, of complexity and of responsibility in face of the essential limits of all knowledge…

However, one might also see this predicament as a challenge for future research in communication…

4 Quality assurance Quality assurance assumes a central role if one accepts the non-representational epistemological claims of constructivism. More precisely, this epistemology does not deny the importance of representations but rather emphasises the practices that give rise to them (the work done in order to arrive at the undisputed ‘facts’) 516 . Furthermore, the fundamentally collective (combined social and material) character of these practices, the primary criteria of interest become those attaching to the processes, procedures and contexts in which knowledge is produced and deployed, rather than criteria pertaining to the sources and content of knowledge. Goorden (2003) provides us with yet another reason for emphasising qualitative process-related criteria. Goorden notices a certain ‘ambivalence’ in the minds of policy makers or political planners with regard to participatory processes. On the one hand, according to the ‘primacy of the political’, politicians carry the ultimate responsibility for the plans they approve of, and hence likely have strong preferences for certain outcomes. On the other hand, stakeholders and ‘emancipated’ citizens (especially in the case of controversies regarding local projects) 516

Latour in his works often likes to remind us of the French play of words “…Les ‘faits’ sont faits…”.

332

Chapter 5

often demand involvement in an early stage of the planning process, in order to maximise their chances of influencing the outcome. According to Goorden, this paradoxical situation often leads policy makers to the choice of a ‘safe’ format of participation – i.e. one that is expected to deliver a reasonably predictable outcome. Quality assurance then becomes a vital precondition for avoiding this tendency and for assuring participation on an equal footing. The following sections will therefore be devoted to a discussion of the most pertinent quality criteria to be upheld by our model process. Our discussion will however remain on a fairly general level, since much of these criteria have been commented upon in greater detail elsewhere. Furthermore, a high level of abstraction seems to be appropriate for the time being since the suggested criteria will themselves become the subject of negotiations (informed by learning experiences) between involved actors once the model process gets under way. Articulation of quality criteria for deliberation will also depend on the context of application or the specific subject under discussion, and can of course benefit from the ‘lesson learnt’ in other participative settings over time (cf. our discussion of the NCE-settings in chapter 2 – section 3.3.2). In the following sections, we mainly refer to Webler’s (1994) insightful application of the Habermasian framework for a ‘free and unconstrained dialogue’ to environmental discourse, which includes a practical ‘quality checklist’ for evaluating participatory processes in terms of fairness and competence that could serve as a starting point for a discussion 517 .

4.1

Fairness

In general, fairness refers to the distribution among the ‘actants’ involved of opportunities to act meaningfully. ‘Acting meaningfully’ incorporates four fundamental expectations 518 : a) being a participant in the discourse; b) initiating speech acts; c) challenging and defending claims; d) influencing collective outcomes of the discussions (i.e. achieving ‘closure’). These expectations can be related to our discussion on ‘problem structuring’ (Section 3.1.2): fairness requires that a) all relevant actants are identified; b) the knowledge regarding these actants is gathered; c) opportunities are offered for identifying limitations to the existing knowledge (e.g. gaps, errors, inherent uncertainties, etc.); and d) a collective (albeit provisional) agreement is reached that the ‘best possible jury’ has been established to speak for the actants. In particular, promoting fairness means that the model process must provide ample opportunities (also in terms of resources – e.g. time and money) for these expectations to be met, including consensually-approved means to resolve conflicts in case closure cannot be reached on certain issues 519 .

517

Which, in the sense developed by Webler, encompass our notions of ‘transparency’ and ‘uncertainty management’. 518 Here, we translate Webler’s (1994) quality criteria into Latourian parlance. 519 Without such rules for enabling closure (e.g. on the type of indicator, the characterisation of the ‘state of knowledge’ with regard to the indicator, etc.), the model process would be open to the dangers of ‘filibustering’ – e.g. participants in the deliberation could conceivably demand that deeper and deeper justifications are given for a particular validity claim.


4.2

333

Competence

Criteria for establishing competence are intimately related to fairness criteria. Competence generally refers to the construction of the best possible understandings and agreements given what is reasonably knowable. Expectations of competence include: a) access to ‘relevant’ knowledge in light of the participating actants; and b) use of the best available procedures for knowledge selection (e.g. peer review). Competence thus depends on individual qualities (e.g. qualifications of individual experts), group characteristics (e.g. composition of a scientific advisory group) and group dynamics (e.g. rules governing the discussions). In particular (as mentioned in our introduction to this chapter), because politics abounds with dangers to deploy uncertainties strategically, competence will also crucially depend on having a fair representation of the limitations in the scope of representational knowledge – hence the importance of having procedural guidelines in place for uncertainty management.

4.3

Uncertainty management

As our model process will be applied in the context of scientific controversy and the acquisition of new knowledge, a clarification will be needed on what is precisely understood by ‘uncertainty’ and what type of uncertainties are particularly relevant for influencing the results of deliberations. We refer here to the ‘typology of uncertainty’ we have developed in chapter 2 for reflecting the state of affairs in science. Having such a typology of uncertainty at hand becomes all the more relevant since the precautionary principle establishes a normative link between the nature of uncertainties and possible adverse effects and subsequent policy actions (cf. Section 3.1.1). We agree with von Schomberg (2006) that uncertainty management should be based initially on a qualitative assessment of the available knowledge – e.g. the knowledge concerning established causeeffect relationships and the degree of necessity to know those relationships in order to make a judgment (i.e. the sensitivity of the policy-relevant outcome to this relationship) – rather than a quantitative one (relating to an ‘amount’ or ‘degree’ of uncertainty), quite simply because such information might not be available under conditions of uncertainty. Rather, uncertainty assessment and management should start from a careful review of available knowledge (e.g. documenting ranges and distributions in literature, including information on the key causes of uncertainty; assessing the ‘strength’ of the evidence on which such distributions are based, etc.) before deciding whether quantification of uncertainties is appropriate for that particular case. This is also the approach we have followed in chapter 3, where we assessed the general ‘robustness’ of external cost calculations as an example of a highly aggregated indicator of sustainable development. However, our reflections on external cost calculations also make clear that such uncertainty assessment, if performed for every single indicator, can quickly become quite elaborate and burdensome. A conflict with the transparency requirement (cf. infra) might surface in many cases. Solutions to this dilemma could for instance include the scoring system with letter codes developed by the ExternE-network (cf. Chapter 3). However, this

334

Chapter 5

solution has the disadvantage of forcing uncertainties into a quantitative frame by characterising them according to a degree of spread around a central value. Another possible candidate for reconciling requirements of ‘transparency’ and ‘uncertainty management’ could be the NUSAP method developed by Funtowicz and Ravetz (1990) 520 . An interesting aspect is that with this method it is possible to address in a stylised way uncertainties at the general level of a scientific study, while not going into the exhaustive details of all types of uncertainties involved (as we have done for external cost calculations in chapter 3). The ideal way forward might very well be a combination of an uncertainty assessment at a fairly aggregate level (e.g. using the NUSAP method), combined with a detailed ‘traceable account’ of how the assessment was produced (e.g. including reasons for adopting a particular probability distribution, lines of evidence used, standards of evidence applied, identification of critical uncertainties, etc.).

4.4

Transparency

In a way, it can be said that much of the content that belongs under the heading of ‘transparency’ has already been covered in the above sections 521 . Since the ultimate goal of a transparent communication – in the enlarged sense of the word used by Latour (2004a) – refers to ‘speaking for an actant as it would speak for itself if it were able to speak without speech impedimenta’, this notion encapsulates the other notions of uncertainty (i.e. if the actant speaks in a hesitant voice, the dialogue should reflect this), competence and fairness. However, in our case of governance for sustainable energy, we believe the importance of transparency should be heightened by the fact that for our model process to be perceived as legitimate, transparency criteria will need to be determined not only for and among the participants to the ongoing deliberations, but also, and crucially, for the ‘wider public’ as well (and should therefore be supported by adequate legislative provisions). Transparency therefore also refers to the effectiveness with which a certain assessment is represented and communicated to the ‘outside world’ (Stirling 1999) – e.g. by ensuring that all participants in the deliberations understand why a certain indicator is measured in that specific way, or by including provisions for decision takers to justify their final decisions with regard to the advice issued by the deliberations. All else being equal, the greater the number of ‘variables’ or ‘parameters’ which remain unexplored or undeclared in appraisal, the less

520

NUSAP stands for ‘Numeral’, ‘Unit’, ‘Spread’, ‘Assessment’ and ‘Pedigree’. The idea is to characterise each part of an analysis in these terms. Numeral, unit and spread are rather familiar terms and enable to characterise an estimate in quantitative terms (e.g. for external costs, the numeral is the ‘expected value’, the unit ‘Euros’, the spread is characterised by a letter code establishing an order of magnitude). Assessment and pedigree represent levels of uncertainty that go beyond technical uncertainties: the pedigree establishes a systematic multi-criteria evaluation of different stages of production of the knowledge base (e.g. quality of model/theory – ranging from ‘established theory’ to ‘definitions’, quality of data – ranging from ‘experimental data’ to ‘uneducated guesses’, and degree of peer acceptance – ranging from ‘total’ to ‘none’). The assessment grade is based on the average pedigree score. Recently, the ‘pedigree’ score has been extended further in order to reflect societal issues (Craye et al. 2004). 521 On achieving transparency and openness in regulatory matters, see e.g. a recent advice of the Flemish environmental and nature council (MiNa-raad) on the implementation of the Aarhus convention in the Flemish region (‘Advies van 24 februari 2005 over de implementatie in Vlaanderen van het Verdrag van Aarhus’ – available at ).


335

transparent will be the outcome 522 . In this way, the inevitable complexity of the questions raised by sustainable energy governance and the requirement of transparency might be seen as being in tension. No doubt, the requirements of ‘competence’ (i.e. having the ‘best science’ available) and ‘transparency’ alike demand the striking of some judicious balance. Achieving transparency might be particularly challenging in situations involving large amounts of data or difficult scientific reasoning. Of course, there are no general rules, but there may be merit in involving non-experts (e.g. scientific journalists) or non-specialist scientists working together with the scientific specialists in order to test the accessibility of the advice and to help with its presentation. Moreover, there is a difference between publishing a considered draft and exposing all the ongoing ‘work-in-progress’ to public scrutiny. In the latter case, public concerns could be unnecessarily fuelled by early hypotheses which are rejected later on in the process. Furthermore, too much public exposure at an early stage may cause scientific advisers to be excessively cautious, and thus compromise the quality of the advice (OXERA 2000).

5 Roles of different actors An important contributing factor to the overall quality of the model process will of course be the involvement of the ‘right’ kind of actors in the deliberations, and the division of labour this implies 523 . Now, we realise that there are inherent dangers to specifying the contributions of different actors, in the sense that, here again, a possible conflict between fairness and competence criteria seems to loom at the horizon. Stressing fairness criteria implies that, in principle, a mutuality of participation in the occupation of roles should be upheld to a maximum degree possible. On the other hand, practical competence concerns of solving a problematic situation within given limits of time and resources (and without re-inventing the wheel every time) imply that certain role constraints should be adopted. Here again, the answer lies in finding a judicious balance 524 . We firmly disagree with OXERA (2000) who argue that the legitimacy of the process depends entirely on a strict separation of functions, with each ‘profession’ taking on different functions 525 . This solution is yet another variant of the old maxim that ‘scientists provide the facts, stakeholders the interests, and decision takers (aided by policy makers) the closure’, which we have criticised over and over again in the preceding chapters of this dissertation. We believe Latour’s (2004a) solution of defining a set of (idealised) roles which jointly contribute to the different tasks (‘perplexity’, ‘problem structuring’, etc.) according to their

522

Hence, ‘transparency’ as a guiding principle should be clearly distinguished from ‘simplicity’. An ‘actor’ (i.e. individuals, organisations, etc.) should be seen as one (albeit important) node in a more general network of ‘actants’. 524 E.g. some stakeholders might not trust certain experts or members of other interest groups, but this does not mean they want to rediscover scientific methods all by themselves – they just want to have sufficient insight in the methodology to determine whether it is intentionally biased or not. Or they might not trust politicians or agree with political decisions, without however wanting to overthrow all democratic procedures – a (legal) guarantee that the decision taker justifies his/her decision might suffice in this case. 525 OXERA (2000) identifies four categories: ‘scientific experts’, ‘stakeholder representatives’, ‘policy makers’ and ‘decision takers’. 523

336

Chapter 5

own capacities shows much more merit 526 . This means for instance that a ‘scientific expert’ (by profession) might very well take on the role of a ‘politician’ (e.g. when devising scenarios). Roles can also be attributed to groups or entire organisations. Ideally, each task should be performed by a network in which each role is represented. However, in view of a transparent functioning, the protocol guiding the deliberations should clearly specify which role is supposed to be played by which actor (or group of actors).

5.1

Scientific experts 527

The competence of the model process, in all of its stages, will of course depend to a great extent on scientific expertise (especially for a complex subject matter such as energy policy). In the past, science has already provided ample examples of its capacity for inducing ‘perplexity’, e.g. in the case of the detection of anthropogenically enhanced global warming, or the essentially ‘serendipitous’ discovery of the hole in the ozone layer (Harremoës et al. 2002). ‘Perplexity’ can, by definition, not be planned, but having independent (i.e. free from special economic and/or political interests) and sufficiently funded research institutes in place seems to be a necessary (albeit not sufficient) condition for creating a favourable atmosphere 528 . Scientific expertise will of course also have a major part to play in setting up and structuring the set of criteria and indicators for guiding sustainable energy policy. Constructing scientifically valid indicators of sustainability can mean hiring consultants for putting into place the necessary research protocols and searching relevant data (and explaining them), but it might also be as simple as copying a report on already existing indicators. Individual competence when speaking as a ‘reliable witness’ for a particular actant is certainly one element to be taken into account when selecting experts. However, individual competence should in our case also extend to an awareness of the wider state of knowledge in the research field in question and an aptitude for communication. Frequently, a group of experts will be responsible for developing

526

The concept of a ‘role’ in Latour (2004a) should be understood as the ability to represent a network of actants whose interpretation is broadly shared. For instance, a ‘scientific expert’ is associated with notions of ‘representing external reality’, ‘mastering formalised knowledge’, ‘objectivity’, etc. Each actor present in the deliberations will tend to have a somewhat different representation of the role of a ‘scientific expert’; however, all that matters is that there exist enough similarities in these different representations in order to serve as critical vantage points. A ‘scientific expert’ by profession (i.e. the physical person) can then be compared to this ‘idealised’ role – e.g. political commitments, ties to special interest groups, etc. can all be considered as indicators of a deviation from the expected performance. 527 We are not only referring here to experts in natural sciences; other disciplines such as economics, history, political sciences, etc. can reasonably be expected to provide a valuable input to the analysis of energy policy. In some cases, lay knowledge (e.g. concerning the actual use made of technical appliances, or on expectancies of comfort, etc.) can provide useful information. Of course such lay knowledge should be subject to the same intensity of critical scrutiny as expert knowledge, e.g. concerning the reproducibility or reliability of methods to elicit lay knowledge (e.g. focus groups). However, it might be expected that for our case (i.e. devising a model process to guide reflection on the level of national policy making), in view of better opportunities for communicating progress in international fora, more value will be placed in indicators derived through formal science-based approaches (Miller 2005). 528 E.g. it can be argued that the discussions taking place within the walls of the NCE were more open, equitable and pluralistic through the provision of energy policy information developed by independent research groups (from universities and the federal planning bureau) as a counterweight to the expertise of the electricity monopolist and the associated engineering bureau (cf. Chapter 2 – Section 3.3.2).


337

indicators (and in the case of external costs, even an entire international research network!). In that case, group competence will also depend on having a balanced representation of the range of scientific opinion on the particular topic. Some scientists who have the potential to contribute useful information (owing to their knowledge or general experience) may even be employed by industry or pressure groups; such scientists should not be excluded a priori. Rather, their contribution should be managed and balanced by other sources of expertise 529 . Scientific experts also dispose of specific methods for achieving ‘closure’ (or an intersubjective consensus on the ‘facts’) in their domain, building on logic, empirical data, established experimental protocols, peer review, etc. – thus potentially providing a solid basis for the final decisions. This capacity aids in finding a scientific consensus on the best conceivable method for designing particular indicators of sustainable development. However, where there exists significant disagreement among scientists (e.g. different ‘paradigms’), such disagreement should not be obscured by striving for consensus at all cost – instead, the differences of opinion should be openly discussed. The task of drawing up ‘scenarisations of the collective’ can be entrusted (of course, in close collaboration with the other roles) to ‘scientific generalists’ in the field of energy policy. A ‘scientific generalist’ is an esteemed scientist, who not only transcends disciplinary boundaries and synthesizes knowledge from several fields but also understands the specific sensitivities of science for policy and the policy issues at stake (OXERA 2000, p. 26). Lastly, when a certain degree of tractability has been reached through the previous steps (e.g. the policy options are known and require a balanced decision, weighting up the costs, risks and benefits of each option), scientific experts can also contribute to assessing the impacts of particular proposed policy measures.

5.2

Ethicists

Ethicists’ role and involvement in energy policy has traditionally been very limited – to our knowledge, no official advisory mechanism on energy policy has ever made room for contributions by ethicists 530 . Nevertheless, ethicists play a vital role in our model process 531 . The common (academic) distinctions between procedural, substantial or consequential ethics become less important if we consider them from the point of view of collective experimentation, as we propose to do. Then it becomes apparent that these

529

E.g. through a mandatory declaration of interests by scientific experts. Other provisions of our model process guard against intentional bias – e.g. the overriding duty to the ‘good common world’ and the explicit consideration of uncertainties. In assessing the likely bias of a scientific expert, the whole of his/her professional profile should be considered (employment, affiliations, consultancy, employment history and experience, etc.) and not just the financial interest in the subject. Scientific adherence to particular schools of thought might also introduce bias. 530 Again, we stress that we are not referring necessarily to ‘academic’ ethicists – although an ‘academic’ ethicist would of course be a likely candidate for this role (perhaps ‘applied philosophy’, in the sense of Schrader-Frechette (1980) comes closest to our idea). In chapter 1, we have argued that environmental NGO’s have to some extent played the role of ‘ethicists’ in conflicts by stressing the ‘rights’ of the unrepresented entities (e.g. nature, future generations) or pointing out institutional power imbalances. 531 Schrader-Frechette has been one of the most vehement defenders of the role of applied philosophy in ‘science for policy’ (see e.g. Schrader-Frechette 1980).

338

Chapter 5

schools do not so much oppose each other, but rather designate successive stages of the model process, while making an effort to characterise the particular virtue of the stage they address. The ethicist’s task in inducing ‘perplexity’ and structuring problems correspond largely to the tasks set out by Habermas’s discourse ethic (at least, in the enlarged constructivist sense): helping to develop procedures and institutions that guarantee equal access to collective deliberations and fair and competent representation of all ‘actants’ to ensure that the decisions made are based on the ‘force of the better argument’ (rather than on the force of money, power, persuasion, etc.). Here, ethicists can be aided by methods such as risk analysis, environmental impact assessment, stakeholder analysis, and institutional analysis. In the ‘definition, comparison and selection of policy options’-phase, the main activities include providing, clarifying and explicating various arguments for or against particular courses of action (e.g. explaining the ethical foundations of the principle of ‘optimality’). In particular, ethicists should check whether moral principles, when they are invoked to defend particular courses of action, are applied coherently and consistently to all similar cases. Scenario-building on the other hand can be aided by the ethicist’s capacity for checking whether our common moral vocabularies, including the current institutional arrangements, are still suitable or have to be revised in the light of the possible future scenarios (aided by Boltanski and Thévenot’s commonwealth model). Contra Keulartz et al. (2004), we argue that it is not the ethicist’s task to aid political decision making through conflict management by clever ‘moral engineering’ 532 . Rather, ethicists should stay faithful to the purity of principle (i.e. attempting the impossible task of subsuming the entire situation within one commonwealth), and where political compromises are struck, the ethicist’s task should precisely be to point out that the solution is, in fact, a compromise. This is essential, since without this contribution, the model process would loose the capacity of keeping track of excluded actants – hence, ‘invisible enemies’ are created rather than ‘adversaries’ with whom we might resume a dialogue at a later time. In the words of Latour (2004a), ethicists should constantly remind us that the construction work is never finished: everything will have to be done all over again. We realise that building in this ‘obstructing’ role in an official governance mechanism might represent a significant challenge, in view of the general expediency for politicians to acquire an ‘ethical label’ by passing over difficult decisions to ‘ethical committees’ 533 …

532

Keulartz et al. (2004, pp. 22-25) suggest two strategies: ‘gradualisation’ and ‘common-ground dialogue’. ‘Gradualisation’ involves thinking in terms of degrees instead of rigid boundaries (e.g. in the case of debates between animal protectionists and nature conservationists about the moral problems associated with the introduction of domesticated large herbivores into newly developed nature areas, gradualisation implies exploring the moral middle ground between ‘domesticated’ and ‘wild’ expressed in a ‘respect for potential wildness’). ‘Common ground dialogue’ stresses the value of leaving core commitments of the discussion table and searching for areas where some agreement might be reached (e.g. in the case of debates between ‘pro-life’ and ‘pro-choice ‘ representatives in the abortion debate, the expediency to reduce abortion by preventing unwanted pregnancies (e.g. through educational programs) could prove to be acceptable to both sides). 533 For a critique of ethical committees and their functioning in political arenas, see De Dijn (2003) and Badiou (2005).


5.3

339

Stakeholder representatives

The stakeholder role is a very ambiguous one. At first glance, the definition of ‘a stakeholder representative’ seems to be quite unproblematic altogether: OXERA (2000, p. 19) define this as a “…person or organisation defending the interests and opinions of a group with an interest in the outcome of a particular policy decision…”. However, as mentioned before, the situation quickly becomes very complicated in a policy field such as energy policy. Extending the OXERA definition to include not only ‘persons’ but ‘actants’ in general means that future generations, the atmosphere, the ozone layer, ‘the people’ (as users of energy-consuming equipment, car drivers, citizens, etc.), industry etc. can all legitimately claim to be given a seat at the negotiation table. It is clear that the apparently unambiguous role of a ‘stakeholder representative’ is quickly shattered: in fact, this role is made up of a mixture of the other roles we discuss in this section. Indeed, a ‘stakeholder representative’ could claim that he/she represents a particular group of ‘actants’, based on a combination (to a greater or lesser extent) of 1. his/her competence to do so (i.e. taking up the role of a ‘scientific expert’); 2. his/her ethical conviction of having to do so, in particular for underrepresented ‘actants’ (i.e. taking up the role of an ‘ethicist’); or 3. his/her mandate to do so (i.e. taking up the role of a ‘politician’). Therefore, the role of a stakeholder representative will always be a hybrid one. However, since a large part of our discussion so far has been concerned with the functioning of ‘central agency’ located near the national legislative bodies, we will concentrate here on stakeholder representatives with an ‘official’ (i.e. formal) mandate to have a seat in this agency. Ideally, ‘problem representativeness’ for some aspects of energy policy (e.g. in the mission statement of an organisation) should be a criterion for receiving a mandate; but clearly, the number of organisations somehow active in the field of energy is simply too large to be practicable. Pragmatic compromises will have to be reached, ensuring that different perspectives (and in particular, previously unrepresented or underrepresented ones) are or become ‘sufficiently’ represented 534 . Conventions will play a large role here (which can be revised as part of the learning experience) – e.g. the major groups defined in Agenda 21 (also represented in the FRDO) could be considered, or the ‘traditional’ groups historically involved in energy policy (labour unions, representatives of SME’s, employers’ organisations, industrial federations, etc.). However, if the ‘central agency’ is to have real influence, it would be wise not to exclude ‘powerful’ actors from the deliberations, or actors which can be expected to contribute to future solutions (e.g. ‘niche’ players) 535 .

534

In the deliberative arena, the question of which perspectives will contribute to collective understanding will always be contested. But it is nevertheless feasible to make a persuasive though tentative case that a given representative deliberation included too few of the appropriate perspectives (Laes et al. 2004c). 535 This can be investigated through ‘reputational analysis’ (i.e. asking a number of people active in the energy policy field for their opinion on influential players) or ‘decisional analysis’ (i.e. analysing legislation in the field of energy policy over the last few years in view of the players which have been consulted).

340

Chapter 5

In any case, stakeholder representatives receive (or, if sufficiently powerful, in some way exact) a right to participate, but they do not have a duty to contribute and therefore cannot be held responsible for any part of the process. This is important, since a ‘boycott’ of the deliberations by one or more of the stakeholder groups represented might in itself be a (powerful) signal that the process is not considered to meet the quality criteria 536 . However, although stakeholder representatives cannot be held accountable for the results of the process, the quality and general effectiveness of the results will also depend on the quality of stakeholder representation (i.e. assuring that there are reasonable grounds for validating the stakeholder representative’s claim that he/she really represents the ‘wants’ and ‘needs’ of his/her constituency’), as well as on the ‘trickling down’ of the collective agreements reached towards the basis (e.g. the operational department of a business firm) of the stakeholder organisation (i.e. assuring that the learning experience is in fact mutual and goes in the same direction) 537 . Answering these questions lies beyond our present purposes, but nevertheless we give two ideas (one ‘soft’ and the other one ‘hard’) which might be taken up for further reflection and more detailed analysis 538 . The ‘soft’ idea is to develop a variety of specific strategies to enhance the flow of communication between an organisation and the external dialogues in which its representatives are engaged (Hunt et al. 2003). For instance it would be helpful to establish forums within organisations for communicating experiences from external dialogue situations. Not least, this would help to ensure that the representatives do not feel isolated or dissonant as a result of their experiences. Involving several people from the organisation in an external dialogue would also help in this regard. Ideally these people would be drawn from a range of different organisational areas. The ‘hard’ idea is to have legally prescribed ‘duties’ for stakeholder groups next to the ‘right to participate’ (Schmitter 1995). In fact, Schmitter proposes a fixed ‘constitution’ (or ‘general provisions’, as he calls it) for all groups, which describes their internal structure, their rights and the means to check on them from outside. Schmitter mentions for example a guarantee for democratic procedures within the organisation, a guarantee that public authorities will not intervene in the internal deliberations and choices, and a commitment to full public disclosure of associational revenues and expenditures (Schmitter 1995, p. 175). Only associations who would fulfil their duties according to those general provisions could gain and maintain a semi-public status. Schmitter concentrates less than other theorists of associative democracy on ‘traditional’ productively-defined interests (capital and labour); instead, he focuses on non-profit organisations (Schmitter 1995, p. 170). For regulating their influence in society, Schmitter even proposes a more ‘radical’ idea: the introduction

536

E.g. as was the case for the CCEG, which was boycotted by the socialist union ABVV in the years from 1977 till 1983. 537 This is a particularly contentious issue for social groups who ‘serve’ interests of constituencies but who are not ‘accountable’ to them (e.g. with regular interactions between ‘leaders’ and ‘members’) – NGO’s active in the field of environmental protection or North-South relations are excellent examples. 538 This is not meant to imply that we agree with these ideas.


341

of a system of ‘vouchers’. In this system, the idea is that every individual citizen gets a fixed amount of vouchers, with which he/she can ‘vote’ for associations which have a semi-public status, by giving them one or more of the vouchers. In this setting, being a member is not a requirement for voting for the association, and vice versa, membership does not oblige to vote 539 . By introducing ‘voucherism’, Schmitter establishes some kind of democratic mechanism, parallel to the traditional representative democracy, but the links between participatory and representative democracy are, because of the scrutiny provisions explained above, clear and strong.

5.4

Administrators

Of all the roles to be taken up in the model process, that of the administrator is the most straightforward one. Administrators are mainly concerned with the procedural side of the deliberations: assuring that the agenda is being respected, deadlines for reports are being met, the mandatory advisory channels are respected, adequate information is being provided, quality control is being upheld, etc.

5.5

Facilitators

Contrary to the administrators, it is very difficult to precisely circumscribe the facilitator role. Nevertheless, in many ways the facilitator functions as a cornerstone in the model process. The best description of a facilitator we have found is the one given by Rip et al. (1995, p. 353) in the epilogue to their book, i.e. “…an intermediary between the present situation and the future world…”. Broadly, this means that the facilitator firmly belongs to the collectivity as it is made up today (with an accepted state of knowledge, set of laws, procedures for interaction, etc.), but, in view of maintaining the learning dynamic, seeks to bring in new actants in a new iteration (e.g. providing incentives for the construction of new sustainable energy indicators). The facilitator role is more active than that of the administrator: instead of merely checking whether deliberations follow the agenda or enforcing (agreed-upon) rules of interaction, the facilitator goes further in actively participating in the debate, e.g. by making proposals, presenting new information or arguments which are felt to be missing so far, etc. In order to be able to do so, the role of the facilitator has to be taken up by professionals who serve a mediating role between the other roles (and therefore, should be regarded with trust by other actors) and who are accountable both to the ‘political system’ and the ‘scientific world’. Now, contrary to some other participatory models building on Habermas’s ideas of deliberative democracy 540 , we are not postulating here the existence of a so-called ‘guardian of the system’, i.e. an organisation with independent means of assessing and intervening on account of the legitimacy, fairness, competence – in short, the overall quality of the

539

Schmitter has thought of quite detailed arrangements concerning the distribution of the vouchers; his idea is that they could be administrated jointly with the yearly income tax. 540 Most notably, we are referring here to the RISCOM model intended as a transposition of Habermas’s ideas of ‘ideal communication’ in the context of radioactive waste management (see e.g. Espejo 2003).

342

Chapter 5

collective learning experience. Such requirements impose upon the facilitator the somewhat ‘heroic’ assumptions of a capacity to oversee the entire organisational system and to assess the quality of the interactions going on within it 541 . There can be no question of ‘neutrality’ either: a facilitator exerts influence on the collective construction process; and to exert influence means being an actor oneself. Furthermore, they will be acted upon by the other actors involved in the deliberations and in this way be drawn into the game. The facilitator plays a precarious role, as hard choices cannot be avoided: ‘Which new (or previously excluded) actants should be allowed to enter into the collective?’, ‘How should they be represented?’, ‘Which new pieces of information are trustworthy and important enough to be taken up for consideration?’, etc. The facilitator will have to take up responsibility for his/her choices; in any case, these choices will (and should) be subject to intense scrutiny of the other participants in the debate.

5.6

Politicians

The final role is reserved for politicians – i.e. elected or appointed institutional actors in the ‘political system’ (parliament, administrative bodies, ministry officials, political parties, parliamentary discussion groups and committees, advisory bodies, etc.). These ‘politicians’ are the only actors legitimately capable of setting up the governance structures we propose in this chapter. Now of course, politicians might not like to be ‘perplexed’ by new perspectives arising from these new structures. We will not discuss this difficulty here further, since this is one of the possible pitfalls we will address in section 7. Here, we merely point out that, in any case, politicians will make use of a lot of other complementary channels besides our proposed governance structure (e.g. party meetings, discussions with other politicians or trusted advisors, lobbying, etc.) for providing the information needed. This observation is not only meant as a statutory reminder for nurturing modest expectations; the point is also that political representatives are the most legitimate actors for translating the general public’s experience (which is not directly represented in the governance structures) with the performance of our proposed model process into policy. After all, political representatives have to be ‘experts’ (that is, if they want to be re-elected) at detecting and translating such ‘gut feelings’ regarding legitimacy, authenticity and/or effectiveness of a certain mode of governance. The value granted by ‘the public’ – which, in a sense, can be considered as an ‘external’ stakeholder to the process – emerges from the quality of the direct or indirect experiences with the process, most likely mediated through mass media channels. More generally – since we cannot simply assume knowledge of the instruments or goals of energy policy – ‘cultural repertoires’ governing the understanding of technological risk in the public mind (cf. Chapter 2) 542 can be expected to play an

541 Nevertheless, facilitators should be generally knowledgeable about most aspects of energy policy, including a conceptual understanding of scientific questions, the ‘room for manoeuvring’ of the different actors, notions of ethical reasoning and dilemmas involved, etc. 542 Of course, decision makers can be aided in this task by social science experts, who by now have developed an impressive ‘toolbox’ of public consultation methods – e.g. consensus conferences, focus groups, extended focus groups, deliberative opinion polls, public juries, etc. (for an overview, see Slocum 2003).


343

important role in setting up political priorities. As developed in section 3.3, politicians are the only legitimate actors for negotiating compromises and giving ‘closure’ to the debates in the form of new institutions (e.g. laws, standards, regulations, etc.).

6 A practical proposal So how can such an (admittedly) demanding governance structure be implemented in the realities of Belgian politics? It should be clear that we cannot formulate a fully-fledged proposal within the confines of one section of a chapter. Nevertheless, we believe a number of things can be said about general possibilities (possible pitfalls will be discussed in section 7). First of all, what are these political realities? While energy policy generally languished in the ‘80s and a large part of the ‘90s (due to the generally cheap supplies of oil and gas), recently the international environmental dimension (the Kyoto protocol) and the liberalisation of the electricity and gas market have propelled energy issues back on the political agenda. The focus for industry became on maximising the efficiency of the existing assets, while the dominant role on the policy side came to be played by the economic regulators (on the federal and regional levels). These are (at least formally) ‘independent’ of government, in the sense that they can carry out their duties and functions of promoting competition and regulating monopolies without regard to governmental concerns. At the federal level, the CREG is responsible for ensuring free competition on the electricity market. Policy intervention in this new economic reality focuses on providing financial incentives for renewable energy and cogeneration on the production side, and curbing energy consumption through a variety of instruments (for an overview and assessment of these instruments, see Fraunhofer Institute 2003). However, a Belgian governance model for sustainable development has to take into account the federal state structure, which means that certain competences and responsibilities relevant to sustainability are divided over the federal and regional governments. In the case of energy policy for instance, decisions on economic instruments such as fiscal policy (for a large part), labels and standards for appliances belong to the federal authority, while other aspects of energy policy (e.g. rational use of energy, promotion of renewable energy, etc.) fall under regional authority. This split in responsibilities has created a situation where no institution at the national level is able to create a strategic energy policy with a long-term view. It is of course not our intention to question this arrangement here; therefore, we accept this institutional division as given. In any case, it is clear that if the government is to develop an energy strategy, it will need some institutional focus to take the initiative forward. For reasons discussed above, it is not enough to simply rely on the initiative of business firms, steered by the ‘public’ networks already in place (e.g. consumer preferences for environmentally friendly

344

Chapter 5

products, financial incentives, etc.) 543 . It is also important for our proposal that a national strategy is developed, which implies that this strategy is developed and approved both by federal and regional authorities – in contrast to the present federal plan for sustainable development, which only applies to federal competences 544 . Fortunately, Belgian political practice already includes some institutional coordination mechanisms at the legislative, executive and administrative levels which could also be harnessed for our purposes 545 . One of the most promising instruments for providing the necessary institutional focus, coordination and political impact could be the organisation of an interministerial conference on sustainable energy. Such interministerial conferences are composed of ministers delegated by the different (regional and federal) governments, who discuss the specific themes for which they were set up in the first place 546 . Belgium’s commitment (made at the Rio+10 conference) to develop a national strategy for sustainable development, or more specifically climate change policy (Kyoto and beyond), could provide the necessary impetus behind such initiative. It would of course also be beneficial if the interministerial conference would be presided by an important political figure – e.g. the prime minister. Of course, the purpose of such conference would not be just to generate enough political goodwill or an agreement on general high-level principles or strategies. As we have argued over the preceding sections, one of the major improvements needed in energy policy would be the foundation of a national energy body capable of absorbing the learning experiences occurring in various ‘collectives’. Ideally, political deliberations in the interministerial conference should result in an agreement of cooperation between the federal state and the regions 547 . The principal advantage offered by an agreement of cooperation is that it concerns an instrument with a firm legal (constitutional) basis, not susceptible to short-term political trends and thus capable of ensuring the needed continuity in certain policy fields. On the downside however, negotiating an agreement of cooperation is a quite cumbersome process, involving inter alia an advice by the Council of State, approval by the federal and regional parliaments, official signatures of the participating governments, etc.

543

Because of the institutional split between ‘promotion’ and ‘control’, possible ‘information asymmetries’ between private companies and the government (or the public), the short-term focus prevalent in liberalised energy markets, etc. 544 This situation is bound to change however, since Belgium has committed itself at the ‘Rio+10 Conference’ (Johannesburg, 26 August – 4 September 2002) to the development of a national strategy for sustainable development by 2005. 545 See the FRDO’s advice on the vertical integration of sustainable development and multi-level governance (“Advies over de verticale integratie van duurzame ontwikkeling en multi-level governance” – FRDO, 18 Dec. 2003). 546 For instance, interministerial conferences on foreign policy or the environment already exist. 547 Such ‘agreements of cooperation’ are enshrined in the constitution (Art. 77). Their function is to promote the collective execution of separate competences, development of joint initiatives and/or the foundation of institutions in which the different policy levels are represented. Examples are the ‘advice council for bio security’ (resulting from the agreement of cooperation of 25 April 1997), the ‘coordination committee for international environmental policy’ (resulting from the agreement of cooperation of 5 April 1995), or the ‘national climate commission’ (resulting from the agreement of cooperation of 14 November 2002). Also, at the administrative level a consultation mechanism on energy matters between the federal state and the regions exists (ENOVER/CONCERE).


345

There are many possible routes for the creation of the ‘central’ energy body we propose. Although it is beyond the scope of the present chapter to provide a detailed assessment of all possible options, there are a number of broad scenarios that could be investigated. These could range from adding on to the capacities of the energy administration (part of the ‘federal government service for the economy, SME’s, the self-employed and energy’) through to the setting up of a new body. The first option has the disadvantage that it will inevitably carry over the interests, culture and ‘historical baggage’ of the original form. For instance, the energy administration is part of the government service for the economy, SME’s and the self-employed; which implies that its approach to energy policy will be focussed on safeguarding the competitiveness of Belgian enterprises in an international context. The point is not to criticise this institution or the policy solutions it emphasises, but merely to point out the possible consequences when using it as a basis for a new form of energy policy. A new institution for sustainable energy policy could take many different forms. However, the most important point here is that the new body should be able to play its role as a ‘boundary organisation’ between the different ‘spheres’ (the political system, the scientific world, organised stakeholders, etc.) and ‘levels’ (ranging from national energy policy to life world experiences). This can be achieved by a number of institutional provisions. The first would be to incorporate an overriding priority to sustainable development. While this might sound obvious, having sustainable development as an explicit mission statement could have the strategic advantage of creating enough ‘common ground’ (e.g. hooking up with strategies already committed to under the label of sustainability) to avoid a continuation of the conflicts of the past. It also allows a connection with the evolving discourse and institutional innovations at the international and EU level (and in particular, linking up with the evolving conceptions of ‘precautionary practices’ – cf. Section 3.1.1). This overriding duty to sustainable development could be organised around a medium-term ‘sustainable energy plan’, with (perhaps) an overall assessment to the interministerial conference of the goals realised on an annual basis (cf. the current four-yearly ‘federal plan on sustainable development’ and the two-yearly federal reports). Secondly, the institutional set-up would have to allow for participation of all the different roles we have discussed in section 5. This set-up could be similar to the one currently in place in the CREG (at least in formal terms): a ‘general council’ made up of the most important officially-recognised stakeholders in the energy field and representatives from different governments, overseeing the activities of a ‘management committee’ responsible for carrying out the tasks assigned to the new energy body, assisted by strong and independent expertise. The management committee should mainly take up the role of ‘facilitators’ and ‘administrators’ (individuals who mainly play the mediating roles). Accountability to the ‘scientific world’ should be secured by also including securing the energy body’s functioning as a centre of expertise within government (i.e. securing the roles of ‘scientific experts’ and ‘ethicists’).

346

Chapter 5

Providing the opportunity for the creation and use of ‘boundary objects’ and ‘standardised packages’ (e.g. energy models, indicators of sustainable energy provision and use) is in itself the third necessary condition for performing ‘boundary work’. Thus, besides of course having ‘scientific quality’ as a criterion, an at least equally important success factor for these ‘boundary objects’ would be features of non-exclusiveness and non-rivalness (Ishii 2002), meaning that the knowledge embodied in these objects can be shared equally among all participating parties. This condition helps in attributing notions of fairness among participants to the knowledge base and avoiding struggle for getting more benefits than others from its use. The general understanding should be that the energy body does not exist for the purposes of furthering the interests of one particular group (e.g. industry), but rather for deriving universal benefits for the ‘collective’ (however defined). The requirement of ‘collective learning’ – i.e. linking up with collective learning practices occurring at different depths of planning – also implies some careful institutional management. Several provisions could ensure a link with the ‘political system’. Firstly, it could be useful to give ‘politicians’ (e.g. representatives of different administrations) a seat in the general council of the new energy body as ‘observers’, serving as ‘political antennas’ for their respective policy fields. Secondly, the ‘products’ of the energy body’s activities – e.g. scenarios, indicator sets, policy advice, etc. – should be legitimated by receiving an official ‘recognition stamp’ by the government of the day. As mentioned before, this could be achieved by having a formal review procedure in place, perhaps organised around the regularly-recurring publication of ‘energy plans’ or ‘reports’ to the interministerial conference, with a number of obligatory consultation channels (e.g. the FRDO, ICDO) also being drawn into the process. The interministerial conference would then have a duty (perhaps reinforced by the commitments made through the Aarhus convention) to justify the policy measures they finally decide upon in light of the advice received. Furthermore, this arrangement is likely to afford the energy policy framework a relatively high degree of stability, since it would be difficult for a government that had actively promoted a specific set of strategic objectives to justify establishing ineffective policy measures or pursuing policy measures clearly in contravention to the previously ratified objectives. Thirdly, we would also argue in favour of a strong parliamentary involvement in energy policy, for instance through a recurrent parliamentary review of government initiatives (prepared by the relevant parliamentary working groups). Having a formal parliamentary review of energy policy in place represents a significant, albeit necessary, rupture with past political


347

practice 548 . Fourthly, the new energy body might also be given some administrative powers, particularly concerning international commitments made by the Belgian state. For instance, a leading role with regard to the registration and observance of carbon credits obtained by the flexible mechanisms proposed in the Kyoto protocol (i.e. ‘joint implementation’ and the ‘clean development mechanism’) or carbon trading. Linking up with ‘life-world practices’ can take advantage of a wide variety of public networks on energy-related issues already in place. These take both official and unofficial forms, and are both formal and informal. Examples of all of these include COGEN Vlaanderen (an organisation for the promotion of cogeneration in Flanders), ODE Vlaanderen (an organisation for the promotion of renewable energy in Flanders), research institutes – both public (e.g. VITO, SCK•CEN, etc.) and private, universities, etc. Public networks at the local level include e.g. local demonstration projects (e.g. the ‘Ecohouse’ in Antwerp), local Agenda 21 initiatives, local mediation concerning (mostly) renewable energy projects (for an overview, see Neyens et al. (2004)). To some extent, all of these types of public networks already exert influence on energy policy; however, in view of the arguments presented in the above sections (e.g. the control vs. promotion paradox), we suspect that it is not enough to rely solely on these received approaches 549 . The way forward in this area would seem to be a careful mapping of existing initiatives keeping in mind the requirements of the model process. Where appropriate, some of these local networks might be modified or strengthened in light of these requirements (e.g. one could argue in favour of a more independent assessment of the costs of decommissioning nuclear reactors), or others might be left over largely to the initiative of private enterprise where this seems to be the most promising route. New networks might have to be set up (e.g. a dialogue process on high-level waste management involving the most important stakeholders), or there might even be an opportunity to prune back the plethora of bodies that have been created over the past years, e.g. where learning experiences are ineffective or duplicated. In any case, the public networks thus built up should contribute to the enhancement of the legitimacy, influence and expertise of the collective learning process.

548

Belgian federal parliament however has not very often intervened in the energy sector. In 1981 the Belgian government presented its policy reaction to the 1973 energy crisis to the parliament that failed to present any strong global conclusions either to the government or to the general populace. The latest debate was announced for 1989 but never took place. In 1993, a parliamentary debate on the use of ‘Mixed-Oxide’ fuel (MOX) in the Belgian nuclear power plants was organised. This seems to have been one of the rare occasions were parliament, after a well-structured and in-depth debate, was willing to intervene in nuclear energy policy with the adaptation of a five-year moratorium on new fuel reprocessing contracts. However, the follow-up debate (which had to take place as foreseen in 1998) never took place, because this was considered to be inopportune in light of the decision to phase out nuclear power. 549 However, the influence of local participatory experiments in environmental or sustainability matters is, according to Bruyninckx (2002, p. 11), very limited: “…Very few cities have a local Agenda 21 that is in fact operational and having a significant influence on local policy making. To associate local Agenda 21 to the large input of participatory processes for this moment seems therefore like stretching reality. Recent research on local participation in local environmental policy making suggests in addition that the interests and motivations of many actors is not always very high. Only about one half of the members of the municipal environmental council for example are present at meetings. This number increases to absence rates of 70 to 80% for a number of groups represented in these councils…”.

348

Chapter 5

7 Possible pitfalls It is probably the case that many objections can be raised against our proposal. After all, as any innovation in political practice, the model process proposed in this chapter could possibly face some opposition because it can be seen as a quite ‘radical’ departure from established institutional and organisational routines 550 . We will address here what we consider to be the four most reasonably conceivable problematic issues (or objections) and attempt to respond to each of them in turn.

7.1

‘Your proposal will suffer from a lack of political and/or social support’

It is evident that institutional innovations need to be carried forward by political actors, and consequently, a lack of sufficient political and/or public support will seriously thwart their advancement. Indeed, a number of contemporary ‘symptoms’ could be read as indicators pointing at a lack of political support for the measures we propose: there are no dramatic external ‘shocks’ (e.g. an oil crisis, a nuclear accident) to the energy system which might mobilise enough support for taking political action; energy firms operate according to a short-term logic (awaiting a stabilisation of liberalised energy markets); uncertainties concerning long-term technological solutions are large (hence, it is too early to pick a ‘winner’); government’s response to the international climate change commitments has been generally reactive rather than pro-active, etc. In a way, even the Fraunhofer Institute’s (2003) message that the Kyoto goals can be reached by a simple benchmarking approach (i.e. copying proven energy efficiency solutions from foreign experiences) could be counterproductive to more radical changes. These are indeed serious challenges, and we are not denying their impact on the present energy policy agenda. For the time being, it is clear that some political voluntarism will be needed for implementing the changes we propose. Therefore, it is clear that a gradual implementation process offers more chances of success than sweeping changes. Small policy steps in the right direction might even be tested out in a protected environment at first – i.e. an application of the ‘strategic niche management’ idea to the governance model in itself. Niches could for instance be created by a first government initiative for developing sustainable energy indicators, setting in motion a process of building a knowledge network, discussions with interested stakeholders, etc. Or the federal planning bureau could be asked to develop long-term energy scenarios (since this institute is already developing short to mid-term scenarios), giving a first impulse to a wider debate. Such initiatives could thrive on a widespread (albeit still diffuse) understanding that the present energy system is not sustainable – e.g. in view of the ill-suitedness of most of the existing capital stock to meet the long-term requirement of a low-carbon economy 551 ; or, less 550

Since it proposes at the same time a new focus on uncertainty, a different framing of the problem and a new division of the roles between experts, laymen, politicians, stakeholders, etc (compared to ‘traditional’ practice). 551 The first (cautious) steps towards a Belgian commitment in the post-Kyoto period our now being taken – see e.g. the FRDO’s advice on a strategy to prevent climate change after 2012 (‘Advies inzake een strategie ter voorkoming van klimaatveranderingen na 2012’ – 26 November 2004).


349

idealistically, on an equally widespread understanding of a long-term need for energy security in an energy-intensive country such as Belgium. In general, the message should be that the resistance from particular stakeholder, or even government, can be potentially high; however, the cost of not acting – in light of the post-Kyoto commitments – might still be higher…

7.2

‘Your proposal is unethical’

A second possible criticism goes deeper (in a normative sense) than simply pointing out a potential lack of support. Rather than emphasising the pragmatic success factors, it is also conceivable that some actors might qualify our model process as being ‘unethical’. The point we wish to raise here is the following: ‘Can one assume contrafactually that everyone would or should be able to accept the outcome of a discourse organised along the lines discussed in the preceding sections?’ For instance, many would dismiss our move as a deliberate delaying of the decision-making process through discourse. From this point of view, the main causes of unsustainability (depletion of resources, threats to the environment, risks of nuclear proliferation, etc.) require solutions, and not just ‘diplomatic’ recognition and careful experimenting. Hence, in view of the perceived need for rapid and decisive action, our suggestions could be perceived as a dilatory strategy. Our answer here would be that our proposal does not necessarily lead to inaction; one could work at both fronts of (slowly) building up the required governance infrastructure whilst still taking measures which are generally accepted to be beneficial from the point of view of sustainability. In particular, we are referring here to policy measures directed at the demand-side of the energy system, where political action seems to be long overdue (Fraunhofer Institute 2003). Or a pragmatic reconciliation of ‘action’ and ‘deliberation’ could be first setting up an ‘energy efficiency agency’ (as recommended by the Fraunhofer Institute), coordinating the different governance levels and initiatives in the field, before further extending the mandate of this agency towards ‘sustainable energy’ as a whole.

7.3

‘Your proposal expects too much of intentional political planning’

A further critique could be that we paint a somewhat overly optimistic picture of intentional political planning. In fact, this critique can take many forms. Firstly, critics could argue that there is no such thing as an ‘intentional policy vision’; rather, such visions are exposed as being post factum reconstructions of the essentially emergent results of many ‘warlike’ (i.e. strategic) interactions between social and political actors. Or, secondly, one could argue that even if a consensual vision can be found, its transposition in reality will be largely ineffectual due to the much larger influence of ‘deterministic’ factors (e.g. prices on international energy markets, globalisation, liberalisation, etc.).

350

Chapter 5

Both objections are, we concede, quite tenacious; however, they are contradicted by historical evidence 552 . The picture emerging from historical studies is that, despite the autonomous character of certain developments and/or the influence of unintentional or unforeseen ‘surprises’, the evolution of the energy system (in terms of its speed or extent) is pointedly influenced by government intervention. To give but one example, the early choice for a nuclear power programme in Belgium was clearly predicated upon technological and political expectations: Belgian industrial activity in virtually all areas of the nuclear fuel cycle combined with the expected technological evolution towards ‘breeder’ reactors was supposed to ensure an autarkic and economically-beneficial energy future with virtually no resource constraints. If not for a political-industrial vision, one would be hard pressed to explain the choice for constructing nuclear power plants in 1966, at a time when oil resources were still abundantly available and cheap. Now, of course we are not saying that there are no constraints to policy intervention. On the contrary, broader dynamics such as the dynamics of international energy resource markets, liberalisation, an increasing shift of political influence to the international level (EU legislation), technology development in large multinational firms, etc. severely restrict the scope of such interventions. However, rather than changing the ‘weight’ of the state as a major player in energy markets, these dynamics have served to modify the form of policy intervention. The mechanisms of influence have shifted from the boardrooms of corporatist decision-making to more explicit policy instruments (e.g. taxes, levies, green current certificates, etc.) and regulatory control (e.g. emission standards, permits, spatial planning, etc.). Moreover, as Helm (2002) argues, the idea that governments could retreat from the energy policy scene and leave it to competitive markets is simply an illusion, because energy is just too important to the economy and society at large. Thus, our answer to the first and second objection would be that the government still has an – albeit changed – role to play as a ‘director’ of different actants. In the above sections, we have tried to give a realistic description of this role, which includes setting up the required dialogue platforms and institutional infrastructure, encouraging experiments (e.g. through strategic niche management), seeking out flexible and diverse solutions, etc. This role can even extend to international policy-making initiatives. Of course, it would be unrealistic to transform Belgium into an island of sustainable energy in otherwise unchanged surroundings. An aspect of governance which has perhaps been stressed less in the previous sections is that the networks we propose to build up can easily be extended to foreign experiences. For instance, international discussion platforms could be erected around promising technologies, networks of local participatory initiatives could be built up, etc. But perhaps the most serious (third) problem or question falling under this heading is whether the scope of reaching some form of consent within the sphere of political planning can be partially replaced or constrained by the discursive learning process we envisage. In

552

See e.g. Rotmans et al. (2000) for the history of Dutch energy policy; Helm (2002) for the UK; and Laes et al. (2004c) for the Belgian case.


351

particular, the general public could be in a bad position to understand such move. The question could be raised to what extent a decision reached through following the model process requirements could be ‘enforced’ in case of severe acceptance problems in society. Pragmatic political considerations would argue in favour of suspending the decision in such cases. However, severe acceptance problems might be considered as an opportunity for instigating yet other learning processes – this time in order to review the model decisionmaking process with respect to its suitability. A learning process might in that case result in a different, modified procedure. In the end, in the case of situations where accepted democratic decision-making procedures do not work, it is finally up to the sovereign (the citizens) to reverse this situation – acceptance of such modifications (and giving reasons for them) is ultimately based on meeting certain basic democratic mechanisms.

7.4

‘Your proposal quite simply comes too late’

A last possible objection could be that ‘the collective’ is already too profoundly committed to certain technological or institutional regimes, so that any new governance scheme can at best marginally contribute to any solution. Contrary to the previous objection, this one also has a strong moral component, and comes down to the following statement: ‘participatory decision making without real impact only serves as an ethical cover-up for decision made elsewhere and earlier’. This objection is not new. By now, many other commentators have championed new models of promoting public engagement in technological innovation with the intention of broadening the scope of discussion and involving ‘the public’ in debates on the technological shaping of society. A consistent finding has been that participation in the early stages of development has greater chances of success. For instance, Hunt et al. (2003) – in an examination of mechanisms for developing transparency and greater participation in radioactive waste management – conclude that it is essential to integrate ethical and social considerations in decision making. They argue in favour of early or ‘front-end’ consultation, and conclude that such upstream processes are more likely to lead to wider acceptability of risky technologies. The term ‘upstream’, Hunt et al. (2003, p. 6) note …designates the idea of conducting participatory consultation early and before the “waters have been muddled” by institutional commitments to a particular course of action…

This belief in upstream processes is founded in their consistent findings throughout different dialogue settings involving ‘the public’ that laypeople frequently wish to step back from immediate technical questions in order to ‘frame’ those within broader ethical contexts. Such findings seem to underline the importance of covering a sufficient number of aspects regarding the sustainability of energy systems – i.e. the possible contribution of nuclear power can only be discussed within a broader framing of the implications of energy provision and use on a global and intertemporal scale, covering a wide range of possibilities and options (e.g. demand-side management). In fact, this has been a demand

352

Chapter 5

of critics of the nuclear option ever since the first construction of nuclear power plants in Belgium (Laes et al. 2004c). Now of course, the problem is here that any innovative governance proposal does not take place in a political vacuum. Furthermore, governance processes need some institutional focus, but some actors might be willing to extend discussion towards other issues not necessarily falling within the remit of these processes. In its most extreme form, critics will argue that no true ‘local’ success is possible as long as international developments in the same area (e.g. development of nuclear power programmes elsewhere in the world) are not thwarted. We have already faced this objection on a theoretical level (in Chapter 1 – Section 3.3.2, where we have called this position ‘level-3 constructivism’), where we have tried to give convincing arguments in favour of a constructivist approach to governance. However, of course we cannot assume that everyone will be convinced by these arguments or will act upon the conclusions. Furthermore, one does not even have to accept the entire (neo-)Marxist theoretical apparatus to recognise that ‘big capitalist developers’ in large parts of the world almost always manage to push through their interests at the cost of sustainability concerns. Our answer to these concerns would be threefold. Firstly, on a ‘material’ level, one does not even have to accept the sustainable development logic to recognise that someday, nonrenewable resources will be depleted. Hence, the realisation that most of our energy infrastructure is unsuitable in view of the ultimate goal of a establishing a low-carbon economy is widely shared. This realisation, combined with international environmental pressures expressed in principles such as sustainable development or precaution, opens up a window of opportunity for more radical changes – although no consensus seems to exist yet regarding the precise technological path to follow, or when these changes might enter a breakthrough phase (Rotmans et al. 2000). Secondly, on a ‘procedural’ level, we submit that although our model process does promote a consensual solution wherever this is feasible, it does not impose this upon the participants. Thus, actors adhering to the views discussed in this section can voice their concerns and are not forced into accepting compromises they cannot defend towards their constituencies. Thirdly, participation in the ‘official’ governance structures proposed in this chapter of course does not exclude other forms of democratic action – e.g. protest marches, petitions, etc. In the previous section, we have already argued that such protests can even be seen as other instances of learning (i.e. antagonistic learning) the results of which might even be productively incorporated. Still, we recognise that our governance mechanism will only function properly in the ‘normal’ mode of democratic societies – severe upheavals could of course result in its suspension altogether.

8 Summary and conclusions In this chapter we have argued that enabling sustainable development in the energy sector requires a shift in energy governance. Objectives and means are not given in advance,


353

since none of the ‘actants’ involved – i.e. the companies, the authorities, the liberalised energy market, future generations, the environment etc. engaged in flexible but uncertain network bonds – allow to have a precise idea of what a sustainable energy production and consumption system might look like, and how it might be realised in practice. Therefore, models which envisage unilateral solutions (e.g. the rather singular focus on nuclear energy in the past; or perhaps a singular focus on efficiency measures and renewable energy in the future) have great chances of not reaching their objectives, since their realisation depends, in fact, on a high number of variables of very different nature on which the actors involved often have limited margins of freedom. Nevertheless, the very idea of sustainability implies that a longer-term basis for energy policy is established. Combining both uncertainty and sustainability means we have to develop new methods of ‘due process’ to complement the more traditional economic and social objectives of energy policy. Moreover, these processes have to be recognised and organised openly as collective experiments, aimed at facilitating a gradual and progressive specification of (policy) objectives and flexible and diverse measures. Therefore, the model process outlined in this chapter goes far beyond the traditional role of technology assessment as conceived in advisory science, with its in-built presumptions of linearity from problem structuring to ‘optimal’ policy recommendations. Rather, our aim has been to influence interactions while trying to implement a dynamic relation between knowledge and action. This chapter elaborated on some practical requirements for a successful implementation of such dynamics. In particular, we have argued that the quality of such dynamics will crucially depend on the creation of a ‘platform’ for collective learning, based on shared rules of interaction and clearly defined roles. This platform, which we have referred to as a new ‘energy agency’, would have to serve as a consultative body with an output towards the formal democratic institutions; but still, should be located at some ‘distance’ from the deliberations going on in these institutions.in order to capture learning experiences and advice government and parliamentary bodies on an ‘independent’ basis. The energy agency’s functioning would crucially depend on meeting some fundamental requirements, set out at the beginning of chapter 5: relying on sound science, bridging the science-policy boundary, an overriding duty to the ‘good common world’ through collective learning (based on temporary, not permanent, closure), an inclusive dynamic (i.e. actively seeking out ‘under-represented’ perspectives) resting on multi-tiered learning experiences, quality control and adequate representation of uncertainties. These requirements could for instance be set out in the agency’s mission statement or in the legislative texts calling for its creation and describing its functioning. However, it is of course not sufficient to merely proclaim an adherence to such ‘pious’ principles – they have to be realised ‘in the field’ through actual socio-material practices. Here, we have argued that the work performed by our new ‘energy agency’ can be aided by creating network links with existing practices, in particular: • Harnessing the emergence of an international discourse on precaution (arguably the most ‘developed’ subsidiary sustainability principle) as a design basis for the new governance structure;

354

Chapter 5

• A broad framing of the sustainability debate, supported by the creation of ‘boundary objects’ (e.g. criteria and indicators of sustainable development) accountable to both the ‘scientific’ and ‘political’ world; • Setting framework conditions for sustainability in the energy sector based on a broad cost-benefit analysis informed by socio-technical foresight exercises (‘vision assessment’); • Enabling precautionary learning through the choice of transformable standards and flexible and diverse measures, trying to avoid the creation of irreversible situations; • The indirect influencing of ongoing technological changes through a multi-actor (actant) decentralised form of control; • The creation of ‘forceful sites’ for learning experiences (e.g. strategic niches); • The formalisation of a review procedure with regard to the implementation of policy measures, the knowledge base and the process itself; • Better anticipation of potential ‘surprises’; • Quality control based on finding an adequate balance between fairness, competence, management of uncertainties and transparency; • A representation of different ‘roles’ in the governance structure. The actual functioning of the ‘energy agency’ could for instance be organised around publication of a mandatory rolling ‘sustainability-of-energy’ statement. The energy agency could even incorporate some of the numerous bodies that have been created in recent years, or weed out others. No doubt we still have a long way to go in improving our governance processes for sustainable development, and we are certainly not claiming that we have given any definitive answers. But we believe one of the crucial messages is that although society has spend a lot of time, money and resources on trying to find ways to make energy policy problems technically ‘structured’ (e.g. through the use of mathematical forms of risk analysis and evaluation, complex economic modelling efforts and foresight exercises, etc.), the area of managing these problems in complex societal settings remains comparatively underdeveloped.

CHAPTER 6 CRITERIA

AND SCENARIOS AS A SUPPORT FOR SUSTAINABLE ENERGY

GOVERNANCE In chapters 6 and 7 of this dissertation, an attempt is made to put the first stages of our model governance process (problem structuring and scenario building) into operation in the form of a limited ‘pilot exercise’. In the present chapter, a comprehensive list for evaluating energy systems is derived from the interviews with representatives of the FRDO (cf. Chapter 4). A hierarchical representation of criteria was logically structured for each interview result separately, and then aggregated into a ‘combined value tree’ (Section 2). With the aid of the energy model MARKAL we developed four broadly conceived longterm energy strategies for Belgium with contrasting economic, social, environmental and political implications. Section 3 contains an in-depth discussion of these scenarios in terms of their central assumptions, hypotheses and results. The capacity of both criteria and scenarios for functionally supporting the energy governance framework proposed in chapter 5 will be empirically tested in the next chapter (Chapter 7).

1 Introduction In chapter 5, we set out the fundamentals of a new governance structure that would be able to ensure some measure of stability (in view of achieving the long-term goal of sustainability) while still taking into account the rapidly changing dynamics of the energy system. In our attempt to negotiate the difficulties engendered by these contradictory givens, a large role was reserved for both the institutional and the formal analytic dimension. The institutional dimension was concerned with the creation of ‘boundary organisations’ – i.e. organisations with a sufficient representation of different roles (e.g. scientists, ethicists, politicians, stakeholders, etc.) in order to uphold accountability both to the formal ‘political system’ and the ‘scientific world’. The formal analytic dimension on the other hand entailed the creation of ‘boundary objects’ (e.g. sustainable energy criteria, scenarios, models, etc.) as a result of the negotiations going on in the ‘boundary organisations’. The present chapter is concerned with setting out a suggestion for two of these ‘boundary objects’, that is a structured value tree of sustainable energy criteria and a set of long-term energy scenarios supported by a modelling tool for discussing the strategic dimension of long-term energy governance. Fundamentally, the use of such formal analytical tools should help to ensure that the quality criteria set out in chapter 5 can be met – i.e. the approach should be more conducive towards a fair, transparent, competent and uncertainty-sensitive treatment of policy options. This is necessary since the simultaneous consideration of quantitative, qualitative and intuitive aspects of a complex problem such as future energy policy is usually a difficult

356

Chapter 6

task for an untrained human mind, as different cognitive biases loom on the horizon. For example, a politician making up his mind on energy policy measures may easily escape the duty of taking into account all relevant actants and facing trade-offs by focussing exclusively on part of the information which is new, easy to understand or compatible with earlier preferences and knowledge (Hämäläinen 1990). In this case, a decision aid such as a structured list of ‘actants’ to be taken into account in arriving at a decision is imperative for assuring some consistency and coherence. It can for instance also enable focussing the debate on the crucial issues by eliminating unimportant ‘actants’, and make different ways of reasoning understandable to others. However, all of these advantages should not make us blind to the fact that such formal analytic methods also potentially face some difficulties, partly due to methodological foundations (which constrain the types of models than can be justified), but also partly due to the inherent ‘power play’ in political debate and deliberations. For instance, the use of formal decision analytic tools shifts the discourse into a more structured framework and, by doing so, inevitably affects the stakeholders’ negotiation positions. Thus, besides the ‘formal’ part of decision aids – i.e. the ways in which scientific information and knowledge about ‘actants’ is managed from a modelling perspective – the ‘interactive’ part – i.e. the interactions and deliberations through which the different actors contribute to the analysis and take part of its results –is also of major importance (Rauschmayer 2000). In this chapter, the ‘formal’ part will be discussed, as we will explain and defend our choices for particular decision analytic methods, namely a value tree analysis (cf. Section 2) and engineering-economic modelling of long-term energy options (Section 3). Discussion of the ‘interactive’ part will be left over for chapter 7. However, before going into the details of the formal part of the analysis, we want to warn against one possible major misconception. In no way should the methods proposed here be considered as definitive (be it just for the fact that the ‘protected experimental settings’ of PhD research differ fundamentally from the ‘real-world settings’ of a concrete governance initiative); on the contrary, they should merely be regarded as a first instigation – a limited ‘pilot exercise’. Chapter 7 will conclude with the lessons that can be learnt from this exercise, pointing out ways for solving some of the difficulties encountered.

2 A structured value tree for sustainable energy criteria

2.1

Methodological approach

One of the first tasks we have set out for a new energy agency (which occupies the ‘central’ role of collecting and transmitting learning experiences occurring in the enlarged ‘public sphere’) was to draw up a list of ‘actants’ (including the criteria they should uphold) which should have a seat in the jury deciding about the sustainability of our energy system. Ideally, the initiative should be taken by an interministerial conference (presided

Criteria and scenarios as a support for sustainable energy governance

357

by a high-level government representative) and entrusted to a group of experts of international standing which take up their role as ‘facilitators’ (cf. Chapter 5, and also Boulanger et al. 2003). This procedure is of course entirely unrealistic within a strictly research-oriented framework. For our research purposes, we have chosen to apply a method which has already proven its value also in energy-related questions (Keeney et al. 1987), namely the ‘value tree’ methodology. A value tree identifies and organises the values of an individual or group with respect to possible decision options. In the process of structuring a value tree, representatives of different stakeholder groups are asked to identify their criteria and objectives for evaluating different options. Translated in constructivist parlance, we could say that a value tree helps in establishing the ‘jury of actants’ who are considered to be relevant for the decision at hand (i.e. the long-term evolution of the energy system)553. A value tree structures the elicited values, criteria, and corresponding attributes in a hierarchy, with general values and concerns at the top, and specific criteria and attributes at the bottom. Theoretically, there are two approaches to generating value trees, namely top-down and bottom-up. The topdown approach is deductive-analytic, starting from general principles going down to more operational criteria, while the bottom-up approach is inductive and synthetic (Salo 1999). For our purposes, we chose to combine both top-down and bottom-up probing concepts, with however, in a first step, an emphasis on the latter. This first step entails a construction of individual value trees, derived from the results of interviews with representatives of stakeholder groups participating in the FRDO554. These individual value trees have been constructed on the basis of an interview protocol (see Annex 1) probing into questions such as what attributes or measures would be appropriate for differentiating between energy options, or asking why a specific alternative was considered to be ‘good’ or ‘bad’. In a first step, this approach has generated a list of criteria, at first without a concern for logical consistency or redundancy, but rather for completeness. To encourage interviewees to generate such a broad spectrum of criteria, at the time of the interviews we stressed that no attempt would be made to assess the relative importance of different values. Logical structuring was introduced only after the interview sessions. This process involved clarifying meanings, identifying means-ends relationships, eliminating redundancies, etc. – an account of which can be found in chapter 4 (Section 4). However, one important consequence of the bottom-up approach is that indicators of sustainable development are chosen and evaluated according to other criteria than when they are defined by a top-down approach. Criteria such as accuracy, standardisation and consistency will likely play a more limited role in the selection of indicators than can be expected from a top-down approach. As the upshot of ‘bottom-up’ participation and involvement, the creation and use of indicators of sustainable development serves as 553

Although of course Latour would resent the denomination of a ‘value tree’ (with the implication that ‘facts’ can be separated from ‘values’), we have chosen to keep this denomination in view of its simplicity of communication towards a broader audience. On a similar note, ‘evaluation criteria’ should be read as a translation of Latour’s ‘matters of concern’. 554 These are on file with the author.

358

Chapter 6

tangible evidence that the ‘members of the collectivity’ in question have asserted control over self-reflective definitions of the ‘good common world’, which likely includes attributed meanings beyond strictly quantitative indicators. This is an explicitly soughtafter quality, since it will likely enhance the (political) legitimacy of the value tree in the eyes of the participants in the discussion.

2.2

Construction of the combined value tree

No matter how expedient, this fundamental quality of the bottom-up approach is at the same time one of its main weaknesses. For if a list of criteria is to function as a ‘boundary object’ (towards the scientific world, international discourse on sustainable energy, etc.) measures of standardisation, coherence, etc. are precisely needed. Therefore, when combining the individual value trees into a combined tree, we have explicitly expanded the list in order to include criteria for any ‘holes’ that could be identified. For this purpose, we have used international reference documents (e.g. UNCSD 1996), overviews of existing indicator frameworks (e.g. Boonekamp 2002; Zuinen 2004) and official planning documents of Belgian policy (e.g. FPB 2002; ICDO 2000, 2004). The combined value tree has been discussed in a joint meeting555 in order to still ensure its validity towards the participants. In this workshop, we explained that we were not looking for a hierarchical ordering of values, but rather for the creation of a value tree where a) the relationship between the lower-level criteria and higher-level categories is one of inclusion; b) interdependencies between categories are avoided; and c) an exhaustive and non-redundant list of criteria is created. The latter requirement should however be qualified in view of the purposes of the exercise (i.e. a discussion on the long-term strategic orientations of the Belgian energy system): of course, certain topics could be split into much more detailed criteria and indicators (e.g. energy use over different economic sectors), but these subcriteria and indicators are more relevant for discussing policy measures in particular policy fields (and could therefore be assigned to the mandate of specific subsidiary institutions). In view of the three above-mentioned requirements, the group present in the workshop was encouraged to discuss the proposed tree with their colleagues, to provide comments for revisions, and, in particular, to add values that might have been left out. As a result of these discussions, inevitably changes had to be made to the proposed structure. The final combined value tree (included at the end of this chapter) has incorporated these changes. The combined value tree has been constructed on the basis of the insights gained from applying Boltanski and Thévenot’s commonwealth model (cf. Chapter 4 – Section 4.3). As explained in chapter 1, Boltanski and Thévenot’s model shows the interesting feature of being able to combine the ‘top-down’ (i.e. the development of coherent and intersubjectively plausible ethical structure) and ‘bottom-up’ (i.e. an ability to include all possible judgments of a particular actor) approaches to structuring ‘matters of concern’. The combined value tree starts from high-level categories (i.e. le principe du bien

555

Called a ‘scenario workshop’ (23 april 2003) – see Annex 1 for the attendance, and Keune et al. (2004) for an evaluation of the results.


359

commun), which cover concerns voiced in the language of the ‘industrial’ commonwealth (category I – ‘Environmental & human health and safety’ – i.e. mostly quantifiable impacts on the environment and/or human health and safety aspects), economic concerns mobilising ‘actants’ in the ‘industrial-market’ world (category II – ‘Economic welfare’) and criteria mostly stemming from the ‘civic commonwealth’ (category III – ‘Social, political, cultural and ethical needs’) – i.e. criteria having to do with governance (laws, rights, democracy, etc.) and institutions (civil society, media, etc.). Within each high-level category, equity concerns (i.e. le principe de commune dignité) were also covered556. For instance, in the ‘Environmental & human health and safety’ category, we included the concern for an equal distribution of risks and benefits – although this argument probably reveals more of a ‘civic’ concern for equal rights. Therefore, we must also admit however that sometimes we had to sacrifice the ‘purity’ of the different commonwealths in order to arrive at a somewhat workable result. ‘Aesthetic impacts’ for instance probably belong more in the ‘inspiration’ commonwealth; and the ‘need for longterm management of high-level waste’ is in itself a problem of such complexity that it probably involves arguments from all commonwealths (thus transcending the boundaries of a simple concern for radiological health impacts)557. We separated ‘Health and safety’ concerns into effects on individuals and effects that potentially threaten society (or even mankind). This separation allows a clearer accounting of the importance (for some) of catastrophic risks as expressed in the interviews. The second category ‘Economic welfare’ covers concerns about the costs, efficiency, security and market consequences of energy systems. We separated this category into the overall economic implications of the energy system on the market economy (e.g. security of supply, overall costs of the energy system, etc.) and issues of concern to more specific segments of the market economy (i.e. producers and consumers). This reflects a concern for distributional equity also clearly voiced in the interviews – i.e. fluctuations in energy prices, investments in new infrastructure, etc. affect different economic actors differently; hence it is desirable that the energy system puts a somewhat equal burden on different actors. The ‘Need for government intervention’ reflects the oft-raised objection that certain components of the energy system have been (or will be in the future) dependent for their success on government subsidies (see e.g. the objections raised against nuclear power from a historical point of view).

556

Other aspects of the ‘commonwealth grammar’ will become more clear in chapter 7, where the combined value tree is used in a multi-criteria mapping exercise. For instance, concerning equity, this exercise made a difference between mid-and long-term impacts, allowing users to make their assessment of intergenerational equity more explicit. The scoring performed in this exercise allowed for a measurement of ‘grandeur’ (principe de l’ordre de grandeur) according (in most cases) to a particular actor’s framing of the criterion in question. This also allowed issues of spatial equity (i.e. framing concerns on a local, regional or global level) to enter into the debate. 557 See Schrader-Frechette (1991) for a summary overview of the ethical dilemmas involved in radioactive waste management.

360

Chapter 6

The third category ‘Social, political, cultural and ethical needs’ was admittedly the most difficult one to assign criteria to. We broadly broke this category down into criteria regarding individual ‘civic’ development (in the sense of having the possibility to exert ‘sub-political’ influence on the shaping of the energy system – e.g. through consumer choices or direct citizen participation in decision making); institutional (or formal political) needs/benefits relate to the impacts of energy systems on democracy and its institutions, as well as the characteristics of the process by which energy decisions are made. The fourth category ‘Diversification’ will be discussed under section 3.4.3.4. Most of the labels in the combined value tree are, we believe, reasonably selfexplanatory; for others, we gave some clarification. We discuss the qualitative criteria (i.e. criteria for which at this stage we had not yet defined quantitative measurement) here; quantitative criteria are discussed under section 3.4.3: • Economic risks: fluctuations in energy prices or availability might affect economic development in the sense that abrupt price changes not only have direct costs, but also lead to perturbations of the economy which generate indirect costs. This criterion refers to these indirect costs; • Ability to provide specialist market: this criterion reflects the perspective that some developments in the energy system might be more conducive to the creation of new market opportunities, e.g. delivering energy services instead of energy as a commodity, highly efficient equipment, etc. • Strategic factors for export: contrary to the previous criterion, this one stresses the position of Belgian industry in the global market, i.e. certain developments in the energy system might allow certain economic sectors to acquire a leading position on the global market (possibly also contributing to the ‘Belgian image’); • Consumer choice: reflects a concern for the possibility of exerting ‘sub-political’ action through consumer behaviour – i.e. having a free choice of energy products or services from a variety of providers; • Citizen participation: reflects a concern for the possibility of exerting ‘formal’ democratic influence on the shaping of the energy system; e.g. certain developments would score better if the parliamentary influence is perceived to be higher, or if more political parties are actively involved in the energy debate (political pluralism), or if the influence of corporatist decision making is mitigated, etc.558; • Contribution to rational energy use559: reflects the concern that a focus on certain technological solutions to the problems raised by the energy system (e.g. end-of-pipe

558

International comparative measures of democracy already exist, see e.g. the ‘Freedom House index of political rights and civil liberties’ (available at http://www.freedomhouse.org/research ). Such indices could serve as a basis for the more specific energy policy context. 559 In countries with a high share of nuclear power production (e.g. Belgium and France), there is always the risk of attempts to promote electricity consumption in order to facilitate nuclear power’s penetration into the market. For instance in the ‘70s, Belgian and French utilities took intensive commercial actions to promote the thermal applications of electricity (Laes et al. 2004c; Romerio 2005).


•

•

•

•

361

solutions, nuclear power) diverts the political attention from the need for a more rational use of energy (i.e. from this point of view, a lock-in into a sub-optimal solution occurs); Control and concentration of power: this criterion reflects the concern that large energy companies might control the market through the use of market power (cf. London Economics (2004)), thus disrupting the functioning of the ‘pure’ market commonwealth; Influence on political decision making: in addition to the previous criterion, this one reflects the concern that market power might also be translated into political power (e.g. through lobbying); Need for socio-political stability: reflects the point of view that certain components of the energy system (e.g. nuclear power) can only function properly if enough stability is ensured. As such, socio-political stability is related to both social and formal political institutions. Institutional stability needs to be checked together with the degree of political democracy: dictatures can provide for a large amount of stability, but are not accepted as being generally conducive to sustainable development. Institutional stability could for instance be made operational through indicators of social justice, degree and evolution of economic development, perceived legitimacy of government initiatives (e.g. demonstrations, frequency of government changes), freedom of the press and external factors (degree of regional integration, relationships with neighbouring countries, etc.); Need for direct political intervention: this criterion describes the amount of effort exercised by government to influence the energy system evolution. Indicators could include the amount of government expenditures on energy policy compared to GDP, the fraction of energy use falling under benchmarking agreements, the fraction of taxes in energy prices or different sectors, etc. (Boonekamp 2002). It is likely that this criterion will be perceived differently by different actors – i.e. some will consider direct political intervention a necessity, others will adhere more to the rhetoric of the free market.

All in all, we have tried to utilise ‘substructures’ of individual value trees to the maximum extent possible in the combined value tree in order to maintain the logic and content of most people’s reasoning. Subsuming these ‘substructures’ under a generalised perspective ensures the comprehensiveness of the effort and principally allows a confrontation of viewpoints (e.g. when using it as a multi-criteria mapping tool – cf. Chapter 7). While of course the limited number of interviews and the individual perspectives of the interview participants (and inevitably also our own biases) make it doubtful that the tree represents the views of ‘Belgian society’, we feel that it is reasonably comprehensive. Nevertheless, our combined value tree should be regarded as a first attempt: if new (or other) entrants in the debate do not agree on its logical structuring or completeness, it can be used as a starting point for improvements. Also, as mentioned before, many low-level criteria can probably be split into further sub-criteria, in view of proposals for measurement. Furthermore, we could not exclude that the understanding of the different criteria still differed somewhat from one participant to another. Whether there still was ‘sufficient’ common understanding of the criteria to function properly is an issue we will explore further in chapter 7.

362

Chapter 6

3 Sustainable energy scenarios

3.1

Introduction

The second step proposed in chapter 5 was to use the list of ‘actants’ in order to draw up scenarios of possible futures (by assigning a ‘role’ to each of the actants and/or excluding others), in order to enable a strategic discussion about the long-term orientation of the energy system. Such scenarios also function as ‘boundary object’, hence the need for different lines of accountability to different ‘worlds’. For instance, scenarios will need to be ‘realistic’ enough in order to convince scientific experts; their scope will have to be broad enough in order to address key concerns of different stakeholders; they will have to be simple and transparent enough in order to promote understanding in different groups which are not necessarily specialists in scenario techniques, etc. Thus, these requirements for long-term scenario explorations being established, one could wonder what type of questions should be addressed in such a study (the ‘content’), and how they should be addressed (the ‘approach’). Long-term energy scenarios on a global scale have already been developed by institutes such as the IIASA (1995) and the IPCC (2001). Key issues such as growth in world population, economic growth, depletion of fossil resources, and the risks linked to global warming are discussed and should of course also be addressed here (population growth being less important on a more local or regional scale). In the Belgian context however, most effort is invested in finding nearterm solutions in order to comply with the obligations contained in the Kyoto protocol560. Demand-side measures, cogeneration and renewable energy are proposed as solutions, although other (technical) options for GHG emission reductions, such as photovoltaic power, fuel cells, and carbon capture and storage could show significant potential but will not capture the market at short or mid-term notice. Also, technical breakthroughs in nuclear electricity generation addressing issues such as safety and risks of proliferation could be possible on a longer timescale. As greenhouse gas emission reduction targets will most probably by tightened over the coming decades, any long term option commands careful consideration. Concerning the approach, reports by the Federal Planning Bureau (Gusbin and Hoornaert 2004) and the modelling exercises reported in CES (2001) could qualify as exemplary foresight exercises on a mid-long term (horizon 2020-2030). However, for the present study (whose main point of departure lies in the explicit consideration of value perspectives), we have chosen a somewhat different approach. We recall here the distinction we have introduced in chapter 5 (Section 3.2.1.) between types of scenario exercises which are meant to be primarily descriptive or exploratory scenarios, i.e.

560

This statement should now be qualified somewhat in view of a recently published report containing four long-term scenarios (horizon 2050) aimed at giving an input to the Belgian position in the upcoming postKyoto negotiations (Econotec and Vito 2005).


363

scenarios describing possible developments starting from what we know about current conditions and trends, and primarily normative, anticipatory or backcasting scenarios, i.e. scenarios which are constructed to lead to a future with a specific subjective value defined by the scenario developers. As explained, they differ in terms of overall purpose. The choice between exploratory and anticipatory approaches depends on the objectives of the scenario development exercise. Anticipatory scenarios represent organised attempts at evaluating the feasibility and consequences of trying to achieve certain desired outcomes (or avoid the risks of undesirable ones). Exploratory scenarios (or ‘what-if’ analysis), on the other hand, try to articulate different plausible future outcomes, and explore their consequences. The emphasis is mostly on prioritising technological choices, the analysis is performed in a relatively closed process by technically or economically schooled experts, and the government (or administrative bodies) mostly assumes the role of client. The above-mentioned examples seem to fall mostly in this category. For our purposes, a combination of backcasting from a set of ‘minimal’ objectives and foresight from initial conditions and drivers has been chosen. However, different transition routes (based on major technological options) towards achievement of the desired end-state have been modelled. It is important to be clear from the outset that any consideration of prospects over a 50 year timescale must be very uncertain. Our projections and technology assessments will inevitably turn out to be inaccurate. But this does not invalidate the exercise. Our approach should primarily be viewed as an approach to bring underlying value perspectives into sharper focus, and hence, to promote a more informed debate. Uncertainty is unavoidable and should be factored in the analysis. The scenarios we have developed for our exercise have been discussed and modified in a ‘scenario workshop’ (for attendance, see Annex 1) in order to ensure the general transparency of the effort. The practical multi-criteria exercise we have carried out subsequently (cf. Chapter 7) allowed us to test this hypothesis further. 3.1.1


We have chosen to follow the backcasting methodology as described by Anderson (2001). Anderson also points at some significant advantages of backcasting when considering policy planning for sustainability: the explicit consideration of value perspectives (whereas traditional foresight is less candid) and the possibility to formulate a strategic normative vision on long-term technological options (whereas traditional foresight focuses more on short-term market forces). The methodology distinguishes six distinct steps or building blocks of scenario development: 1. definition of strategic goals; 2. description of present-day supply and demand structure; 3. choice of end year; 4. demand-side analysis; 5. supply-side analysis; 6. development of a set of policy measures to meet the strategic objectives.

364

Chapter 6

How these ‘building blocks’ have been used to define the energy scenarios in the present report will be the subject of the following sections. 3.1.1.1 Definition of strategic goals The main goal of the present chapter is of course to identify possible scenarios (or strategic options) for a sustainable energy future. We have based our judgement on what is considered to be the present-day ‘minimal consent’ on the concept of a sustainable energy system. That is, some of the criteria described in section 2 were given priority over others as ‘meta-criteria’ which must be met by any sustainable energy scenario. Hence, none of the scenarios described here can claim to represent a definitive vision of sustainability – all scenarios involve (often difficult) tradeoffs. The choice of ‘meta-criteria’ is an unavoidably subjective one; therefore, it must be well documented. We have been guided here by the following considerations: • The ‘minimal’ objectives should be in line with strategic objectives as set out by official policy documents. The most notable exception is of course the government decision to phase out existing nuclear power plants, since we want to discuss the possible future of nuclear energy in the first place; • Energy should in any case remain reliable and affordable. Our choice of an economic optimisation model (MARKAL, cf. infra) guarantees the choice of a least-cost energy system (under different constraints and economic assumptions). However, with this model, it is impossible to assess the reliability of the energy system (in terms of forced outages, matching demand and supply, quality of electric power, etc.). Therefore, for our present purposes the reliability of the energy system simply has to be assumed, pending more detailed analysis with the aid of electricity sector models; • The ‘minimal’ objectives should be ‘politically stable’ on the long term; • The ‘minimal’ objectives should of course be internally consistent (i.e. noncontradictory). Furthermore, as a small nation, Belgium will of course be very dependent on evolutions in the international energy policy scene. Therefore, we also have to assume that these international conditions will be in line with the goals set out in the Belgian context. Thus, these considerations led us to adopt overriding criteria in the domains of security of energy supply, economy and ecology. These criteria will be developed further under section 3.3.1. 3.1.1.2 Description of the present-day supply and demand structure The so-called initial conditions of the energy system: • Technical, economic and environmental parameters for the different technologies making up the energy system; • An adequate description of the entire energy chain: from the use of primary energy (natural gas, oil, coal, natural uranium, or renewable energy) through secondary energy


365

carriers (fuels, electricity, etc.) to the provision of energy services (transportation, space heating, lighting, etc.). 3.1.1.3 Choice of end year Pragmatic considerations have led us to the adoption of a 50 year time horizon for our scenarios: this allows for a broad range of structural and technological changes to come into play to meet the requirements set out above, whilst not being too remote in order to influence present-day decision making. 3.1.1.4 Demand-side analysis For this issue, Anderson (2001) proposes to: • Use an exogenous estimation of the demand for energy services, common to all scenarios under consideration. This is done in order to make the scenarios more comparable on an equal basis. The rationale is that deliberate policy measures will have less impact on the actual demand for energy services rather than on the ways these demands are met; or, in any case, that the demand for energy services will be very difficult to predict (and hence difficult to manipulate by deliberate government intervention), especially on a long time scale and on a national scale; • The resulting demand for final and primary energy should however be determined endogenously (within the model); • The demand-side analysis should proceed from an evaluation of the state-of-the art technological options to provide for a certain energy service, and an evaluation of the policy instruments necessary to introduce these technologies into the market. In this chapter, the baseline demand for energy services in different sectors (industry, commercial, households, transportation) has been determined with the aid of a general equilibrium macroeconomic model for the 15 countries of the EU (the GEM-E3 model) for 2030 by the ‘Centre for Economic Studies’ (CES, KULeuven). For 2050, results have been extrapolated (and sometimes reviewed by the author). The drawback to this approach is that shifts in the economic activity of industrial sectors (e.g. a shift to an economy with increased recycling of used materials, shift to less energy-intensive sectors such as services), or shifts in level of energy services demanded by consumers (e.g. conscious efforts to reduce the number of kilometres driven by car, lower indoor temperature, etc.) are not explicitly addressed (econometric modelling implicitly assumes that historic trends, expressed as statistical correlations, will still prevail in the future). Nevertheless such profound structural changes are often demanded by stakeholders and are seen as a precondition for sustainability. Some studies do try to accommodate for these more profound structural changes in society. For instance, the UK White Paper (2003) uses key assumptions of four qualitative storylines about the future (with variable rates of GDP growth, population and household numbers) to derive an estimation of the potential gap in CO2 emissions that has to be

366

Chapter 6

bridged in 2050 by policy actions561. However, this approach does not allow scenarios to be compared on an equal basis. Furthermore, it can be argued that our approach is in line with a precautionary view on the problem at hand: the derived demand estimates for energy-intensive services (based on historical evidence) will most likely be at the higher end of other possible estimates, especially if the world economy is confronted with ever more stringent carbon constraints. Regarding the demand-side analysis, it is difficult to assess the scope for energy efficiency measures on a 50 year time scale on a similar basis as for energy generation options. This is linked to the number of technologies involved, new technologies and alternative interpretations of unrealised but apparently cost-effective potentials. We limited ourselves to the existing technical potential (taken from diverse sources). This technical potential has been considered to be equal for most technologies for all scenarios. However, the level of uptake of energy efficient technologies was varied across two broad scenario groups (cf. Section 3.2.4). 3.1.1.5 Supply-side analysis The supply-side analysis includes: • An analysis of state-of-the-art technologies for energy ‘production’. In our scenarios,

we only retained technologically proven options, rather than some options which are now only in a conceptual phase (e.g. fusion, some fission technologies); • An analysis of the technical, economic and environmental parameters of these technologies, and possible constraints to the development of the technology under consideration. Of course, there is a range of uncertainties attached to various costs. To accommodate for this inevitable uncertainty, we have drawn from a range of sources (cf. infra). In a first step, and as part of our research strategy to build the strongest possible case for different value positions, some worldviews (cf. Section 3.2.4) have built in ‘optimistic’ estimates for certain technologies – e.g. the lower cost estimates and less stringent constraints were used for distributed electricity generation options in the worldview called ‘Rational Perspective’. Next, a range of cost estimates can be used in detailed sensitivity analyses. Again, our approach is motivated by reasons of transparency: for instance, some models incorporate ‘learning curves’ which endogenise technological innovation. However, it seems that a ‘grand theory’ to give robust forecasts of technological development is not available. All one can do then is at least to strive for consistency and transparency.

561

World markets (a world based on individual consumerist values, a high degree of globalisation and scant regard for the environment); global sustainability (based on predominance of social and ecological values, strong collective environmental action and globalisation of governance systems); provincial enterprise (based on individual consumerist values, reinforced governance systems at national and sub-national level) and local stewardship (based on communitarian and strong conservation values, diverse political systems and economic regionalisation).


367

3.1.1.6 Policy measures to meet the strategic objectives The last building block entails the development of a coherent and internally consistent package of policy measures in order to reach the strategic objectives as defined in section 3.1.1.1. 3.1.2

Choice of energy model

MARKAL is a generic model that represents all energy demand and supply activities and technologies for a country with a horizon of up to 50 years. Here, it is used as a technicaleconomic bottom-up model which assembles in a simple but economically consistent way technological information (conversion-efficiency, investment and variable costs, etc.). As the model is formulated as a dynamic optimisation model, it can produce alternative developments for energy supply and demand options that satisfy a certain given level of demand for energy services, under certain constraints (for instance a CO2 emission reduction goal), at least cost. Simultaneously, the model makes prospective energy and emission balances, tests the potential of new energy technologies and can contribute to R&D policy formulation. One of the main advantages is that results are easily verifiable: they can immediately be related to assumptions regarding technological data and economic parameters. This approach is thus more suited to our purpose of ‘mapping’ different technological options (fuel choices by users, energy efficiency options adopted, energy supply options chosen, etc.). The main disadvantage is that economic processes in energy markets and the behaviour of economic actors are not modelled (demand for energy services is determined as an exogenous model variable). This can for instance be done with national general equilibrium models (e.g. GEM-E3 model of the EU): these are economic models which allow for instance to evaluate the macroeconomic impact of a CO2 tax. These models can study such questions as the use of the revenue from a CO2 tax, the double dividend discussion, the total impact on employment, etc. and deliver a basic forecast for the demand for energy services (an input for MARKAL). Table 12 summarises the strengths and weaknesses of different modelling approaches in the representation of energy use in a national economy.

368

Chapter 6

Driving forces for emissions Economic

Economic

Level of

Energy

Fuel choice

Final

Energy sector

Primary

activity

activity per

energy

efficiency

by user

energy use

transformation,

energy use

level

sector

services per

choices by

transport and

user

user

distribution

Reduction possibilities for GHG emissions (excluding carbon sinks) Lower level

Switch to

Less energy

More

Substitute

Reduce losses

of economic

less energy

intensive

energy

between

and use less

growth

intensive

production

efficient

coal, oil, gas

carbon-

activities

in industry

processes

and

intensive fuels

renewables

in electricity production

Lower

Better

indoor

insulation,

between

temp., less

more

coal, oil, gas

km. driven

Substitute

efficient

and

appliances

renewables

Modelling domain GEM-E3 model Yes

18 sectors

Implicit and joint

Simple

Simple

Simple

Simple

Modelling domain of MARKAL model Constant

Implicit

39

Detail

Categories

Table 12. Strengths and weaknesses of different modelling approaches to the representation of carbon emissions and energy use of a national economy (Source: Proost et al. 2000)

3.1.3

Limitations of the modelling approach

To summarise, with this scenario exercise, we can address the impact of different strategic objectives for the Belgian energy sector on the following issues: • • • •

The evolution of the demand for different energy vectors (electricity, fuels, gas, coal); The evolution of the demand for primary energy; The technological evolution of the electricity generation sector; A comparison between scenarios of the evolution of the production cost of electricity and the total energy system cost (this should however not be viewed as an absolute cost – to calculate this, econometric modelling approaches are more suited); • The contribution of different sectors to the CO2 emissions.


369

Subject to the following limitations, since some important aspects are not considered in this analysis: • The opening of the electricity market and the impact it can have on the production and the investment in the electricity sector is not modelled in detail – in fact we have either explicitly assumed (in 6 of the scenarios studied) that transmission constraints will not make intensive trade possible (hence, a negligible amount of electricity is imported), or either that Belgium will increasingly rely on net electricity imports in the future (in the remaining 2 scenario cases); • The possibility of reaching greenhouse gas reduction targets through international actions (tradable permits, CDM projects) is not considered. However, it can be argued that on the longer term, a 30% reduction target (as proposed here) by domestic measures is not exaggerated (cf. supra – further sensitivity analysis includes a 50% reduction target in 2050); • The overall effects of stringent carbon constraints on the Belgian economy (e.g. effects on economic activity of different sectors, employment effects, shifts in demand for certain energy services) cannot be studied – however, this limitation is rather linked to the inherent uncertainty of any econometric modelling on a (very) long time scale; • The MARKAL model is an energy sector model and is not appropriate to study the contribution of decentralised power to the overall reliability of power supply; a model specific for the electricity sector would be more appropriate to study these issues. This is particularly relevant, since generally, it can be said that the smaller the centrally controlled supply system is and the more the power generation is stochastic (e.g. wind and solar energy), the more difficult it gets to manage the power supply system and to maintain a satisfactory voltage level.

3.2

Background information on framing assumptions

Here, we give a brief introduction to the scenarios which will be developed in the following sections. A scenario consists of a particular combination of ‘framing assumptions’ and ‘building blocks’. Framing assumptions describe the external factors (the ‘environment’) influencing the Belgian energy system. In particular, we defined two worldviews – the ‘Rational Perspective’ (letter code R) and the ‘Market Drive’ worldview (letter code M) (cf. Section 3.2.4). Within each worldview, there is one ‘baseline’ scenario, which will be used for economic cost comparisons. The baseline scenario does not comply with (post-)Kyoto requirements and allows all energy production technologies. All other scenarios will have to comply with post-Kyoto commitments: -7.5% reduction of GHG emissions (compared to 1990 levels) in 2010, -15% in 2030, - 30% in 2050 (letter code K). Further sensitivity analysis with more stringent post-Kyoto commitments has also been conducted for this report: - 7.5% in 2010, -30% in 2030, -50% in 2050 (letter code KK).

370

Chapter 6

Building blocks describe the internal structure of the Belgian energy system (strategic objectives, energy saving potentials, supply options, energy policy). In particular (for this exercise), we focus on the technological supply options. We suggest four broadly defined options. For analytic simplicity, it is assumed that these major options will be pursued (almost) to the exclusion of the other options: 1. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity (letter code P.CS – Phase out & Carbon Storage); 2. phase out of existing nuclear power plants (40 yr. lifetime), new nuclear build allowed, limited possibilities for carbon capture and storage technology, limited import of electricity (letter code LCS – Low Carbon Storage); 3. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity (letter code P.LCS – Phase out & Low Carbon Storage); 4. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity (letter code P.LCS.I – Phase out, Low Carbon Storage & Import of electricity). In principle then, the combination of framing assumptions and technological options leads to 8 possible scenarios: M.K.P.CS, M.K.LCS, M.K.P.LCS, M.K.P.LCS.I (and a ‘Market Driven’ baseline scenario) and R.K.P.CS, R.K.LCS, R.K.P.LCS, R.K.P.LCS.I (and a ‘Rational perspective’ baseline scenario). 3.2.1

Demographics

Demographic developments influence the demand for energy services in a number of ways. Firstly, the number of households is one of the key determinants of residential energy consumption, since this parameter determines the number of household appliances in use and the specific surface that has to be heated. Secondly, the number of people influences the use of transportation services. The evolution of the population is taken from NIS-FPB (1996). In this report, Belgian population was 9.97 million in 1990, and was projected to reach 10.3 million in 2030. Actually, according to the latest data, Belgium already counted 10.3 million inhabitants in 2000, and is projected to count 10.9 million in 2050562. However, in view of the much larger uncertainties involved in long-term energy modelling, we have chosen to continue working with the original figures, rather than updating the entire existing database.

562

http://www.statbel.fgov.be/figures/d23_nl.asp#2


households population

1990 4.00 9.97

2000 4.27 10.21

2010 4.53 10.33

2020 4.61 10.34

2030 4.68 10.30

2040 4.68 10.30

371

2050 4.68 10.30

Table 13. Number of households and population (millions)

The number of persons per household is based on the reference scenario in NIS–FPB (1997). It goes from an average of 2.49 persons per household in 1990 to 2.28 in 2011. The trend was extrapolated until 2030 assuming the declining trend would continue, though at a slower pace (-0.02 persons per household every five years). For the period 2030-2050, we assumed that the trend would level off at 2.20 persons per household. These estimates still correspond well with the latest figures of an average of 2.35 persons per household in 2002563. 3.2.2

Assumptions for energy service demands

The demand of energy services differs from the final energy demand: the demand of energy services corresponds to the demand for heat in houses or heat with certain characteristics (temperature, pressure) industrial processes or the demand of vehicle-km. in case of transportation, whereas the final energy demand corresponds to the delivery of energy products to the consumers (Figure 7). Final energy is one of the inputs into the production of energy services, other inputs are e.g. heating equipment or house insulation. As explained in the methodological section, a macro-economic activity evolution model (GEM-E3) was used by the CES to determine the shift in the demand (curves) of energy services until 2030. In the industrial and service sectors, the demand function shifts at the same rate as the production or the value added of these sectors, taking into account the evolution of the relative energy service price and technical progress. For the households, the demand function shifts as a function of the evolution of income and relative energy prices, with an income elasticity of 0.3 for heating demand, 0.5 for hot water and cooking demand and 1 for specific electricity demand and a price elasticity of -0.3 for all categories of demand. For the transportation sector, passenger transport is a function of income whereas freight transport is a function of the general activity level, with a price elasticity of -0.3. The derived evolution of the demand for energy services used in both scenario groups is summarised in the table below. For the period 2030-2050, we made linear extrapolations of these projections, except for space heating (residential), warm water use (residential) and private car use, where we assumed that the demand would be saturated after 2030564. 563

http://www.statbel.fgov.be/figures/d24_nl.asp#3 For space heating, the unit consumption (i..e. final energy use per dwelling, with climate corrections) has remained almost constant over the period 1990-1998 (Odyssee Database, April 2000). Combining this with our demographic assumptions for the number of households (and hence, dwellings), the no-growth assumption after 2030 seems reasonable. For private car use, the no-growth assumption is based on the saturation effect of transportation infrastructure. The FPB (2001) assumes a maximum of 20,500 km/person.yr. In the existing MARKAL database, private car use already reaches the level of 21,200 km/person.yr. in 2030, so the nogrowth assumption for the period 2030-2050 seems justified.

564

372

Chapter 6

International energy prices

Other exogenous assumptions: - Overall technological progress - Overall growth expectations

GEM-E3 produces baseline levels of economic activity

Levels of energy services required in reference scenario

MARKAL

Energy efficiency use

Input MARKAL

Exogenous assumptions on technological development and cost of alternatives

output

Figure 7. Determination of the demand for energy services (Source: CES 2001)

Since the evolution of the long-term demand for energy services is inherently very uncertain, we have also included a sensitivity analysis with lower demand in industrial and transportation sectors (-30 % compared to the 2050 demand levels for all categories in both sectors in Table 14, with linear interpolation in the period 2000-2050). This could for instance correspond to a society with greater relative emphasis on services (compared to industrial production) and innovative solutions for the growth in demand for transportation (e.g. tele-working, carpooling, spatial planning, etc.). Information on demand for energy services is included in Annex 2.


2000-2010

2010-2030

2030-2050

Building materials (kton / yr.)

0.4

0.4

0.5

Glass (kton / yr.)

0.4

0.4

0.4

Limestone (kton / yr.)

0.4

0.4

0.4

Electrical motors (PJ)

0.8

0.6

0.6

Ammonia synth. (PJ)

0.5

0.5

0.4

Chemical industry

0.4

0.6

0.5

Chlorine (PJ)

0.5

0.6

0.6

Iron & steel prod. (kton / yr.)

0.2

0.1

0.0

Iron & steel proc. (kton / yr.)

0.2

0.1

0.0

Other industry (PJ)

0.9

0.9

0.9

Lighting (PJ)

1.0

0.8

0.8

Space heating (PJ)

0.9

0.9

0.8

Warm water (PJ)

0.9

1.0

0.8

Electricity use (PJ)

1.3

1.3

1.0

Space heating (PJ)

0.6

0.7

0.0

Warm water (PJ)

0.8

0.9

0.0


1.9

2.0

1.6

Private car use (billion veh. km.)

2.5

1.9

0.0

Rail (million veh. km.)

1.8

1.7

1.3

Freight (truck) (billion veh. km.)

1.6

1.4

1.1

373

Industrial sector

(steam + process heat) (PJ)

Service sector

Residential sector

Transportation sector

Table 14. Evolution of the demand for energy services – all scenarios (% average annual growth)

3.2.3

Resource availability and energy prices

The world’s oil, gas and coal resources are a key factor for future energy supply. Price levels are very uncertain and depend on a number of factors: cost of energy resource

374

Chapter 6

technologies, worldwide demand for fossil energy sources and strategic behaviour of exporting countries. As explained above, we assume in our scenarios that all fossil fuels will remain physically available over the next 50 years. For our assumptions regarding energy prices, we made use of two distinct (and rather extreme) prospects (Charpin et al. 2000)565. The average of these two prospects is taken as the standard value for energy prices in all scenarios (Table 15). Furthermore, a sensitivity analysis has been conducted by making use of the higher price estimates. All prices are net prices at the Belgian border, i.e. without additional taxes or duties. Price levels for 1990-2000 have been calibrated to the historic values. The price of gas is assumed to remain tied to oil product prices. The price of hard coal is supposed to remain relatively constant over the entire time horizon: coal resources are amply available and will certainly not be exhausted in the next century. Also, the uranium price is supposed to remain constant over the entire 50 year period. We have assumed a value of 3.3 MEuro/ton uranium, including the costs of mining the natural uranium and fabrication and delivery of fuel elements to power plants. Even assuming a rise in resource prices, the total nuclear fuel cycle costs would not rise significantly, since uranium is only a small fraction of the fuel cycle cost. In any case, this price effect will be very small compared to for instance uncertainties on the investment cost for new nuclear energy generation. The import price of electricity (again, net of taxes) corresponds to (in more familiar units) 4 Eurocent/kWh in 2000 and 10 Eurocent/kWh in 2050. No profit margin for the producer is included in this price. Cost (MEuro1995 / PJ)

1990

2000

2010

2030

2050

Crude oil High Sulphur Heavy Distillate Low Sulphur Heavy Distillate Gasoline Light Distilate Oil LPG Natural Gas Coal (non-residential use) Import of Electricity

3.12 2.21 2.53 4.88 4.21 3.07 2.48 1.34 10.33

2.42 1.72 1.96 3.59 3.20 2.76 1.69 1.16 11.81

3.11 2.52 2.88 4.86 4.55 3.07 2.47 1.19 14.76

4.75 3.86 4.40 7.28 6.92 3.93 3.41 1.24 20.66

6.40 5.19 5.92 9.69 9.28 4.78 4.35 1.28 26.56

Table 15. Resource prices for all scenarios

3.2.4

Discount rates

For the discount rates, we have distinguished two alternative ‘worldviews’ (inspired by the approach outlined in Lako et al. (1998)). Both worldviews should be considered to be 565

Resource prices for crude oil might seem to be rather low at first sight – e.g. crude oil is expected to cost about 32 $ per barrel in 2030 (in $ of 2000). However, according to the ‘International Energy Agency’ (IEA 2004), it is assumed that the oil prices will drop down to about 22 $ per barrel in 2006 (measured in $ of the year 2000). The price would then remain flat until about 2010 after which it would increase linearly till about 29 $ in 2030 (again in $ of 2000). This is only an assumption, but it shows that the energy analysts – ignoring temporary price fluctuations – expect that oil will remain quite affordable over the next 25 years.


375

‘technically’ possible – thus answering the demands of the ‘industrial commonwealth’ in Boltanski and Thévenot’s words. The ‘Rational Perspective’ worldview can be characterised as being more driven by ‘civic’ concerns; while in the ‘Market Driven’ perspective, these concerns take a backseat towards those of the ‘market commonwealth’. It is assumed that the process of global economic integration will lead to more collective public action. Strong penetration of new, more efficient demand and supply technologies (also at the end-use level) is facilitated. This strong penetration can for instance be achieved by setting efficiency standards, removing existing barriers for the introduction of efficient and decentralised technologies, and active energy service companies which carry out cost-effective efficiency improvements for third parties. The two distinctive storylines can be described as follows: In the ‘Market Drive’ worldview, the market mechanism is seen as the best way to generate prosperity and handle uncertainty. The penetration of new, more efficient demand and supply technologies is seen to depend more on market forces and the behaviour of actors. The environmental protection agenda is also more influenced by market actors than by public policy. Moreover, energy policy is driven by the desire to minimise government control and to maximise efficient operation of free markets. Governments will only intervene in issues of overriding importance: e.g. securing diversity of energy supplies or enforcing internationally binding agreements (e.g. under the UNFCCC). Barriers will persist in the uptake of efficient equipment. Efficiency gains will only be made for competitive reasons. The ‘Market Drive’ worldview assumes that in reality more stringent investment criteria apply for many energy related decisions and that hidden costs and market barriers do play a role. This scenario group assumes different discount rates per type of sector. The discount rates reflect representative hurdle rates applicable to that sector or kind of end-use (cf. Table 16). In the ‘Rational Perspective’ worldview, a single discount rate is applied. By applying a uniform rate, all technologies at the demand side of the energy system are allowed to compete with energy supply technologies like in a perfect market. RP assumes a market that works ‘rationally’ without barriers and with perfect information so that any difference in pay back opportunities will automatically be removed. The discount rate in RP amounts to 5% per year, which is typical of a low-risk investment climate. The above-mentioned general perspectives have been modelled (very schematically) by varying the discount rate across the scenario groups. A discount rate is required to annualise the capital cost in order to compare the costs of alternative technologies with different ratios of initial capital expenditure to annual running costs. The formulation of the MARKAL model allows choosing one uniform discount rate applied to all technologies or different discount rates applied to the different sectors.

376

Chapter 6 Sector

Discount rate

Power generation

8%

Industrial cogeneration, refineries, biomass conversion, all processes in industry

10%

Space and water heating in residential sector and commercial sector

15%

Trucks, vans and buses Electric appliances

20%

Passenger cars

25%

Table 16. Sector specific discount rates in ‚Market Drive’ scenario group

In addition to the differences between the worldviews relating to their perspective towards decision criteria on energy investments by individual actors (represented by different discount rates (Lako et al. 1998)), assumptions regarding the available technical potential of energy saving technologies and regarding cost and constraints on technologies for electricity generation were varied in line with the overall worldview under consideration (cf. Table 17) . Rational Perspective (RP)

Market Drive (MD)

Decision criteria

Uniform 5% discount rate for all energy decisions across all sectors

8 % discount rate for power generation, higher discount rates for end use

Technical potential energy saving

In some sectors higher than in MD

In some sectors lower than in RP

Technologies for electricity generation

Favourable towards decentralised electricity generation options

Favourable towards centralised electricity generation options

Table 17. Key differences between ‚Rational Perspective’ and ‚Market Drive’ scenario


3.3

3.3.1

377

Using the ‘building blocks’ and ‘framing assumptions’ to construct scenarios

Strategic goals

Overriding strategic goals (which apply to all scenarios) are: Security of energy supply • On a world-wide scale, it is assumed that sudden shortages of energy supply (energy crises) can be avoided – we do however take into account the possible consequences of a gradual depletion of fossil energy resources on energy prices; • On the level of the individual consumer (in Belgium), it is assumed that energy supply will be reliably, flexibly and qualitatively tailored to the demand for energy services as required by (at least) present-day demands for comfort (however, this does not imply that the same level of end-use energy demand will be sustained !); • On the level of the individual consumer (in Belgium), it is assumed that social energy provisions (social tariffs, minimal provision of energy, etc.) can be maintained. Economic criteria • It is assumed that worldwide ratio of economically proven (fossil or uranium) reserves (R [PJ]) over the production of primary energy carriers based on these reserves (P [PJ/a]) will allow for enough time (e.g. 20 years?) for a transition to an energy system based on virtually inexhaustible energy resources (e.g. wind, sun, fusion) when needed (beyond the time horizon under consideration); • The intensity of energy use (energy use per unit of GDP) should not increase in any scenario. Ecological criteria • SO2, NOx and NMVOC emission levels will have to be limited (the Göteborg Protocol emission reduction targets can be used as a guideline) ; • Greenhouse gas emission levels will be significantly reduced through follow-ups to the Kyoto Protocol. All ‘sustainable’ scenarios will have to comply with reduction goals of 7.5% in 2010 (Kyoto requirements), 15% in 2030 and 30% in 2050. These requirements are not very ambitious compared to long-term visions developed for other countries: the UK White Paper (2003) sets out a 60% reduction goal based on the principle of ‘contraction and convergence’ (in line with a potential global agreement which would set an upper limit of 550 ppm for the carbon dioxide concentration), while the Wuppertal Institute (Fischedick et al. 2002) even aims at an 80% emission reduction goal for Germany, both on a time scale of 50 years. However, we did investigate further reduction targets (-30% in 2030, -50% in 2050) in our sensitivity analyses. Furthermore, we only study energy-related emissions. These however account for 87% of the total Belgian GHG emission level in 1990 (Proost et al. 2000).

378

3.3.2

Chapter 6

Demand side analysis

3.3.2.1 Industry As explained above, the demand for energy services was kept identical for both scenario groups. However, the technical potential for energy savings in different sectors was sometimes varied in both scenario groups. We have drawn from a range of sources, including the original MARKAL database (described in CES (2001)), WEC (2000), and Fischedick et al. (2002)). We have chosen to limit ourselves to the proven existing technological potential, rather than simply assuming the continuation of certain trends in energy productivity (energy end use/energy service) over the entire time horizon (as is the case in the UK White Paper for instance). Technical potential

Rational Perspective

Market Drive

Cement

20%

20%

Construction materials

20%

20%

Glass flat

15%

15%

Glass hollow

37.5%

37.5%

Glass fibres

27.5%

27.5%

Lime

10%

10%

Industry - power

35%

20%

Industry electro thermal

20%

20%

Industry lighting

50%

20%

0-25%

0-25%

Steam (chemical industry)

20%

20%

Process heat (chemical industry)

10%

10%

Chlorine electrolysis

10%

10%

-

-

Iron & steel processing

20%

10%

Low temp. steam & heat

10%

10%

10%

10%

energy savings

Ammonia synthesis (gradual increase till 2050)

Steel production

(other ind.) Other industry

Table 18. Technical energy saving potential (industry)


379

In MARKAL, for each sector at least one conventional end use technology is modelled which represents the present mix of installations. In addition, one or more alternative processes or saving options are provided. These saving options can occur in addition to a change in end-use technology and thus contribute to the overall efficiency of the process. Only some of these saving options were reviewed for our exercise. 3.3.2.2 Residential and service sector For electricity saving potentials in the residential and service sector, we employed the following figures (cf. Table 19). For the heating demand and the potential for energy savings in the residential sector, a detailed bottom-up modelling philosophy is applied, starting from the basic unit: the number of houses. The house stock is characterised by the following variables: age, type (open, half-open, closed and flats), heating system (centralised or local) and insulation level. Considering the evolution of the population, the size of the households and the number of demolitions, the evolution of the number of houses in each category can be calculated. The demand for heating is then computed per type of house, and then corrected for losses and average temperature. Total heating demand takes into consideration the number of houses, the heating demand per house, and the effects of rising income on heat demand (for more details, see CES-VITO 2001). Technical potential


Market Drive

Residential electricity use

0-50%

0-10%

Small service sector elec. use

0-50%

0-10%

Large service sector elec. use

0-50%

0-10%

energy savings (gradual till 2050)

Table 19. Technical energy saving potential (residential and service sector)

The availability of insulation measures and their cost depends on the construction year of the building (3 insulation levels are considered for existing buildings). In both scenario groups, we have assumed that the K55 insulation level for new buildings applies in the residential sector since 1990. Furthermore, for new buildings, we have defined an additional new building standard with very low heating demand (the so-called ‘passivhaus’ (Fischedick et al. 2002), with specific heating demand of some 20 kWh/m2.a)566. Comparing this standard to the average heating requirement of a K55-open building of some 135 kWh/m2 shows this is very ambitious indeed! Both scenario groups then have built in different assumptions about the speed of retrofitting existing houses to the K55

566

The name derives from the fact that in this type of house the same level of heating comfort can be reached almost without making use of an ‘active’ heating unit. In a sense, our scenario assumptions are unrealistic (or rather optimistic), since we do not consider more detailed intermediary insulation standards over time (such as the K40 standard). In view of the goals of the scenario exercise, this is considered to be acceptable.

380

Chapter 6

isolation standard, and about the percentage of new houses built according to the ‘passivhaus’ standard567. 3.3.2.3 Transportation For the transportation sector, it is recognised that there are technological developments ongoing which are significantly improving engine and fuel efficiency, primarily through the introduction of a wide range of fuel saving technologies already developed (e.g. direct injection engines, variable transmission systems, and lightweight materials). There is also a range of emerging vehicle technologies which can further improve the energy efficiency of vehicles over the longer term (such as hybrid electric and fuel cell technology). Likewise, other measures such as reducing traffic congestion or the encouragement of modal shifts away from the car to more sustainable alternatives could have a large impact on the CO2 reduction potential in the transportation sector. No robust estimates of the long term potential for modal shift and their subsequent impact on carbon emissions are available, and given the wide range of potential measures in the transportation sector which could reduce carbon emissions, it is not possible at present to provide a comprehensive assessment of different long term options for this report. Therefore, we have only included technological options for CO2 reduction in the transportation sector. Vehicles are supposed to become much more fuel efficient, up to a factor 4 in 2050. 3.3.3

Supply side analysis

3.3.3.1 Brief description of key technological options The MARKAL database contains a great number of different technologies for power generation. Power generation options which are expected to play a key role in the future energy system are described in the following sections. Figures are taken from the original MARKAL database (described in CES-VITO 2001), the AMPERE report (Section I on the evaluation of the current and future costs of electricity production and external costs), the report by Lako et al. (1998) and the WEC (2000), and are given in Annex 3. The residual capacities of existing power plants were adapted to those proposed in the electricity sector equipment plan 1995-2005. Nuclear For existing nuclear power plants the average availability factor and the technical lifetime are very important parameters. The availability factor used in MARKAL is 85%, the Belgian historical average for this type of plants. It is assumed that the decided investment on existing nuclear plants will allow maintaining this availability factor in the future. The operation time is set equal to 40 years, in accordance with the Belgian phaseout law. In the scenarios with letter code P, no new investments in nuclear power are

567

The ‘Rational Perspective’ worldview assumes 2% of the existing housing stock will be retrofitted per year starting from the year 2000, resulting in a complete renovation in 2050. In the ‘Market Drive’ worldview, investment in isolation measures is solely driven by cost considerations.


381

allowed. However, in other scenarios we have left open the possibility of new investments. For this report, we have considered the use of two new reactor types: the EPR (European Pressurized Water Reactor) – an evolutionary design, or the MHTGR (Modular HighTemperature Gas-cooled Reactor) – a more innovative concept, described in more detail below. Future costs for these reactor types remain uncertain until one has actually been built and operated. High-temperature gas-cooled reactors (HTGRs) typically involve large numbers of uranium fuel pellets encased in layers of carbon, silica, or both (designed to contain the fission products from the reaction). Modern HTGRs are designed to be passively safe, offering the potential to avoid many of the complex, expensive safety systems used in LWRs. It is hoped that such features could lead to lower costs and improved safety. The key to enhanced safety for the so-called ‘pebble-bed modular reactor’ (PBMR) and the ‘gas turbine-modular helium reactor’ (GT-MHR) is a design that ensures that the highest temperature in the reactor core—under any conceivable operating or accident condition— never exceeds the operating limit of the fuel. This requirement limits the thermal output for a single module to 250 MWth and the electrical output to 100 MWe—a factor of 10 smaller than for a typical LWR – hence the modular characteristic of this technology. The spent fuel of the PBMR would be high-burn-up material in many tiny spheres, making it a comparatively unattractive source from which to recover weapons-usable material (WEC 2000). In our choice of nuclear technologies to be included in the MARKAL database, we only considered these two (in a sense) ‘extreme’ cases: very centralised (the EPR would have a typical installed capacity of 1400 MWe per reactor) vs. more decentralised (the MHGTR would have an installed capacity per module of 100 MWe). This does not imply that these reactor types represent the only nuclear technology options available for the future568. Fossil Fuel Based Power Plants Two types of coal power plants are considered in the model: an ultra super-critical coal power plant (USC) and integrated coal gasification combined cycle (IGCC). These types of power plants can become the main base load technology in the longer run when natural gas becomes more expensive. IGCC is a technology which is still at the demonstration level: important technical progress has been foreseen in the energy efficiency (from 42% in 1995 to 51% in 2030). Both technologies are, at least at the end of the horizon, in close competition with the cost and efficiency figures assumed here. Gas-fired plants are either gas turbines or steam and gas power plants (STAG). In addition, the option of CO2 capture and storage (CCS) is included in the database. Carbon dioxide capture and storage is essentially a process whereby CO2 is removed from the fossil fuel used to generate electricity (either pre- or post-combustion) and stored in

568

An ‘intermediate’ size new nuclear reactor type (600 MWe) would for instance be the advanced light water reactor developed by Westinghouse, known as the AP-600.

382

Chapter 6

natural underground reservoirs, preventing it reaching the atmosphere. It can achieve an 80% reduction in CO2 emissions to the atmosphere. Carbon dioxide storage will only be an effective way of avoiding climate change if the CO2 can be stored for several hundreds or thousands of years. The most promising storage options are depleted and producing oil and gas reservoirs (where CO2 injection for enhanced oil recovery is possible), deep saline reservoirs, unminable coal beds or aquifers. However the legal status and public or political acceptance of disposal in sub-sea strata is questionable. Carbon capture and storage technologies (CCS) are best suited to large-scale sources of CO2 such as power stations (coal or gas). However, the large capital costs involved in adapting existing generation plants to CCS and the resulting loss in efficiency, mean the technology is best applied to new plants, where it can be incorporated into initial construction. In addition, if hydrogen became established as a major fuel for cars, electricity production and heat and power generation; centralised, large-scale production of hydrogen from fossil fuels would be possible from precombustion capture of CO2 emissions. All the individual components of the technology exist and are commercially proven and could be deployed. However, there appears to have been no systematic probabilistic analysis of risks and environmental consequences, or systematic assessment of the available data on slow release (UK Energy White Paper 2003). In view of the uncertainties involved with this technological option, some scenarios will have constraints imposed on the amount of CO2 that can be stored in deep geological layers (in the scenarios with letter code LCS, a limit of 5 Mton CO2 /a has been applied). Cogeneration technologies In the industrial sector, gas turbines of 4MWe and of 35MWe and STAG of 30MWe are considered for high temperature steam and a backpressure turbine (20MWe) for low temperature steam. STAG units of 30 and 100 MWe and gas turbines and diesel engines of 30MWe for cogeneration in the residential and tertiary sector are modelled. A number of gas turbines on biomass, as well as fuel cells for cogeneration have been included in the model (fuel cells for residential use were added). Fuel cells either use hydrogen or employ a gas reformer. Small gas engines of 1MWe for decentralised cogeneration in this sector are also considered. Limits have been imposed on the penetration of cogeneration in the different sectors (see under constraints). Renewables Conventional renewables like hydro power, a large range of biomass-fuelled power plants and wind turbines (both onshore and offshore) are addressed in the database. Photovoltaic electricity generation has been added as an option for the long-term. The long-term potential of both wind energy (offshore) and photovoltaics is subject to uncertainty and thus constitutes a variable in different scenario groups. In scenarios with both a nuclear phase out and low potential for CCS (letter code P.LCS), cogeneration and renewables will play a comparatively more important role in energy provision.


383

Import of electricity Some scenarios (with the letter code I) rely on a significant import of foreign electricity to meet energy demands. In these scenarios, electricity imports of approximately one third of the total demand for electricity have been allowed in 2050 (linear interpolation from 1990 levels). To determine the impacts of this import of electricity, it is of course necessary to know the electricity generation mix of the imported electricity mix. We have taken as a reference the global energy perspectives developed by Nakicenovic et al. (1998). For the ‘Market Drive’ worldview, we have chosen the electricity generation mix of scenario A3 for Western Europe569. This mix amounts to (in 2020/2050): 20/35% gas fuelled power generation, 25/25% renewables (wind, solar, hydro, biomass and waste) and 25/40% nuclear. For the ‘Rational Perspective’ worldview, we have chosen the electricity generation mix of scenario C1570. This mix amounts to: 20/40% gas fuelled power generation, 25/45% renewables (wind, solar, hydro, biomass and waste) and 25/15% nuclear. 3.3.3.2 Constraints Some of the technologies incorporated in the MARKAL model are well-known, e.g. proven coal and gas-fired power technologies. A lot of technologies are in the stage of development, demonstration or early implementation. A first criterion to select technologies for uptake in the database is technical feasibility. Based on this criterion, more speculative options such as fusion power or a nuclear fuel cycle with advanced partitioning and transmutation can be excluded (at least during the time horizon under consideration). Constraints to the application of certain technologies can result from a number of factors. For instance, technical-economic constraints apply to certain renewable energy options. Wind energy has been split up in several categories (onshore at seaside, onshore polders, onshore inland, offshore) depending on wind regime and availability of space. Each category has an upper bound in a specific period. Photovoltaics suffer from an upper bound on capacity due to limiting factors such as availability of suitable surfaces and solar irradiation. However, as these bounds are rather uncertain on the long term, different assumptions have been used in both scenario groups (‘optimistic’ in scenario group ‘Rational Perspective’, ‘pessimistic’ in scenario group ‘Market Drive’). In the RP worldview, a higher potential has been assumed for offshore wind energy, but the

569

This scenario is described by the authors as “...an illustration of of a case where a ‘rich and clean’ energy future resolves some of the challenges of global warming without recourse to stringent environmental policy measures…” (p. 73). 570 This scenario is described by the authors as “...a challenging pathway of transition away from the current dominance of fossil sources to a dominance of renewable energy flows. Ambitious policy measures accelerate energy efficiency improvements and develop and promote environmentally benign, decentralised energy technologies…” (p. 74).

384

Chapter 6

additional 1 GWe (compared to the MD worldview) can only be realised at a 30% increase in investment costs571. For nuclear power generation, we assumed that only existing sites can be used for possible new reactor units. This limits the maximum installed nuclear capacity to 8 GWe (in all scenarios). If new nuclear power plants are allowed in the ‘Rational Perspective’ worldview, we assumed that new reactors would be of the MHTGR type (in line with the promotion of smaller-scale technologies). Of course for nuclear energy, public acceptance of the technology, and possible long lead times, high capital costs, waste management issues, uncertainty about back-end costs and, at present, lack of political readiness (in most OECD countries) to promote nuclear as an option, are the main barriers to new nuclear construction. All of these issues might lead to a political or a ‘de facto’ nuclear phase out (letter code P). Public acceptance might however also be an issue for other technologies, such as wind power (visual intrusion). This could well lead to some overestimation of the potential for wind power. However, one might argue that the resulting smaller potential for onshore wind energy could then be offset by a larger application of offshore technology. The only difference then is the total economic cost. Carbon capture and storage technologies might also suffer from a lack of public or political acceptance, or are even still subject to technological uncertainties. We considered two cases: either a constraint (of 5 Mton, letter code LCS) to the amount of CO2 which can be stored in an acceptable way has been applied, or the CCS potential was considered to be sufficiently large for (no constraint). Worldview RP

2000

2030

2050

Wind (offshore) (GWe)

1.0

1.6

2.0

Wind (onshore) (GWe)

0.9

1.2

1.4

Photovoltaics (TWh/yr)

-

10

10

1.0

2.7

3.9

0.2

0.5

0.6

0.4

1.2

1.6

0

8

8

Cogeneration (industry) – HT steam (GWe) Cogeneration (industry) – LT steam (GWe) Cogeneration (service + res) (GWe) MHTGR nuclear reactor (GWe)

571

A recent in-depth study (published after we developed our scenarios) has calculated an offshore economic potential of 2.1-4.2 GWe installed capacity (Van Hulle et al. 2004). Thus our ‘optimistic’ assumption might turn out to be too ‘pessimistic’ after all.

Criteria and scenarios as a support for sustainable energy governance Worldview MD

2000

2030

2050


1.0

1.0

1.0


0.75

0.75

0.75


-

10

10

1.0

1.8

2.0

0.2

0.3

0.3

0.4

0.8

0.8

0

8

8

Cogeneration (industry) –

385

HT steam (GWe) Cogeneration (industry) – LT steam (GWe) Cogeneration (service + res) (GWe) EPR nuclear reactor (Gwe)

Table 20. Constraints on technological options in both worldviews

Table 20 summarises the default constraints adopted in both worldviews. For cogeneration technologies, maximum capacities are taken from the AMPERE report (2000). However, as these capacities are largely uncertain on a long term, the possibility of a doubling of maximum capacity has been allowed under the ‘Rational Perspective’ worldview. 3.3.4

Policy measures

This kind of analysis does not deal explicitly with the policy measures needed to attain the strategic objectives. For instance, (post-)Kyoto targets can simply be ‘imposed’, without however considering the exact mix of policy measures needed to arrive at these targets. It is simply assumed that policy measures taken will be in line with the defined objectives and the major technological options chosen within different scenarios (e.g. policy measures to promote cogeneration in scenarios with nuclear phase out, continuation of nuclear related R&D in scenarios without a phase out, R&D for options with carbon capture & storage, etc.). 3.3.5

Visions beyond 2050

As already stated, none of the scenarios can claim to provide a definitive sustainable solution to energy provision in 2050. For fossil-fuel based power generation, carbon capture and storage can only be used as a transitional solution towards more sustainable energy supply options, as fossil fuels will inevitably run out (however, coal could still provide a resource base for some centuries to come). Proponents of the nuclear option propose (as an option for a very far future, e.g. 20402060) to recycle the majority of the fission products with long half-lives (mainly 99Tc and 129 I) and of the actinides (mainly Pu and Am, elements with long half-lives) to fission in fast-spectrum reactors. The remaining waste would then be mainly fission products with half-lives of decades or shorter, and after 300 years the total fission-product inventory would decline to a radio toxicity level lower than that of the original ore. Of course, it is very uncertain whether (and when) this option will ever become technically and

386

Chapter 6

economically feasible. Also, innovative fuel cycles based on thorium (largely available) are investigated in response to possible uranium resource limitations. Furthermore, fusion energy might become available after 2050. One could also envisage an energy system largely based on renewable energy sources (combined with a reduction in the demand for energy services) – see e.g. the report by the LTI research group (LTI 1998). Optimal integration of a large share of different renewable energy production options in the energy system is however only conceivable on a European scale.

3.4

Results

This chapter gives results for the 8 scenarios developed for this exercise. In section 3.4.1, key findings will be discussed for each option (under the ‘Market Drive’ and ‘Rational Perspective’ worldview). Then, some striking differences between the scenarios will be illustrated (Section 3.4.2). Finally, complete results for quantifiable parameters for all scenarios will be discussed (Section 3.4.3). The assumptions behind these scenarios have been described in section 3.2, but we recapitulate here for practical purposes: • M.K.P.CS – Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity; • M.K.LCS – Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), new nuclear build, limited possibilities for carbon capture and storage technology, limited import of electricity; • M.K.P.LCS: Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity; • M.K.P.LCS.I: Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity; • R.K.P.CS – Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity; • R.K.LCS – Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), new nuclear build, limited possibilities for carbon capture and storage technology, limited import of electricity;


387

• R.K.P.LCS: Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity; • R.K.P.LCS.I: Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity.

In addition, two baseline scenarios (for the ‘Market Driven’ en ‘Rational Perspective’ worldview) have been calculated for cost comparison purposes. 3.4.1

Selected results per option

3.4.1.1 Nuclear phase out (scenario R.K.P.CS and M.K.P.CS) Primary energy use The total primary energy demand increases slightly for the M.K.P.CS scenario, while – after an initial increase - remaining almost equal to the present primary energy demand (some 2300 PJ) in the R.K.P.CS scenario (notwithstanding economic growth). Several changes in the fuel mix occur over time. Until the year 2020, oil consumption remains rather constant (compared to 1990 levels), gas consumption rises rapidly (more than a doubling!), and coal consumption decreases. Trends for oil, gas and coal are more or less the same for both worldviews (although the absolute rise in gas consumption is less pronounced in R.K.P.CS).

Primary energy use 3000

2500

PJ

2000

Renewable Energy Gaseous Liquid Solid Nuclear Energy

1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050

388

Chapter 6


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 8. Primary energy demand – scenarios M.K.P.CS (top) and R.K.P.CS (bottom)

After 2020, appreciable changes occur. Nuclear energy is phased out, and coal use rises sharply. Gas use remains high (more or less levelling at 2020 demand) for the M.K.P.CS scenario, while in the R.K.P.CS scenario it decreases to some 700 PJ in 2050. Oil use decreases (due to the introduction of fuel efficient vehicles). Renewables gain a modest share of some 5-6% in 2050572. Electricity production The trends for electricity production are more pronounced than those of primary energy demand. In the M.K.P.CS scenario, coal shows a rather steep decline until 2015. After that, coal replaces the nuclear generation capacity and becomes the dominant fuel choice for electricity generation in 2050. Natural gas shows a steady increase until 2035 and declines a bit afterwards. The coal gasification combined cycle (IGCC) and steam and gas combined cycle (STAG), both equipped with carbon capture technology, become the dominant electricity generation options from 2025 onwards. Electricity demand is very high. This is logical: since electricity generation in this scenario is essentially carbon-free, switching to electricity as energy carrier presents a robust and economically attractive (relative to other options) solution to meeting CO2 reduction targets.

572

Primary energy use is calculated based on the net calorific values of fuels. For electricity from renewable sources, we applied the “substitution equivalence” method, by which primary energy equivalents are calculated assuming a conversion efficiency of 38.6 %. For nuclear electricity, we used a value of 33%.


389

Electricity production 180

160

140

120 Renewables

100 TWh

Cogeneration Gas Coal

80

Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 9. Electricity production – scenarios M.K.P.CS (top) and R.K.P.CS (bottom)

In scenario R.K.P.CS, coal becomes even more important (in relative terms) than in M.K.P.CS, and completely replaces gas for electricity production (the remaining cogeneration capacity is still gas-fuelled however): since the discount rate is lower in the ‘Rational Perspective’ scenario, investment in power plants will take into account price effects on a longer timescale (and gas prices show a steady increase compared to coal prices). Gas is reserved for other uses (e.g. process heat for industry) where no electric alternative exists. Overall electricity demand is lower as a result of more energy-saving measures. Cogeneration (mainly for high-temperature steam in industry) gains a relatively

390

Chapter 6

large market share in the period 2010-2035 before declining. Renewables are more important in the R.K.P.CS case (gaining a share of 13% of total electricity production in 2050). CO2 emissions CO2 emissions in both scenarios are somewhat similar. A decrease is most noticeable in the residential & service sector, although the mechanism is somewhat different in both scenarios: in the M.K.P.CS scenario, electric heating appliances (various heat pumps and some electric accumulation heating) gain a large market share, while in the R.K.P.CS case, lower CO2 emissions result mostly from energy saving measures (e.g. better isolation). CO2 emissions [Mton]

2000

2010

2020

2030

2040

2050

Electricity

20

15

15

16

18

19

Industry

34

25

26

24

23

23

Resid. & service

34

28

25

23

21

15

Transportation

26

30

28

28

21

19

Other

4

4

3

3

2

1

Total

118

102

97

94

85

77

2020

2030

2040

2050

CO2 emissions [Mton]

2000

2010

Electricity

19

13

15

20

17

18

Industry

34

27

27

28

29

26

Resid. & service

32

26

22

19

17

13

Transportation

26

30

29

25

21

19

Other

4

4

3

0

0

0

Total

115

100

96

92

84

76

Table 21. CO2 emissions – all sectors (scenario M.K.P.CS (top) and R.K.P.CS (bottom))

Electricity consumption in the residential & service sector amounts to 92 TWh in M.K.P.CS, compared to 47 TWh in R.K.P.CS. Also, the R.K.P.CS case relies heavily on centralised district heating (mainly for urban dwellings): 93 PJ is provided by this technology in 2050 (vs. 31 PJ for the M.K.P.CS scenario). In both scenarios, some of the heating demand in flats is met by combined heat-and-power (CHP): 23 PJ in M.K.P.CS vs. 28 PJ in R.K.P.CS. CO2 emissions in the electricity sector remain relatively low despite the nuclear phase out as a result of the availability of carbon capture and storage technology.


391

3.4.1.2 Low potential carbon capture and storage (scenario M.K.LCS and R.K.LCS) Primary energy use


2500

PJ

2000 Renewable Energy Gaseous Liquid Solid Nuclear Energy

1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


2500

PJ


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 10. Primary energy demand – scenario M.K.LCS (top) and R.K.LCS (bottom)

Again, the total primary energy demand increases some 20% over the entire time horizon for the M.K.LCS scenario, while the increase is less marked for the R.K.LCS scenario. In

392

Chapter 6

both cases, nuclear power plants are constructed up to the full capacity of 8 GWe starting from 2020 (after the existing nuclear power plants are retired). Coal only plays a minor role, while the economy becomes increasingly dependent on imported gas (less dependence in absolute terms in the R.K.LCS scenario: 750 PJ vs. 1075 PJ in 2050 for the M.K.LCS scenario). Renewables again play only a relatively small part: 6% in both cases. In these scenarios there is a much greater emphasis on gas than in the scenarios of the previous section (M.K.P.CS and R.K.P.CS) because coal is basically forbidden by post-Kyoto GHG emission limits and the non-existence of carbon storage, while at the same time nuclear is limited to 8 GWe. In the previous section coal with CS was allowed to expand unconstrained. Electricity production The trends for electricity production are shown in Figure 11. Some trends can be noted. In the M.K.LCS scenario, gas becomes the second choice fuel after the nuclear electricity generation capacity has been fully developed. From 2040 onwards, fuel cells become the preferred option for industrial cogeneration (generating some 13 TWh in 2050). The hydrogen for these fuel cells is produced starting from natural gas (in a steam reforming process); it is assumed that the CO2 generated during the process is captured and stored onsite, while the hydrogen is transported (e.g. by pipeline) to the power plant573. Renewables (mostly wind power and photovoltaics) represent a modest share of electricity generation: some 12% in 2050.


140

120

100 Renewables

TWh

Cogeneration

80

Gas Coal Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

573

For steam reforming, a production cost of 5.92 MEuro per PJ natural gas is assumed (in addition to the natural gas price of 1.69 – 4.35 MEuro/PJ). No detailed cost calculations for transport, storage and distribution of hydrogen have been included. This of course implies that the costs of the hydrogen option are underestimated.


393


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 11. Electricity production – scenario M.K.LCS (top) and R.K.LCS (bottom)

In the R.K.LCS scenario, the use of gas for electricity generation is restricted to cogeneration plants from 2035 onwards. Cogeneration is mainly used to produce hightemperature steam for industrial uses, but some combined heat-and-power is also installed in flats and large service sector buildings (providing 4.4 TWh of electricity in 2050). In this scenario, almost no use has to be made of the (relatively expensive) hydrogen to reach the post-Kyoto requirements. Renewables steadily increase over time, providing some 15% of electricity in 2050. Photovoltaics are not used up to their full potential. The existing nuclear production park is replaced by the more small-scale (100-300 MWe per unit) modular high-temperature gas-cooled reactor starting from 2020, up to the full potential of 8 GWe. CO2 emissions CO2 emissions [Mton]

2000

2010

2020

2030

2040

2050

Electricity

20

14

15

11

12

15

Industry

34

25

26

25

24

24

Resid. & service

34

29

26

27

25

17

Transportation

26

30

27

28

21

19

Other

4

4

3

3

3

3

Total

118

102

97

94

85

78

394

Chapter 6


2000

2010

2020

2030

2040

2050

Electricity

19

13

13

11

8

9

Industry

34

27

28

29

29

30

Resid. & service

32

26

22

24

23

15

Transportation

26

30

30

25

21

19

Other

4

4

3

3

3

3

Total

115

100

96

92

84

76

Table 22. CO2 emissions – all sectors (scenario M.K.LCS (top) and R.K.LCS (bottom))

Similar observations as in the previous case can be made. A decrease is most noticeable in the residential & service sector. More electricity is used in the M.K.LCS scenario in these sectors (although the effect is now less pronounced: 78 TWh vs. 49 TWh for the R.K.LCS scenario). Heating is provided for with a mixture of heating pumps and gas boilers in open and half open residences. Again, in the R.K.LCS scenario central district heating is used to meet some of the heating demand in urban dwellings (45 PJ in 2050), while this option does not become available in the M.K.LCS scenario. Decentral combined heat-and-power provides for another 21 PJ of heating demand in the M.K.LCS case, compared to 29 PJ in the R.K.LCS case. As a result of the lower electricity demand in the R.K.LCS scenario, CO2 emissions in the electricity sector remain very low, while an increasing reliance on gas in the M.K.LCS scenario results in higher CO2 emissions starting from 2030. 3.4.1.3 Phase out and low potential carbon capture and storage (scenario M.K.P.LCS and R.K.P.LCS) Primary energy use For the first time, an absolute decoupling between economic growth and primary energy use can be noticed for the R.K.P.LCS scenario, while in the M.K.P.LCS case, primary energy use more or less returns to the 1990 demand level. Both scenarios show somewhat similar results. Most noticeable is the increasing dependence on imported gas: in the M.K.P.LCS scenario, this dependence amounts to 1110 PJ in 2050, while in the R.K.P.LCS scenario, it is limited to 1034 PJ (due to the overall lower primary energy demand). Also, renewable energy sources gain a relatively large importance: 9% of primary energy demand in the R.K.P.LCS scenario and even 12% in the M.K.P.LCS scenario (this is because in the period 2040-2050 solar boilers are installed in order to meet CO2 reduction targets).


395


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 12. Primary energy demand – scenario M.K.P.LCS (top) and R.K.P.LCS (bottom)

Electricity production The results are more marked for electricity production (cf. Figure 13). In the M.K.P.LCS scenario, the existing nuclear power plants (phased out in 2015-2025) are almost exclusively replaced by STAG power plants (some equipped with carbon capture and storage technology, up to the maximum potential of 5 Mton CO2 stored per year). Renewables (mainly wind energy and photovoltaics) are developed up to their maximum potential, contributing some 14% of electricity demand in 2050. Cogeneration is, starting

396

Chapter 6

from 2025, largely based on fuel cell technology (to generate high pressure steam for industrial demand), contributing some 14% of electricity demand in 2050. Again, the hydrogen for these fuel cells is produced starting from natural gas (in a steam reforming process); it is assumed that the CO2 generated during the process is captured and stored onsite, while the hydrogen is transported (e.g. by pipeline) to the power plant. In the R.K.P.LCS scenario, the large dependence on cogeneration technology immediately stands out. When the existing nuclear power plants are phased out in 20152025, they are still largely replaced by gas-fuelled electricity generation (STAG power plants), although in later years (2030-2050), this technology loses some market share to cogeneration technologies. Renewables gain a market share of 20% in 2050 (all options are used up to their full potential), cogeneration even 37% (remember that the ‘Rational Perspective’ scenarios include ‘optimistic’ assumptions about constraints to both technological options). The dominant technology for cogeneration is the hydrogen-fuelled fuel cell for high-temperature steam in industry (23 TWh in 2050), although some applications can also be found in centralised and decentralised heating for the residential and service sectors.


120

100

Renewables

80 TWh


60

Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050


397


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 13. Electricity production - scenario M.K.P.LCS (top) and R.K.P.LCS (bottom)

CO2 emissions In the M.K.P.LCS scenario, the electricity production sector becomes the largest contributor to total CO2 emissions in 2050 due to the increasing reliance on natural gas as the energy vector of choice. CO2 emission reduction efforts are largely borne by the residential and service sector. Emission reductions are realised through the diffusion of solar boilers (providing for 108 PJ of space heating and warm water in 2050), efficient technologies (heat pumps and natural gas condensing boilers) and conservation measures. Also, there is a substantial decrease in CO2 emissions from the transportation sector in the period 2040-2050: this is because (imported) ethanol is introduced as a fuel for the transport fleet574.


574

2000

2010

2020

2030

2040

2050

Electricity

20

14

17

25

28

29

Industry

34

25

27

23

23

23

Resid. & service

34

33

29

21

10

9

Transportation

26

25

22

22

21

13

Other

4

4

3

3

3

3

Total

118

101

98

94

85

77

The ethanol fuel price is assumed to be 13.2 MEuro/ PJ (compare to the gasoline fuel price of 3.59 – 9.69 Euro/PJ).

398

Chapter 6


2000

2010

2020

2030

2040

2050

Electricity

19

13

16

20

18

16

Industry

34

27

27

25

26

24

Resid. & service

32

26

22

18

16

14

Transportation

26

30

28

25

21

19

Other

4

4

3

3

3

3

Total

115

100

96

91

84

76

Table 23. CO2 emissions – all sectors (scenario M.K.P.LCS (top) and R.K.P.LCS (bottom))

In the R.K.P.LCS scenario, CO2 emissions from the electricity sector are markedly lower (a combination of lower electricity demand and higher potential for renewables and cogeneration). This allows for less stringent (and costly!) reduction efforts in residential and transportation sectors.

3.4.1.4 Phase out, low potential carbon capture and storage and import of electricity (scenario M.K.P.LCS.I and R.K.P.LCS.I) Primary energy use


2500

PJ


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050


399


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 14. Primary energy demand – scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom)

Primary energy use remains rather flat for both scenarios575. As in the previous scenario cases, the increasing dependence on gaseous energy carriers (once existing nuclear power plants are phased out) is obvious. Renewable energy sources provide for 14% of primary energy demand in the R.K.P.LCS.I scenario. Electricity production In the M.K.P.LCS.I scenario, nuclear electricity generation is replaced by gas-fuelled electricity generation (in STAG power plants, up to 40% in 2050), imports of foreign electricity (up to 32% in 2050), electricity from renewable sources (up to 12% in 2050) and cogeneration (up to 13% in 2050). Cogeneration is (again) mainly used to produce high temperature steam for industry. Hydrogen-fuelled fuel cells become the dominant technological choice for this option. In the R.K.P.LCS.I scenario, electricity demand is generally much lower. The reliance on STAG technology is smaller compared to the previous case: nuclear electricity generation is replaced by a mix of cogeneration (up to 26% in 2050), renewables (up to 17% in 2050) and import of foreign electricity (up to 36% in 2050). Gas-fuelled electricity generation in STAG power plants makes up for the rest (up to 18% in 2050). The preferred cogeneration technology is the gas turbine (providing 11 TWh of electricity in 2050), although in the later years (2040-2050) it is increasingly being replaced by hydrogenfuelled fuel cell technology (providing 6 TWh of electricity in 2050).

575 The imported electricity has been subdivided according to origin (renewable, nuclear, gas, coal) and then added to the corresponding primary energy category.

400

Chapter 6


140

120

TWh

100

Import Renewables Cogeneration Gas Coal Nuclear

80

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


100

TWh

80 Import Renewables Cogeneration Gas Coal Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 15. Electricity production - scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom)

CO2 emissions Table 24 represents the sectoral CO2 emissions for both scenarios. CO2 emissions (or other emissions) resulting from imports of foreign electricity are not included in this table; it is assumed that neighbouring countries also have to comply with a comparable CO2 emission reduction target in 2050.



2000

2010

2020

2030

2040

2050

Electricity

19

11

12

15

15

18

Industry

34

25

25

23

23

23

Resid. & service

34

31

26

25

23

13

Transportation

26

30

31

27

21

19

Other

4

4

3

3

3

3

Total

117

101

97

93

85

76

2000

2010

2020

2030

2040

2050

Electricity

18

12

13

16

15

13

Industry

34

27

27

28

28

27

Resid. & service

32

27

23

19

17

13

Transportation

26

30

29

25

21

19

Other

4

4

3

3

3

3

Total

114

100

95

91

84

75


401

Table 24. CO2 emissions – all sectors (scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom))

These imports notwithstanding, CO2 emissions in the electricity sector rise after 2020 in the M.K.P.LCS.I scenario due to the increasing reliance on (domestic) STAG electricity production. In the R.K.P.LCS.I scenario, emissions can be kept low due to the higher potential for cogeneration and renewables. Again, a large part of the necessary reductions can be achieved in the residential and service sectors. The R.K.P.LCS.I scenario relies more on centralised district heating (45 PJ in 2050) and decentralised combined heat-andpower (28 PJ in 2050) than the M.K.P.LCS.I scenario (0 PJ and 22 PJ respectively). 3.4.2

Comparison between scenarios

In this part, a summary of the most important results is given. Some graphs explaining differences in final energy demand and electricity demand in the above-mentioned scenarios will be explored. More complete results on quantifiable scenario indicators (including sensitivity analyses) are given in the next section. 3.4.2.1 Final energy demand The final energy demand projection for all scenarios is considerably lower than in the baseline ‘Market Driven’ case. For the ‘Rational Perspective’ scenarios, this difference amounts to some 30% in 2050. This is due to substantial increases in the efficiency of enduse technologies. In the RP scenarios, more efficient technologies are introduced than in the MD scenarios, due to the lower discount rate applied. In the base case, final energy demand would reach some 2100 PJ in 2050, while most MD scenarios level off at about 1800 PJ (a 10% increase compared to 2000). The M.K.P.LCS scenario is an exception: here, supply-side measures to reduce CO2 emissions are limited in potential, so that in this

402

Chapter 6

scenario reductions in energy demand comparable to most RP scenarios have to be achieved (notice however the much more drastic reduction effort in 2020-2030 as a result of the higher discount rates). In the later years (2040-2050), some supply side measures (ethanol fuelled cars for transport and solar boilers for heating) are again preferred. Most RP scenarios level off at some 1600 PJ (or even lower) – almost equal to final energy use in 2000. This would on average imply that the improvements in energy intensity of the national economy have to keep up with economic growth. 2,200.00

2,100.00

2,000.00 BASE MKLCS MKP

1,900.00 PJ

MKPLCS MKPLCSI RKLCS 1,800.00

RKP RKPLCS RKPLCSI

1,700.00

1,600.00

1,500.00 2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Figure 16. Final energy use (all scenarios)

3.4.2.2 Electricity demand Electricity demand varies considerably between the different scenarios. The increase in electricity demand until 2050 ranges between 30% and 100% (a doubling) compared to the electricity demand in 2000. This means that, in all scenarios, electricity will increase its share in meeting the final energy demand. A shift from other energy vectors towards electricity thus appears to be a relatively robust option for achieving CO2 emission reductions under a variety of assumptions.


403

160.0

150.0

140.0

130.0 MKLCS 120.0

MKP

TWh

MKPLCS MKPLCSI

110.0

RKLCS RKP RKPLCS

100.0

RKPLCSI 90.0

80.0

70.0

60.0 2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Figure 17. Electricity production (all scenarios)

The choice of technological options for the electricity production also clearly influences the level of electricity demand in the MD scenarios. For instance in the M.K.P.CS scenario, a comparatively cheap source of carbon-free energy is available (electricity produced by fossil fuel-based power generation with carbon capture and storage), and MARKAL will correspondingly use this energy carrier to a maximum degree576. This is also true for the M.K.LCS and M.K.P.LCS.I scenarios, but to a lesser extent: the availability of carbon-free electricity is constrained by the 8 GWe limit for domestic nuclear electricity production, and roughly one third of total electricity demand for import of electricity. In the scenario variants with high electricity demand, electricity increasingly replaces fossil fuels, for instance for water and space heating (heat pumps gain a large market share). Impacts of electricity generation options on electricity demand are less outspoken in the RP scenarios, because carbon-free electricity sources have to compete with demand-side measures (e.g. better isolation) for a market share.

576

A sensitivity analysis with 30% higher investment costs for all electricity generation option equipped with CCS showed no significant impact on the level of output of these power plants.

404

3.4.3

Chapter 6

Detailed results (including sensitivity analysis)

This section gives some detailed results on the quantifiable criteria for the different scenarios. For each scenario, two sensitivity analyses were performed: one with higher costs for fossil fuel imports (gas and oil) (cf. Table 25), the other with a more stringent CO2 reduction target (-7.5% in 2010, -30% in 2030, -50% in 2050) and lower demands for energy services in the industrial and transportation sector (-30% for both sectors in 2050 compared to the base scenarios, with linear interpolation from 1990 demand levels). This second sensitivity analysis might for instance represent an economic development with structural changes towards less energy intensive industries (e.g. more service-oriented), less energy-intensive production processes in industrial sectors (e.g. use of new materials to replace steel, use of membrane technology, etc.), higher efficiency gains than presumed in the base scenarios, or the emergence of new consumer values (e.g. less materialism, ecofriendly consumption, etc.). Results are presented on a comparative basis (the exact magnitude of the parameters involved is of less importance), with ‘optimistic’ and ‘pessimistic’ scores for each scenario (i.e. the ‘best’ and ‘worst’ result for the scenario in question, including both sensitivity analyses). Low scores for a certain criterion always represent a ‘bad’ score, even though the primary indicator on which the criterion is based might be higher for the ‘bad’ result. For instance, if the cost for scenario A would be in the range of 80 - 100 MEuro, and for scenario B in the range of 30-50 MEuro, scenario A would have a score of 0 - 20 and scenario B a score of 50 - 70 on a scoring scale where 0 = 100 MEuro and 100 = 0 MEuro.

Cost (MEuro/PJ)

1990

2000

2010

2030

2050

Import of High Sulphur Heavy Distillate

2.20

1.71

2.97

4.95

6.91

Import of Low Sulphur Heavy Distillate

2.52

1.96

3.39

5.64

7.87

Import of Gasoline

4.88

3.59

5.54

9.23

12.90

Import of Light Distillate Oil

4.21

3.19

5.30

8.84

12.35

Import of LPG

3.07

2.75

3.27

4.75

6.24

Import of Natural Gas

2.48

1.68

2.97

4.33

5.67

Table 25. Resource prices (sensitivity analysis)


405

3.4.3.1 Health and environment Impacts on public health caused by air pollution (during power generation stage) Impacts on public health caused by air pollution (SO2, NOx and particulate matter (PM)) mainly arise during the power generation stage. In order to estimate future emissions of these pollutants, emission coefficients of SO2, NOx and PM associated with the different electricity producing technologies have been added to the Belgian MARKAL database on an average basis (therefore they can only be considered as indicative). For our purposes, the impacts of air pollutants on public health have been weighted according to the monetary valuation of the damages they cause. Results are taken from the ExternE project of the European Commission (cf. Chapter 3). The figures used here are based on the framework developed in this project, although at a more aggregated level (cf. Table 26).

Power generation All fuels (Euro/ton) SO2

5,998

NOx

5,000

PM

11,989

Table 26. External costs of air pollutants (Source: AMPERE Principal Report, Section I, p. 6)

Briefly, the ExternE methodology involves the estimation of external costs (based on the willingness-to-pay or willingness-to-accept concept) based on the impact pathway (starting from primary emissions, transportation and atmospheric chemistry, deposition processes and different burdens are taken into account). The damage categories considered in Table 26 are acute morbidity (e.g. emergency room visits, restricted activity days, etc.), mortality and chronic morbidity (e.g. chronic bronchitis, asthma, non-fatal cancers, etc.), but no occupational health effects (these are treated under a separate heading, because they mainly arise during other stages of the fuel cycle, mostly outside Belgium). Also, damages to materials and agricultural crops are included, although these damage categories only contribute to a minor extent to the total cost figure. It is clear that measuring environmental costs at such a global level as in this model raises different problems (which have been discussed in more detail in chapter 3). However, despite all uncertainties involved, external cost calculations can still provide an informative, comparative and quantified indicator of the health impacts of the different scenarios. Other (more precise) possible indicators (such as the absolute emission levels of the pollutants under consideration) give no information on the actual health impacts associated with these emission levels. Results are shown in Table 27. For each scenario, cumulative

406

Chapter 6

(undiscounted) health damages have been calculated for the time periods 1990-2020 and 2020-2050. The differences between the scenarios over the period 1990-2020 are relatively small. However, in 2020-2050, scenarios involving a large reliance on coal technologies generally score lower (M.K.P.CS and R.K.P.CS). The best performing scenarios are those which allow for a continuation of nuclear power (M.K.LCS and R.K.LCS) and those which rely on imports of electricity (M.K.P.LCS.I and R.K.P.LCS.I). Score (0-100): 0=25000 ; 100=2500 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

1990-2020 Best score

30

29

25

30

26

25

25

25

Worst score

26

24

21

27

23

22

23

23

Best score

96

51

86

94

98

60

91

89

Worst score

93

22

81

91

91

26

87

83

2020-2050

Table 27. Impacts on public health caused by air pollution

Impacts on occupational health (coal and gas fuel cycle) Burdens on the workforce employed in the gas and coal fuel cycle arise mostly in other stages than the power generation stage. For the gas and coal fuel cycle, monetary values for these impacts have also been calculated (De Nocker et al., 1998). For Belgium, values of 0.69 mEuro/kWh are used for the coal fuel cycle and 0.07 mEuro/kWh for the gas fuel cycle. Impacts on occupational health mainly arise due to accidents (causing minor injuries, major injuries or death) during coal mining and transport of coal or waste materials, or due to radon exposure (causing lung cancer) or diseases caused by coal dust (chronic bronchitis, chronic cough, etc.). For the gas fuel cycle, impacts are mainly accident-related and generally lower than for the coal fuel cycle. Impacts arising during the nuclear fuel cycle are described under a separate section. Table 28 summarises the cumulative external cost implications (1990-2050) for the different scenarios. Clearly, scenarios where coal is used intensively score rather low on this criterion. Score (0-100): 0=2500 ; 100=0 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

84

53

78

82

85

51

81

84

Worst score

82

8

72

76

82

18

80

81

Table 28. Impacts on occupational health (coal and gas fuel cycle)


407

Radiological health impacts of nuclear fuel cycle Radiological impacts of the nuclear fuel cycle have also been quantified according to the ExternE methodology (De Nocker et al. 1998; AMPERE Principal Report, Section I, p. 6). Due to the long half-lives of some radionuclides, low-level doses will exist into the far future. These low-level doses can add up to larger values when the total population dose is considered for thousands of years. Of course, the uncertainty of the models used to evaluate damages increases over this time span, while the level of doses also falls into the range where there exists scientific uncertainty of resulting radiological health effects. Issues of discounting and the spatial boundaries used in the analysis thus become very important. Also, health impacts include hereditary effects, a category not encountered in health impacts in other fuel cycles. Thus, participants in the multi-criteria scoring exercise are given the opportunity to reflect their judgement on these issues by providing the opportunity to score impacts of the nuclear fuel cycle on a separate basis. Table 29 gives results for the different scenarios based on a central estimate of 0.69 mEuro/kWh (0% discounting, impacts studied over 10000 years). Score (0-100): 0=2500 ; 100=500 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

12

79

79

63

13

78

78

73

Worst score

12

79

79

63

13

78

78

73

Table 29. Radiological health impacts of nuclear fuel cycle

Need for very long term management of high-level waste The output from fission power includes chemical and low-level radioactive wastes, as well as used spent fuel, which is the main disposition challenge. Spent fuel from nuclear power reactors contains approximately 5% fission products (atoms produced by splitting another atom or by radioactive decay of another fission product), of which approximately 10% are long-lived fission products (LLFP); and 1% plutonium and other minor actinides (MA)577. Together, these products are responsible for the radiotoxicity on a very long timescale. Table 30 summarises the cumulative production of MA and LLFP (1990-2050) in the different scenarios. MA and LLFP resulting from the import of nuclear electricity from foreign countries are taken into account. Scenarios which incorporate a nuclear phase out score equal. The differences between the M.K.LCS and the R.K.LCS scenario relate to the different technologies for nuclear electricity production.

577

The exact figures used were, for the LWR: 22 kg Pu/TWh, 4 kg MA/TWh and 12 kg LLFP/TWh; and for the MHTGR: 15 kg Pu/TWh, 2 kg MA/TWh and 9 kg LLFP/TWh (Charpin et al., 2000).

408

Chapter 6

Score(0-100): 0=125 ; 100=0 (ton)

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

1

59

59

46

19

59

59

55

Worst score

1

59

59

46

19

59

59

55

Table 30. Need for very long-term management of high-level waste (HLW)

Impacts on natural ecosystems caused by air pollution Effects of electricity generation on natural ecosystems are very difficult to quantify in monetary values. Ecosystems may be damaged by fuel cycles in a number of ways, but the most serious and widespread effects are caused by acidic and nitrogenous deposition and photo-oxidants. Ideally, one should identify various types of ecosystems in the neighbourhood of the power plant(s) under consideration, and estimate the exceedance of critical loads or levels of pollutants in these ecosystems caused by the emissions of the power plant. Such a detailed analysis could not be performed for this exercise. As a proxy for the acidification damages caused by the power generation stage, we have calculated a composite indicator based on the SO2 and NOx emission levels. SO2 emissions have been given a weight of 2 compared to NOx emissions, since the former provide for two acid equivalents in the acidification process, while the latter provide maximum one (so in fact, we calculated a conservative estimate). Results for the different scenarios are shown in Table 31 (cumulative emissions over the period 1990-2020 and 2020-2050). Score (0-100): 0=5000 ; 100=500 (kton acid eq.) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI


14

13

11

14

10

9

10

10

Worst score

11

9

10

12

9

6

8

8

Best score

98

43

85

95

100

49

93

90

Worst score

95

0

80

92

93

5

88

84

2020-2050

Table 31. Impacts on ecosystems caused by air pollution

Environmental impacts caused by solid waste from coal fuel cycle The treatment of solid waste matter and by-products from the coal cycle deserves special mention, since these impacts have not been captured in external cost calculations. The following types of wastes can be produced: mine wastes (accumulated at surface in slag heaps), furnace bottom ash (FBA), pulverised fuel ash (PFA), flue gas desulphurisation


409

sludge, and gypsum. However, impacts associated with waste utilised elsewhere (which should rather be referred to as by-products) should be considered as part of the system to which they are transferred from the moment that they are removed from the boundaries of the fuel chain (e.g. FBA and PFA can be used by construction industry). It is of course important to be sure that a market exists for any such by-products. The capacity of, for example, the building industry to utilise gypsum from flue gas desulphurisation systems for coal power plants is clearly finite. If it is probable that markets for particular by-products are already saturated, the ‘by-product’ must be considered as waste instead. A further difficulty lies in the uncertainties about future management of waste storage sites. For example, if solid residues from a power plant are disposed in a well engineered and managed landfill there is no impact (other than land use) as long as the landfill is correctly managed; however, for the more distant future such management is not certain. Also, the impact of liquid effluents on the receiving surface waters has not been quantified, but can play an important role in the mining and production stage. As a rough comparative proxy for all these impacts, we have calculated the cumulative use of coal578 (1990-2050) in the different scenarios (cf. Table 32). Score (0-100): 0=35000 ; 100=10000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

88

55

88

87

83

47

86

84

Worst score

83

3

84

84

73

5

80

75

Table 32. Environmental impacts caused by solid waste from coal fuel cycle

Catastrophic risk: nuclear No attempt was made to quantify the catastrophic risk associated with nuclear power use, neither in terms of probability of occurrence multiplied by expected number of casualties, nor in monetary terms (expected damages). Rather, we have simply calculated the total amount of TWh produced by nuclear power plants in the different scenarios as a (rough) comparative measure of the risk of actually experiencing a catastrophic accident (all else being equal – e.g. we do not take into account the reported lower probability of occurrence for new reactor types such as the modular high-temperature gas reactor). It is left up to the participants in the multi-criteria scoring to weight this nuclear risk according to their own framing of the issue (e.g. which might include other dimensions such as irreversibility of the risk, gravity of single events, issues of intragenerational equity, control, trust in the institutions responsible for managing the risk, etc.). Electricity generated by foreign nuclear power plants is included in Table 33. Fatalities in other fuel cycles are covered by the ‘occupational health’ criterion. This way, we were able to make

578

1 PJ of coal equals some 40000 tons of coal approximately.

410

Chapter 6

a distinction between effects on individuals and effects that potentially threaten society as a whole. Score (0-100): 0=3500 ; 100=0 (TWh)

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

7

61

61

49

8

61

61

57

Worst score

7

61

61

49

8

61

61

57

Table 33. Catastrophic risk (nuclear)

3.4.3.2 Economy Intensity of energy use Table 34 summarises the intensity of energy use for the different scenarios. Since it is assumed that all scenarios attain equal economic growth, final energy use (cumulative over the period 1990-2050) is a suitable indicator for the intensity of energy use (defined as final or primary energy per unit of GDP). Intensity of energy use can be seen as a criterion to measure the general ‘rationality’ (in the logic of the industrial/market commonwealth) of the ways in which society makes use of energy resources, e.g. avoiding unnecessary squandering. Score (0-100): 0 = 120000 ; 100 = 90000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

57

59

67

63

76

83

91

84

Worst score

17

18

33

24

38

46

53

47

Table 34. Intensity of energy use

Security of energy supply Imports of gas and oil can be used as a rough indicator for security of energy supply. Security of supply is most commonly defined as the robustness of the Belgian energy system to ensure that Belgian citizens will not be exposed to shortages of energy and that Belgium is less vulnerable to international policy or conflicts in this area. Therefore, besides keeping import of energy resources to a minimum, this criterion also reflects a concern for regional diversification of import sources (especially with regard to regions which might be vulnerable to risks of erupting wars or political upheavals). It is generally accepted that both the coal and the nuclear fuel cycle are less vulnerable to this aspect.


411

Score (0-100): 0=110000 ; 100=85000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

68

40

20

46

95

88

70

73

Worst score

21

12

0

0

58

52

27

35

Table 35. Security of energy supply

Table 35 summarises the cumulative reliance (1990-2050) on imports of gas and oil for all scenarios. Imports of electricity produced by gas are taken into account. ‘Rational Perspective’ scenarios score better than their ‘Market Driven’ counterparts. Scenarios which allow for a nuclear continuation score better on this criterion. Total system costs Table 36 summarises the total annualised system costs for the different scenarios (in 2010, 2030 and 2050), expressed as a percentage of the Belgian GDP in 2000. Total system costs include all cost parameters modelled in MARKAL, i.e. investment costs for demand and supply technologies, fuel costs, delivery costs for fuels, operating & maintenance costs, variable costs for power plant operation, etc. – for all sectors of the energy system. Score (0-100): 0=5% ; 100=0% of GDP 2000 579 MKLCS MKPCS MKPLCS MKPLCSI RKLCS RK.PCS RKPLCS RKPLCSI

2010

91

91

91

91

96

96

96

95

2030

88

73

65

66

94

83

81

79

2050

62

55

0

24

88

78

69

62

Table 36. Total system costs580

The reported costs are relative with regard to the respective baseline scenarios (‘Market Drive’ or ‘Rational Perspective’) and thus represent the additional cost in order to meet GHG emission requirements. Therefore, ‘Rational Perspective’ and ‘Market Driven’

579

GDP2000 equals 247.7 billion Euro. Actually, the total system costs for M.K.P.LCS in 2050 are 8.6% of GDP2000, but because all other costs fell in the range of 0-5% we have chosen to treat this result as an (unrealistic) outlier and have attributed it with a score of 0. Compared to Proost and Van Regemorter (2000), these costs are rather low. These authors foresee a cost of 2.7% of GDP2000 for a post-Kyoto scenario (-15% GHG emissions in 2030) with a nuclear phase out (as planned), while our M.K.P.LCS scenario (which comes closest to the assumptions used by Proost and Van Regemorter) this cost would be 1.7% of GDP2000 in 2030. However, costs in our case rise rapidly after 2030. 580

412

Chapter 6

scenarios are not directly comparable, since both scenario groups differ in the discount rates used. The lower discount rate of the ‘Rational Perspective’ scenario group implies better anticipation of the future conditions facing the energy system, and hence a different investment behaviour. Scenarios where nuclear energy is phased out, with a low potential for carbon capture and storage and no import of foreign electricity (P.LCS) generally show higher economic costs to meet the CO2 emission reduction demands in 2050. Also, generally higher costs are incurred in the period 2030-2050, since much of the existing energy infrastructure will have to be replaced by then. Marginal cost of electricity The MARKAL model also calculates the marginal costs for different energy carriers, i.e. the delivery cost for one extra unit of the energy carrier under consideration. Also for electricity, this cost is calculated (without consideration of cost of distribution or a margin of profit for the producer). Results for the average yearly marginal costs of electricity for the periods 1990-2020 and 2020-2050 are given in Table 37 for the different scenarios. The rapid increase in marginal costs of electricity in the period 2020-2050 is explained by the absolute limits imposed on CO2 emissions: every additional kWh must be produced without adding significantly to the global carbon budget (especially in the ‘Market Driven’ scenarios, were demand-side measures are less readily available than in the ‘Rational Perspective’ scenarios), which implies some rather radical changes in the technologies used for electricity generation and the use of existing installations. Actually, in the MPLCS scenario marginal costs were so prohibitively high that we have chosen to attribute a score of 0 for this scenario. Score (0-100): 0=0.15;100=0.03 (Eurocent/kWh) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI


88

88

82

92

93

93

93

96

Worst score

77

75

68

79

86

86

86

92

Best score

50

65

0

27

83

76

56

67

Worst score

44

61

0

15

80

74

50

62

2020-2050

Table 37. Marginal costs of electricity

3.4.3.3 Social, political, cultural, ethical needs Need for intermediary storage of spent fuel Spent fuel from nuclear power plants will have to be stored for several decades in order to cool down, also implying that appropriate institutions for control have to be kept in


413

place. Table 38 summarises the total amount of spent fuel produced in the different scenarios. Score (0-100): 0=7500 ; 100=3000 (ton)

MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

5

87

87

87

35

87

87

87

Worst score

5

87

87

87

35

87

87

87

Table 38. The need for intermediary storage of spent fuel

The R.K.LCS scenario scores better than the M.K.LCS scenario because the hightemperature gas reactor (deployed in this scenario) operates at a higher burn-up rate than the standard light-water reactor technology. All scenarios with a nuclear phase out have an equal score on this criterion. Use of non-renewable resources The non-use of non-renewable resources (oil, gas, coil and uranium) can be regarded as a measure of leaving development opportunities open for other (Third World) countries. Table 39 summarises the cumulative use of non-renewable resources (1990-2050) in the different scenarios. Score (0-100): 0 = 135000 ; 100 = 100000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

81

38

49

40

97

66

81

64

Worst score

45

0

30

6

67

33

48

33

Table 39. Use of non-renewable resources

Degree of decentralisation of electricity sector Table 40 summarises the installed capacity of decentralised electricity production technology (mainly cogeneration and photovoltaics) in 2050 as a percentage of the total installed capacity in Belgium. Decentralised electricity production is defined as production that is fed directly into local power grids. Decentralisation of the electricity sector might be considered to be important for a number of reasons – e.g. making energy production more ‘visible’ to the users, enabling local participation in energy projects, limiting the powers of large energy companies, etc. Thus, this criterion will tend to overlap with other criteria taken from the third category, although it is not exactly identical (e.g. a large energy company might still be the owner of most of the decentralised power units). Offshore wind power is not counted as a decentral technology. Also, the scores reported in

414

Chapter 6

Table 40 relate to the installed capacity and thus not necessarily to the amount of electricity effectively produced by these decentralised technologies. ‘Rational Perspective’ scenarios generally allow for a higher percentage of decentralised electricity production technologies. Score (0-100): 0 = 0% ; 100=100%

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

38

40

39

44

57

58

58

65

Worst score

38

40

39

44

57

58

58

65

Table 40. Degree of decentralisation of electricity production

3.4.3.4 Diversity of the electricity production park Diversity is generally seen as a good strategy to deal with the many incertitudes facing an energy system. Discussion of diversity in the energy debate has tended to remain confined to the issue of uncertainty concerning the security of future energy supplies. But new insights (Stirling 1994) also relate diversity to a more general protection against ignorance, for instance in financial or environmental performance, and even as a potential way to reconcile different values and interests related to controversial issues. As a measure of diversity, Stirling proposes to use the so-called Shannon-Wiener function581. Table 41 summarises the value of this diversity function (in 2050) for the different electricity production parks (as a result of the different choices made in the scenarios). For the calculation of this function, we used the following broad categorisation of electricity generation options: nuclear, gas, coal, cogeneration, renewables and import. Score (0-100): 0=0.7 ;100=1.5

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

82

63

40

85

67

22

63

96

Worst score

79

31

24

81

46

8

59

91

Table 41. Diversity of electricity production park

581

H = -∑ pi ln pi , with H the value taken by the diversity index for a mix of options, pi the proportional reliance on option i and ln the natural logarithm. This index captures both the variety (the number of options) and balance (the relative importance of the different options in the mix).


415

4 Summary and conclusions In this chapter a first attempt has been made to achieve some tractability in the otherwise ‘unstructured’ debate on which direction a sustainable development of the energy system should take. The first necessary step in this process entails arriving at an agreement on the list of ‘actants’ to be taken into account in deliberations. On this point, while we concede we cannot give absolutely watertight guarantees that we have fully covered all possible points of view, we are still confident that our combined value tree is reasonably comprehensive, as it is based on the inputs of the different perspectives which have been most ‘vocal’ in the general energy debate. In addition, the process by which the tree was constructed (i.e. a bottom-up interview approach) is likely to provide it with some political legitimacy. However, in order to function as a ‘boundary object’, the combined value tree should also meet scientific quality criteria. Assuring scientific legitimacy depends upon requirements such as the logical coherence and non-redundancy of the value tree structure, and the general pertinence, fidelity, specificity and sensibility of the associated indicators. While we have tried to meet criteria of logical coherence through the use of Boltanski and Thévenot’s commonwealth model as a structuring framework, (quantitative) indicators have at this stage only been proposed for some of the most ‘uncontroversial’ criteria. Further work should thus include a discussion about which of the criteria are most pertinent for being measured; how they could be measured and perhaps how different measurement should be aggregated (Chapter 7 will provide some further indications on this topic). In a second step, we constructed a number of scenarios covering a broad range of strategic options for the future of the Belgian energy system. This scenario approach allowed to give an idea of the magnitude of the impacts associated with these strategic options under two different worldviews (one stressing ‘market’ values, the other ‘civic’ values). Again, here we took care to tailor our scenario approach to the needs of different parties involved in the discussion. For instance, we relied on the most simple form of engineering-economic modelling without making use of increasingly complex macroeconomic modelling tools. Choosing this approach might perhaps make the exercise less credible in the eyes of experts in econometric modelling, but, in return is likely to increase its overall transparency. While both the value tree and the scenarios were discussed in a workshop, we chose to further test their capacity for functioning as boundary objects in the ‘unstructured’ debate on sustainable energy policy in a subsequent individual multi-criteria exercise. The next chapter gives the details and conclusions following from this exercise, as well as suggesting areas where common ground can be found.

The combined value tree

I.

High-level criterion

Intermediate level criterion

Low-level criterion

Environmental & human health and safety

Air pollution

Impacts of air pollution on human health: mid-term Impacts of air pollution on human health: long-term Impacts on occupational health (gas+coal) Radiological health impacts (nuclear) Need for long-term management of HLW Visual impact on landscape Noise amenity Impact on natural ecosystems (air pollution): mid-term Impact on natural ecosystems (air pollution): long-term Environmental impact from solid waste (coal) Land use Water use Catastrophic risk: nuclear Geographical distribution risks / benefits

Occupational health Radiological health impacts Aesthetic impacts Other environmental impacts

Resource use Other energy related pressures

II.

Economic welfare

Overall economic benefit

Producer need/benefit

Consumer need/benefit

Intensity of energy use Security of energy supply Distribution of economic benefits / burdens Economic risks Overall cost energy system: mid-term Overall cost energy system: long-term Ability to provide specialist market Required investment in supply technology: mid-term Required investment in supply technology: long-term Net expenditure on fuels: mid-term


International co-operation Need for government intervention III.

Social, political, cultural and ethical needs

Individual/consumer choice/benefit

Consumer choice Citizen participation Contribution to rational energy use

Institutional needs/benefits

Degree of decentralisation Need for intermediary storage of spent fuel Control and concentration of power Influence on political decision-making Need for socio-political stability Need for direct political intervention Reversibility of technology choice Knowledge specialisation Need for institutional non-proliferation measures Potential for technology transfer Leaving resources for development Equity (general) Job opportunities in the energy system

Development opportunities

Jobs IV.

Diversification

Net expenditure on fuels: long-term Marginal cost electricity: mid-term Marginal cost electricity: long-term Strategic factors for export Compatibility with international R&D agenda Amount of direct or indirect subsidies needed

Diversification

417

CHAPTER 6 CRITERIA

AND SCENARIOS AS A SUPPORT FOR SUSTAINABLE ENERGY

GOVERNANCE In chapters 6 and 7 of this dissertation, an attempt is made to put the first stages of our model governance process (problem structuring and scenario building) into operation in the form of a limited ‘pilot exercise’. In the present chapter, a comprehensive list for evaluating energy systems is derived from the interviews with representatives of the FRDO (cf. Chapter 4). A hierarchical representation of criteria was logically structured for each interview result separately, and then aggregated into a ‘combined value tree’ (Section 2). With the aid of the energy model MARKAL we developed four broadly conceived longterm energy strategies for Belgium with contrasting economic, social, environmental and political implications. Section 3 contains an in-depth discussion of these scenarios in terms of their central assumptions, hypotheses and results. The capacity of both criteria and scenarios for functionally supporting the energy governance framework proposed in chapter 5 will be empirically tested in the next chapter (Chapter 7).

1 Introduction In chapter 5, we set out the fundamentals of a new governance structure that would be able to ensure some measure of stability (in view of achieving the long-term goal of sustainability) while still taking into account the rapidly changing dynamics of the energy system. In our attempt to negotiate the difficulties engendered by these contradictory givens, a large role was reserved for both the institutional and the formal analytic dimension. The institutional dimension was concerned with the creation of ‘boundary organisations’ – i.e. organisations with a sufficient representation of different roles (e.g. scientists, ethicists, politicians, stakeholders, etc.) in order to uphold accountability both to the formal ‘political system’ and the ‘scientific world’. The formal analytic dimension on the other hand entailed the creation of ‘boundary objects’ (e.g. sustainable energy criteria, scenarios, models, etc.) as a result of the negotiations going on in the ‘boundary organisations’. The present chapter is concerned with setting out a suggestion for two of these ‘boundary objects’, that is a structured value tree of sustainable energy criteria and a set of long-term energy scenarios supported by a modelling tool for discussing the strategic dimension of long-term energy governance. Fundamentally, the use of such formal analytical tools should help to ensure that the quality criteria set out in chapter 5 can be met – i.e. the approach should be more conducive towards a fair, transparent, competent and uncertainty-sensitive treatment of policy options. This is necessary since the simultaneous consideration of quantitative, qualitative and intuitive aspects of a complex problem such as future energy policy is usually a difficult

356

Chapter 6

task for an untrained human mind, as different cognitive biases loom on the horizon. For example, a politician making up his mind on energy policy measures may easily escape the duty of taking into account all relevant actants and facing trade-offs by focussing exclusively on part of the information which is new, easy to understand or compatible with earlier preferences and knowledge (Hämäläinen 1990). In this case, a decision aid such as a structured list of ‘actants’ to be taken into account in arriving at a decision is imperative for assuring some consistency and coherence. It can for instance also enable focussing the debate on the crucial issues by eliminating unimportant ‘actants’, and make different ways of reasoning understandable to others. However, all of these advantages should not make us blind to the fact that such formal analytic methods also potentially face some difficulties, partly due to methodological foundations (which constrain the types of models than can be justified), but also partly due to the inherent ‘power play’ in political debate and deliberations. For instance, the use of formal decision analytic tools shifts the discourse into a more structured framework and, by doing so, inevitably affects the stakeholders’ negotiation positions. Thus, besides the ‘formal’ part of decision aids – i.e. the ways in which scientific information and knowledge about ‘actants’ is managed from a modelling perspective – the ‘interactive’ part – i.e. the interactions and deliberations through which the different actors contribute to the analysis and take part of its results –is also of major importance (Rauschmayer 2000). In this chapter, the ‘formal’ part will be discussed, as we will explain and defend our choices for particular decision analytic methods, namely a value tree analysis (cf. Section 2) and engineering-economic modelling of long-term energy options (Section 3). Discussion of the ‘interactive’ part will be left over for chapter 7. However, before going into the details of the formal part of the analysis, we want to warn against one possible major misconception. In no way should the methods proposed here be considered as definitive (be it just for the fact that the ‘protected experimental settings’ of PhD research differ fundamentally from the ‘real-world settings’ of a concrete governance initiative); on the contrary, they should merely be regarded as a first instigation – a limited ‘pilot exercise’. Chapter 7 will conclude with the lessons that can be learnt from this exercise, pointing out ways for solving some of the difficulties encountered.

2 A structured value tree for sustainable energy criteria

2.1


One of the first tasks we have set out for a new energy agency (which occupies the ‘central’ role of collecting and transmitting learning experiences occurring in the enlarged ‘public sphere’) was to draw up a list of ‘actants’ (including the criteria they should uphold) which should have a seat in the jury deciding about the sustainability of our energy system. Ideally, the initiative should be taken by an interministerial conference (presided


357

by a high-level government representative) and entrusted to a group of experts of international standing which take up their role as ‘facilitators’ (cf. Chapter 5, and also Boulanger et al. 2003). This procedure is of course entirely unrealistic within a strictly research-oriented framework. For our research purposes, we have chosen to apply a method which has already proven its value also in energy-related questions (Keeney et al. 1987), namely the ‘value tree’ methodology. A value tree identifies and organises the values of an individual or group with respect to possible decision options. In the process of structuring a value tree, representatives of different stakeholder groups are asked to identify their criteria and objectives for evaluating different options. Translated in constructivist parlance, we could say that a value tree helps in establishing the ‘jury of actants’ who are considered to be relevant for the decision at hand (i.e. the long-term evolution of the energy system)553. A value tree structures the elicited values, criteria, and corresponding attributes in a hierarchy, with general values and concerns at the top, and specific criteria and attributes at the bottom. Theoretically, there are two approaches to generating value trees, namely top-down and bottom-up. The topdown approach is deductive-analytic, starting from general principles going down to more operational criteria, while the bottom-up approach is inductive and synthetic (Salo 1999). For our purposes, we chose to combine both top-down and bottom-up probing concepts, with however, in a first step, an emphasis on the latter. This first step entails a construction of individual value trees, derived from the results of interviews with representatives of stakeholder groups participating in the FRDO554. These individual value trees have been constructed on the basis of an interview protocol (see Annex 1) probing into questions such as what attributes or measures would be appropriate for differentiating between energy options, or asking why a specific alternative was considered to be ‘good’ or ‘bad’. In a first step, this approach has generated a list of criteria, at first without a concern for logical consistency or redundancy, but rather for completeness. To encourage interviewees to generate such a broad spectrum of criteria, at the time of the interviews we stressed that no attempt would be made to assess the relative importance of different values. Logical structuring was introduced only after the interview sessions. This process involved clarifying meanings, identifying means-ends relationships, eliminating redundancies, etc. – an account of which can be found in chapter 4 (Section 4). However, one important consequence of the bottom-up approach is that indicators of sustainable development are chosen and evaluated according to other criteria than when they are defined by a top-down approach. Criteria such as accuracy, standardisation and consistency will likely play a more limited role in the selection of indicators than can be expected from a top-down approach. As the upshot of ‘bottom-up’ participation and involvement, the creation and use of indicators of sustainable development serves as 553

Although of course Latour would resent the denomination of a ‘value tree’ (with the implication that ‘facts’ can be separated from ‘values’), we have chosen to keep this denomination in view of its simplicity of communication towards a broader audience. On a similar note, ‘evaluation criteria’ should be read as a translation of Latour’s ‘matters of concern’. 554 These are on file with the author.

358

Chapter 6

tangible evidence that the ‘members of the collectivity’ in question have asserted control over self-reflective definitions of the ‘good common world’, which likely includes attributed meanings beyond strictly quantitative indicators. This is an explicitly soughtafter quality, since it will likely enhance the (political) legitimacy of the value tree in the eyes of the participants in the discussion.

2.2

Construction of the combined value tree

No matter how expedient, this fundamental quality of the bottom-up approach is at the same time one of its main weaknesses. For if a list of criteria is to function as a ‘boundary object’ (towards the scientific world, international discourse on sustainable energy, etc.) measures of standardisation, coherence, etc. are precisely needed. Therefore, when combining the individual value trees into a combined tree, we have explicitly expanded the list in order to include criteria for any ‘holes’ that could be identified. For this purpose, we have used international reference documents (e.g. UNCSD 1996), overviews of existing indicator frameworks (e.g. Boonekamp 2002; Zuinen 2004) and official planning documents of Belgian policy (e.g. FPB 2002; ICDO 2000, 2004). The combined value tree has been discussed in a joint meeting555 in order to still ensure its validity towards the participants. In this workshop, we explained that we were not looking for a hierarchical ordering of values, but rather for the creation of a value tree where a) the relationship between the lower-level criteria and higher-level categories is one of inclusion; b) interdependencies between categories are avoided; and c) an exhaustive and non-redundant list of criteria is created. The latter requirement should however be qualified in view of the purposes of the exercise (i.e. a discussion on the long-term strategic orientations of the Belgian energy system): of course, certain topics could be split into much more detailed criteria and indicators (e.g. energy use over different economic sectors), but these subcriteria and indicators are more relevant for discussing policy measures in particular policy fields (and could therefore be assigned to the mandate of specific subsidiary institutions). In view of the three above-mentioned requirements, the group present in the workshop was encouraged to discuss the proposed tree with their colleagues, to provide comments for revisions, and, in particular, to add values that might have been left out. As a result of these discussions, inevitably changes had to be made to the proposed structure. The final combined value tree (included at the end of this chapter) has incorporated these changes. The combined value tree has been constructed on the basis of the insights gained from applying Boltanski and Thévenot’s commonwealth model (cf. Chapter 4 – Section 4.3). As explained in chapter 1, Boltanski and Thévenot’s model shows the interesting feature of being able to combine the ‘top-down’ (i.e. the development of coherent and intersubjectively plausible ethical structure) and ‘bottom-up’ (i.e. an ability to include all possible judgments of a particular actor) approaches to structuring ‘matters of concern’. The combined value tree starts from high-level categories (i.e. le principe du bien

555

Called a ‘scenario workshop’ (23 april 2003) – see Annex 1 for the attendance, and Keune et al. (2004) for an evaluation of the results.


359

commun), which cover concerns voiced in the language of the ‘industrial’ commonwealth (category I – ‘Environmental & human health and safety’ – i.e. mostly quantifiable impacts on the environment and/or human health and safety aspects), economic concerns mobilising ‘actants’ in the ‘industrial-market’ world (category II – ‘Economic welfare’) and criteria mostly stemming from the ‘civic commonwealth’ (category III – ‘Social, political, cultural and ethical needs’) – i.e. criteria having to do with governance (laws, rights, democracy, etc.) and institutions (civil society, media, etc.). Within each high-level category, equity concerns (i.e. le principe de commune dignité) were also covered556. For instance, in the ‘Environmental & human health and safety’ category, we included the concern for an equal distribution of risks and benefits – although this argument probably reveals more of a ‘civic’ concern for equal rights. Therefore, we must also admit however that sometimes we had to sacrifice the ‘purity’ of the different commonwealths in order to arrive at a somewhat workable result. ‘Aesthetic impacts’ for instance probably belong more in the ‘inspiration’ commonwealth; and the ‘need for longterm management of high-level waste’ is in itself a problem of such complexity that it probably involves arguments from all commonwealths (thus transcending the boundaries of a simple concern for radiological health impacts)557. We separated ‘Health and safety’ concerns into effects on individuals and effects that potentially threaten society (or even mankind). This separation allows a clearer accounting of the importance (for some) of catastrophic risks as expressed in the interviews. The second category ‘Economic welfare’ covers concerns about the costs, efficiency, security and market consequences of energy systems. We separated this category into the overall economic implications of the energy system on the market economy (e.g. security of supply, overall costs of the energy system, etc.) and issues of concern to more specific segments of the market economy (i.e. producers and consumers). This reflects a concern for distributional equity also clearly voiced in the interviews – i.e. fluctuations in energy prices, investments in new infrastructure, etc. affect different economic actors differently; hence it is desirable that the energy system puts a somewhat equal burden on different actors. The ‘Need for government intervention’ reflects the oft-raised objection that certain components of the energy system have been (or will be in the future) dependent for their success on government subsidies (see e.g. the objections raised against nuclear power from a historical point of view).

556

Other aspects of the ‘commonwealth grammar’ will become more clear in chapter 7, where the combined value tree is used in a multi-criteria mapping exercise. For instance, concerning equity, this exercise made a difference between mid-and long-term impacts, allowing users to make their assessment of intergenerational equity more explicit. The scoring performed in this exercise allowed for a measurement of ‘grandeur’ (principe de l’ordre de grandeur) according (in most cases) to a particular actor’s framing of the criterion in question. This also allowed issues of spatial equity (i.e. framing concerns on a local, regional or global level) to enter into the debate. 557 See Schrader-Frechette (1991) for a summary overview of the ethical dilemmas involved in radioactive waste management.

360

Chapter 6

The third category ‘Social, political, cultural and ethical needs’ was admittedly the most difficult one to assign criteria to. We broadly broke this category down into criteria regarding individual ‘civic’ development (in the sense of having the possibility to exert ‘sub-political’ influence on the shaping of the energy system – e.g. through consumer choices or direct citizen participation in decision making); institutional (or formal political) needs/benefits relate to the impacts of energy systems on democracy and its institutions, as well as the characteristics of the process by which energy decisions are made. The fourth category ‘Diversification’ will be discussed under section 3.4.3.4. Most of the labels in the combined value tree are, we believe, reasonably selfexplanatory; for others, we gave some clarification. We discuss the qualitative criteria (i.e. criteria for which at this stage we had not yet defined quantitative measurement) here; quantitative criteria are discussed under section 3.4.3: • Economic risks: fluctuations in energy prices or availability might affect economic development in the sense that abrupt price changes not only have direct costs, but also lead to perturbations of the economy which generate indirect costs. This criterion refers to these indirect costs; • Ability to provide specialist market: this criterion reflects the perspective that some developments in the energy system might be more conducive to the creation of new market opportunities, e.g. delivering energy services instead of energy as a commodity, highly efficient equipment, etc. • Strategic factors for export: contrary to the previous criterion, this one stresses the position of Belgian industry in the global market, i.e. certain developments in the energy system might allow certain economic sectors to acquire a leading position on the global market (possibly also contributing to the ‘Belgian image’); • Consumer choice: reflects a concern for the possibility of exerting ‘sub-political’ action through consumer behaviour – i.e. having a free choice of energy products or services from a variety of providers; • Citizen participation: reflects a concern for the possibility of exerting ‘formal’ democratic influence on the shaping of the energy system; e.g. certain developments would score better if the parliamentary influence is perceived to be higher, or if more political parties are actively involved in the energy debate (political pluralism), or if the influence of corporatist decision making is mitigated, etc.558; • Contribution to rational energy use559: reflects the concern that a focus on certain technological solutions to the problems raised by the energy system (e.g. end-of-pipe

558

International comparative measures of democracy already exist, see e.g. the ‘Freedom House index of political rights and civil liberties’ (available at http://www.freedomhouse.org/research ). Such indices could serve as a basis for the more specific energy policy context. 559 In countries with a high share of nuclear power production (e.g. Belgium and France), there is always the risk of attempts to promote electricity consumption in order to facilitate nuclear power’s penetration into the market. For instance in the ‘70s, Belgian and French utilities took intensive commercial actions to promote the thermal applications of electricity (Laes et al. 2004c; Romerio 2005).


•

•

•

•

361

solutions, nuclear power) diverts the political attention from the need for a more rational use of energy (i.e. from this point of view, a lock-in into a sub-optimal solution occurs); Control and concentration of power: this criterion reflects the concern that large energy companies might control the market through the use of market power (cf. London Economics (2004)), thus disrupting the functioning of the ‘pure’ market commonwealth; Influence on political decision making: in addition to the previous criterion, this one reflects the concern that market power might also be translated into political power (e.g. through lobbying); Need for socio-political stability: reflects the point of view that certain components of the energy system (e.g. nuclear power) can only function properly if enough stability is ensured. As such, socio-political stability is related to both social and formal political institutions. Institutional stability needs to be checked together with the degree of political democracy: dictatures can provide for a large amount of stability, but are not accepted as being generally conducive to sustainable development. Institutional stability could for instance be made operational through indicators of social justice, degree and evolution of economic development, perceived legitimacy of government initiatives (e.g. demonstrations, frequency of government changes), freedom of the press and external factors (degree of regional integration, relationships with neighbouring countries, etc.); Need for direct political intervention: this criterion describes the amount of effort exercised by government to influence the energy system evolution. Indicators could include the amount of government expenditures on energy policy compared to GDP, the fraction of energy use falling under benchmarking agreements, the fraction of taxes in energy prices or different sectors, etc. (Boonekamp 2002). It is likely that this criterion will be perceived differently by different actors – i.e. some will consider direct political intervention a necessity, others will adhere more to the rhetoric of the free market.

All in all, we have tried to utilise ‘substructures’ of individual value trees to the maximum extent possible in the combined value tree in order to maintain the logic and content of most people’s reasoning. Subsuming these ‘substructures’ under a generalised perspective ensures the comprehensiveness of the effort and principally allows a confrontation of viewpoints (e.g. when using it as a multi-criteria mapping tool – cf. Chapter 7). While of course the limited number of interviews and the individual perspectives of the interview participants (and inevitably also our own biases) make it doubtful that the tree represents the views of ‘Belgian society’, we feel that it is reasonably comprehensive. Nevertheless, our combined value tree should be regarded as a first attempt: if new (or other) entrants in the debate do not agree on its logical structuring or completeness, it can be used as a starting point for improvements. Also, as mentioned before, many low-level criteria can probably be split into further sub-criteria, in view of proposals for measurement. Furthermore, we could not exclude that the understanding of the different criteria still differed somewhat from one participant to another. Whether there still was ‘sufficient’ common understanding of the criteria to function properly is an issue we will explore further in chapter 7.

362

Chapter 6

3 Sustainable energy scenarios

3.1

Introduction

The second step proposed in chapter 5 was to use the list of ‘actants’ in order to draw up scenarios of possible futures (by assigning a ‘role’ to each of the actants and/or excluding others), in order to enable a strategic discussion about the long-term orientation of the energy system. Such scenarios also function as ‘boundary object’, hence the need for different lines of accountability to different ‘worlds’. For instance, scenarios will need to be ‘realistic’ enough in order to convince scientific experts; their scope will have to be broad enough in order to address key concerns of different stakeholders; they will have to be simple and transparent enough in order to promote understanding in different groups which are not necessarily specialists in scenario techniques, etc. Thus, these requirements for long-term scenario explorations being established, one could wonder what type of questions should be addressed in such a study (the ‘content’), and how they should be addressed (the ‘approach’). Long-term energy scenarios on a global scale have already been developed by institutes such as the IIASA (1995) and the IPCC (2001). Key issues such as growth in world population, economic growth, depletion of fossil resources, and the risks linked to global warming are discussed and should of course also be addressed here (population growth being less important on a more local or regional scale). In the Belgian context however, most effort is invested in finding nearterm solutions in order to comply with the obligations contained in the Kyoto protocol560. Demand-side measures, cogeneration and renewable energy are proposed as solutions, although other (technical) options for GHG emission reductions, such as photovoltaic power, fuel cells, and carbon capture and storage could show significant potential but will not capture the market at short or mid-term notice. Also, technical breakthroughs in nuclear electricity generation addressing issues such as safety and risks of proliferation could be possible on a longer timescale. As greenhouse gas emission reduction targets will most probably by tightened over the coming decades, any long term option commands careful consideration. Concerning the approach, reports by the Federal Planning Bureau (Gusbin and Hoornaert 2004) and the modelling exercises reported in CES (2001) could qualify as exemplary foresight exercises on a mid-long term (horizon 2020-2030). However, for the present study (whose main point of departure lies in the explicit consideration of value perspectives), we have chosen a somewhat different approach. We recall here the distinction we have introduced in chapter 5 (Section 3.2.1.) between types of scenario exercises which are meant to be primarily descriptive or exploratory scenarios, i.e.

560

This statement should now be qualified somewhat in view of a recently published report containing four long-term scenarios (horizon 2050) aimed at giving an input to the Belgian position in the upcoming postKyoto negotiations (Econotec and Vito 2005).


363

scenarios describing possible developments starting from what we know about current conditions and trends, and primarily normative, anticipatory or backcasting scenarios, i.e. scenarios which are constructed to lead to a future with a specific subjective value defined by the scenario developers. As explained, they differ in terms of overall purpose. The choice between exploratory and anticipatory approaches depends on the objectives of the scenario development exercise. Anticipatory scenarios represent organised attempts at evaluating the feasibility and consequences of trying to achieve certain desired outcomes (or avoid the risks of undesirable ones). Exploratory scenarios (or ‘what-if’ analysis), on the other hand, try to articulate different plausible future outcomes, and explore their consequences. The emphasis is mostly on prioritising technological choices, the analysis is performed in a relatively closed process by technically or economically schooled experts, and the government (or administrative bodies) mostly assumes the role of client. The above-mentioned examples seem to fall mostly in this category. For our purposes, a combination of backcasting from a set of ‘minimal’ objectives and foresight from initial conditions and drivers has been chosen. However, different transition routes (based on major technological options) towards achievement of the desired end-state have been modelled. It is important to be clear from the outset that any consideration of prospects over a 50 year timescale must be very uncertain. Our projections and technology assessments will inevitably turn out to be inaccurate. But this does not invalidate the exercise. Our approach should primarily be viewed as an approach to bring underlying value perspectives into sharper focus, and hence, to promote a more informed debate. Uncertainty is unavoidable and should be factored in the analysis. The scenarios we have developed for our exercise have been discussed and modified in a ‘scenario workshop’ (for attendance, see Annex 1) in order to ensure the general transparency of the effort. The practical multi-criteria exercise we have carried out subsequently (cf. Chapter 7) allowed us to test this hypothesis further. 3.1.1


We have chosen to follow the backcasting methodology as described by Anderson (2001). Anderson also points at some significant advantages of backcasting when considering policy planning for sustainability: the explicit consideration of value perspectives (whereas traditional foresight is less candid) and the possibility to formulate a strategic normative vision on long-term technological options (whereas traditional foresight focuses more on short-term market forces). The methodology distinguishes six distinct steps or building blocks of scenario development: 1. definition of strategic goals; 2. description of present-day supply and demand structure; 3. choice of end year; 4. demand-side analysis; 5. supply-side analysis; 6. development of a set of policy measures to meet the strategic objectives.

364

Chapter 6

How these ‘building blocks’ have been used to define the energy scenarios in the present report will be the subject of the following sections. 3.1.1.1 Definition of strategic goals The main goal of the present chapter is of course to identify possible scenarios (or strategic options) for a sustainable energy future. We have based our judgement on what is considered to be the present-day ‘minimal consent’ on the concept of a sustainable energy system. That is, some of the criteria described in section 2 were given priority over others as ‘meta-criteria’ which must be met by any sustainable energy scenario. Hence, none of the scenarios described here can claim to represent a definitive vision of sustainability – all scenarios involve (often difficult) tradeoffs. The choice of ‘meta-criteria’ is an unavoidably subjective one; therefore, it must be well documented. We have been guided here by the following considerations: • The ‘minimal’ objectives should be in line with strategic objectives as set out by official policy documents. The most notable exception is of course the government decision to phase out existing nuclear power plants, since we want to discuss the possible future of nuclear energy in the first place; • Energy should in any case remain reliable and affordable. Our choice of an economic optimisation model (MARKAL, cf. infra) guarantees the choice of a least-cost energy system (under different constraints and economic assumptions). However, with this model, it is impossible to assess the reliability of the energy system (in terms of forced outages, matching demand and supply, quality of electric power, etc.). Therefore, for our present purposes the reliability of the energy system simply has to be assumed, pending more detailed analysis with the aid of electricity sector models; • The ‘minimal’ objectives should be ‘politically stable’ on the long term; • The ‘minimal’ objectives should of course be internally consistent (i.e. noncontradictory). Furthermore, as a small nation, Belgium will of course be very dependent on evolutions in the international energy policy scene. Therefore, we also have to assume that these international conditions will be in line with the goals set out in the Belgian context. Thus, these considerations led us to adopt overriding criteria in the domains of security of energy supply, economy and ecology. These criteria will be developed further under section 3.3.1. 3.1.1.2 Description of the present-day supply and demand structure The so-called initial conditions of the energy system: • Technical, economic and environmental parameters for the different technologies making up the energy system; • An adequate description of the entire energy chain: from the use of primary energy (natural gas, oil, coal, natural uranium, or renewable energy) through secondary energy


365

carriers (fuels, electricity, etc.) to the provision of energy services (transportation, space heating, lighting, etc.). 3.1.1.3 Choice of end year Pragmatic considerations have led us to the adoption of a 50 year time horizon for our scenarios: this allows for a broad range of structural and technological changes to come into play to meet the requirements set out above, whilst not being too remote in order to influence present-day decision making. 3.1.1.4 Demand-side analysis For this issue, Anderson (2001) proposes to: • Use an exogenous estimation of the demand for energy services, common to all scenarios under consideration. This is done in order to make the scenarios more comparable on an equal basis. The rationale is that deliberate policy measures will have less impact on the actual demand for energy services rather than on the ways these demands are met; or, in any case, that the demand for energy services will be very difficult to predict (and hence difficult to manipulate by deliberate government intervention), especially on a long time scale and on a national scale; • The resulting demand for final and primary energy should however be determined endogenously (within the model); • The demand-side analysis should proceed from an evaluation of the state-of-the art technological options to provide for a certain energy service, and an evaluation of the policy instruments necessary to introduce these technologies into the market. In this chapter, the baseline demand for energy services in different sectors (industry, commercial, households, transportation) has been determined with the aid of a general equilibrium macroeconomic model for the 15 countries of the EU (the GEM-E3 model) for 2030 by the ‘Centre for Economic Studies’ (CES, KULeuven). For 2050, results have been extrapolated (and sometimes reviewed by the author). The drawback to this approach is that shifts in the economic activity of industrial sectors (e.g. a shift to an economy with increased recycling of used materials, shift to less energy-intensive sectors such as services), or shifts in level of energy services demanded by consumers (e.g. conscious efforts to reduce the number of kilometres driven by car, lower indoor temperature, etc.) are not explicitly addressed (econometric modelling implicitly assumes that historic trends, expressed as statistical correlations, will still prevail in the future). Nevertheless such profound structural changes are often demanded by stakeholders and are seen as a precondition for sustainability. Some studies do try to accommodate for these more profound structural changes in society. For instance, the UK White Paper (2003) uses key assumptions of four qualitative storylines about the future (with variable rates of GDP growth, population and household numbers) to derive an estimation of the potential gap in CO2 emissions that has to be

366

Chapter 6

bridged in 2050 by policy actions561. However, this approach does not allow scenarios to be compared on an equal basis. Furthermore, it can be argued that our approach is in line with a precautionary view on the problem at hand: the derived demand estimates for energy-intensive services (based on historical evidence) will most likely be at the higher end of other possible estimates, especially if the world economy is confronted with ever more stringent carbon constraints. Regarding the demand-side analysis, it is difficult to assess the scope for energy efficiency measures on a 50 year time scale on a similar basis as for energy generation options. This is linked to the number of technologies involved, new technologies and alternative interpretations of unrealised but apparently cost-effective potentials. We limited ourselves to the existing technical potential (taken from diverse sources). This technical potential has been considered to be equal for most technologies for all scenarios. However, the level of uptake of energy efficient technologies was varied across two broad scenario groups (cf. Section 3.2.4). 3.1.1.5 Supply-side analysis The supply-side analysis includes: • An analysis of state-of-the-art technologies for energy ‘production’. In our scenarios,

we only retained technologically proven options, rather than some options which are now only in a conceptual phase (e.g. fusion, some fission technologies); • An analysis of the technical, economic and environmental parameters of these technologies, and possible constraints to the development of the technology under consideration. Of course, there is a range of uncertainties attached to various costs. To accommodate for this inevitable uncertainty, we have drawn from a range of sources (cf. infra). In a first step, and as part of our research strategy to build the strongest possible case for different value positions, some worldviews (cf. Section 3.2.4) have built in ‘optimistic’ estimates for certain technologies – e.g. the lower cost estimates and less stringent constraints were used for distributed electricity generation options in the worldview called ‘Rational Perspective’. Next, a range of cost estimates can be used in detailed sensitivity analyses. Again, our approach is motivated by reasons of transparency: for instance, some models incorporate ‘learning curves’ which endogenise technological innovation. However, it seems that a ‘grand theory’ to give robust forecasts of technological development is not available. All one can do then is at least to strive for consistency and transparency.

561

World markets (a world based on individual consumerist values, a high degree of globalisation and scant regard for the environment); global sustainability (based on predominance of social and ecological values, strong collective environmental action and globalisation of governance systems); provincial enterprise (based on individual consumerist values, reinforced governance systems at national and sub-national level) and local stewardship (based on communitarian and strong conservation values, diverse political systems and economic regionalisation).


367

3.1.1.6 Policy measures to meet the strategic objectives The last building block entails the development of a coherent and internally consistent package of policy measures in order to reach the strategic objectives as defined in section 3.1.1.1. 3.1.2

Choice of energy model

MARKAL is a generic model that represents all energy demand and supply activities and technologies for a country with a horizon of up to 50 years. Here, it is used as a technicaleconomic bottom-up model which assembles in a simple but economically consistent way technological information (conversion-efficiency, investment and variable costs, etc.). As the model is formulated as a dynamic optimisation model, it can produce alternative developments for energy supply and demand options that satisfy a certain given level of demand for energy services, under certain constraints (for instance a CO2 emission reduction goal), at least cost. Simultaneously, the model makes prospective energy and emission balances, tests the potential of new energy technologies and can contribute to R&D policy formulation. One of the main advantages is that results are easily verifiable: they can immediately be related to assumptions regarding technological data and economic parameters. This approach is thus more suited to our purpose of ‘mapping’ different technological options (fuel choices by users, energy efficiency options adopted, energy supply options chosen, etc.). The main disadvantage is that economic processes in energy markets and the behaviour of economic actors are not modelled (demand for energy services is determined as an exogenous model variable). This can for instance be done with national general equilibrium models (e.g. GEM-E3 model of the EU): these are economic models which allow for instance to evaluate the macroeconomic impact of a CO2 tax. These models can study such questions as the use of the revenue from a CO2 tax, the double dividend discussion, the total impact on employment, etc. and deliver a basic forecast for the demand for energy services (an input for MARKAL). Table 12 summarises the strengths and weaknesses of different modelling approaches in the representation of energy use in a national economy.

368

Chapter 6

Driving forces for emissions Economic

Economic

Level of

Energy

Fuel choice

Final

Energy sector

Primary

activity

activity per

energy

efficiency

by user

energy use

transformation,

energy use

level

sector

services per

choices by

transport and

user

user

distribution

Reduction possibilities for GHG emissions (excluding carbon sinks) Lower level

Switch to

Less energy

More

Substitute

Reduce losses

of economic

less energy

intensive

energy

between

and use less

growth

intensive

production

efficient

coal, oil, gas

carbon-

activities

in industry

processes

and

intensive fuels

renewables

in electricity production

Lower

Better

indoor

insulation,

between

temp., less

more

coal, oil, gas

km. driven

Substitute

efficient

and

appliances

renewables

Modelling domain GEM-E3 model Yes

18 sectors

Implicit and joint

Simple

Simple

Simple

Simple

Modelling domain of MARKAL model Constant

Implicit

39

Detail

Categories

Table 12. Strengths and weaknesses of different modelling approaches to the representation of carbon emissions and energy use of a national economy (Source: Proost et al. 2000)

3.1.3

Limitations of the modelling approach

To summarise, with this scenario exercise, we can address the impact of different strategic objectives for the Belgian energy sector on the following issues: • • • •

The evolution of the demand for different energy vectors (electricity, fuels, gas, coal); The evolution of the demand for primary energy; The technological evolution of the electricity generation sector; A comparison between scenarios of the evolution of the production cost of electricity and the total energy system cost (this should however not be viewed as an absolute cost – to calculate this, econometric modelling approaches are more suited); • The contribution of different sectors to the CO2 emissions.


369

Subject to the following limitations, since some important aspects are not considered in this analysis: • The opening of the electricity market and the impact it can have on the production and the investment in the electricity sector is not modelled in detail – in fact we have either explicitly assumed (in 6 of the scenarios studied) that transmission constraints will not make intensive trade possible (hence, a negligible amount of electricity is imported), or either that Belgium will increasingly rely on net electricity imports in the future (in the remaining 2 scenario cases); • The possibility of reaching greenhouse gas reduction targets through international actions (tradable permits, CDM projects) is not considered. However, it can be argued that on the longer term, a 30% reduction target (as proposed here) by domestic measures is not exaggerated (cf. supra – further sensitivity analysis includes a 50% reduction target in 2050); • The overall effects of stringent carbon constraints on the Belgian economy (e.g. effects on economic activity of different sectors, employment effects, shifts in demand for certain energy services) cannot be studied – however, this limitation is rather linked to the inherent uncertainty of any econometric modelling on a (very) long time scale; • The MARKAL model is an energy sector model and is not appropriate to study the contribution of decentralised power to the overall reliability of power supply; a model specific for the electricity sector would be more appropriate to study these issues. This is particularly relevant, since generally, it can be said that the smaller the centrally controlled supply system is and the more the power generation is stochastic (e.g. wind and solar energy), the more difficult it gets to manage the power supply system and to maintain a satisfactory voltage level.

3.2

Background information on framing assumptions

Here, we give a brief introduction to the scenarios which will be developed in the following sections. A scenario consists of a particular combination of ‘framing assumptions’ and ‘building blocks’. Framing assumptions describe the external factors (the ‘environment’) influencing the Belgian energy system. In particular, we defined two worldviews – the ‘Rational Perspective’ (letter code R) and the ‘Market Drive’ worldview (letter code M) (cf. Section 3.2.4). Within each worldview, there is one ‘baseline’ scenario, which will be used for economic cost comparisons. The baseline scenario does not comply with (post-)Kyoto requirements and allows all energy production technologies. All other scenarios will have to comply with post-Kyoto commitments: -7.5% reduction of GHG emissions (compared to 1990 levels) in 2010, -15% in 2030, - 30% in 2050 (letter code K). Further sensitivity analysis with more stringent post-Kyoto commitments has also been conducted for this report: - 7.5% in 2010, -30% in 2030, -50% in 2050 (letter code KK).

370

Chapter 6

Building blocks describe the internal structure of the Belgian energy system (strategic objectives, energy saving potentials, supply options, energy policy). In particular (for this exercise), we focus on the technological supply options. We suggest four broadly defined options. For analytic simplicity, it is assumed that these major options will be pursued (almost) to the exclusion of the other options: 1. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity (letter code P.CS – Phase out & Carbon Storage); 2. phase out of existing nuclear power plants (40 yr. lifetime), new nuclear build allowed, limited possibilities for carbon capture and storage technology, limited import of electricity (letter code LCS – Low Carbon Storage); 3. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity (letter code P.LCS – Phase out & Low Carbon Storage); 4. phase out of existing nuclear power plants (40 yr. lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity (letter code P.LCS.I – Phase out, Low Carbon Storage & Import of electricity). In principle then, the combination of framing assumptions and technological options leads to 8 possible scenarios: M.K.P.CS, M.K.LCS, M.K.P.LCS, M.K.P.LCS.I (and a ‘Market Driven’ baseline scenario) and R.K.P.CS, R.K.LCS, R.K.P.LCS, R.K.P.LCS.I (and a ‘Rational perspective’ baseline scenario). 3.2.1

Demographics

Demographic developments influence the demand for energy services in a number of ways. Firstly, the number of households is one of the key determinants of residential energy consumption, since this parameter determines the number of household appliances in use and the specific surface that has to be heated. Secondly, the number of people influences the use of transportation services. The evolution of the population is taken from NIS-FPB (1996). In this report, Belgian population was 9.97 million in 1990, and was projected to reach 10.3 million in 2030. Actually, according to the latest data, Belgium already counted 10.3 million inhabitants in 2000, and is projected to count 10.9 million in 2050562. However, in view of the much larger uncertainties involved in long-term energy modelling, we have chosen to continue working with the original figures, rather than updating the entire existing database.

562

http://www.statbel.fgov.be/figures/d23_nl.asp#2


households population

1990 4.00 9.97

2000 4.27 10.21

2010 4.53 10.33

2020 4.61 10.34

2030 4.68 10.30

2040 4.68 10.30

371

2050 4.68 10.30

Table 13. Number of households and population (millions)

The number of persons per household is based on the reference scenario in NIS–FPB (1997). It goes from an average of 2.49 persons per household in 1990 to 2.28 in 2011. The trend was extrapolated until 2030 assuming the declining trend would continue, though at a slower pace (-0.02 persons per household every five years). For the period 2030-2050, we assumed that the trend would level off at 2.20 persons per household. These estimates still correspond well with the latest figures of an average of 2.35 persons per household in 2002563. 3.2.2

Assumptions for energy service demands

The demand of energy services differs from the final energy demand: the demand of energy services corresponds to the demand for heat in houses or heat with certain characteristics (temperature, pressure) industrial processes or the demand of vehicle-km. in case of transportation, whereas the final energy demand corresponds to the delivery of energy products to the consumers (Figure 7). Final energy is one of the inputs into the production of energy services, other inputs are e.g. heating equipment or house insulation. As explained in the methodological section, a macro-economic activity evolution model (GEM-E3) was used by the CES to determine the shift in the demand (curves) of energy services until 2030. In the industrial and service sectors, the demand function shifts at the same rate as the production or the value added of these sectors, taking into account the evolution of the relative energy service price and technical progress. For the households, the demand function shifts as a function of the evolution of income and relative energy prices, with an income elasticity of 0.3 for heating demand, 0.5 for hot water and cooking demand and 1 for specific electricity demand and a price elasticity of -0.3 for all categories of demand. For the transportation sector, passenger transport is a function of income whereas freight transport is a function of the general activity level, with a price elasticity of -0.3. The derived evolution of the demand for energy services used in both scenario groups is summarised in the table below. For the period 2030-2050, we made linear extrapolations of these projections, except for space heating (residential), warm water use (residential) and private car use, where we assumed that the demand would be saturated after 2030564. 563

http://www.statbel.fgov.be/figures/d24_nl.asp#3 For space heating, the unit consumption (i..e. final energy use per dwelling, with climate corrections) has remained almost constant over the period 1990-1998 (Odyssee Database, April 2000). Combining this with our demographic assumptions for the number of households (and hence, dwellings), the no-growth assumption after 2030 seems reasonable. For private car use, the no-growth assumption is based on the saturation effect of transportation infrastructure. The FPB (2001) assumes a maximum of 20,500 km/person.yr. In the existing MARKAL database, private car use already reaches the level of 21,200 km/person.yr. in 2030, so the nogrowth assumption for the period 2030-2050 seems justified.

564

372

Chapter 6

International energy prices

Other exogenous assumptions: - Overall technological progress - Overall growth expectations

GEM-E3 produces baseline levels of economic activity

Levels of energy services required in reference scenario

MARKAL

Energy efficiency use

Input MARKAL

Exogenous assumptions on technological development and cost of alternatives

output

Figure 7. Determination of the demand for energy services (Source: CES 2001)

Since the evolution of the long-term demand for energy services is inherently very uncertain, we have also included a sensitivity analysis with lower demand in industrial and transportation sectors (-30 % compared to the 2050 demand levels for all categories in both sectors in Table 14, with linear interpolation in the period 2000-2050). This could for instance correspond to a society with greater relative emphasis on services (compared to industrial production) and innovative solutions for the growth in demand for transportation (e.g. tele-working, carpooling, spatial planning, etc.). Information on demand for energy services is included in Annex 2.


2000-2010

2010-2030

2030-2050

Building materials (kton / yr.)

0.4

0.4

0.5

Glass (kton / yr.)

0.4

0.4

0.4

Limestone (kton / yr.)

0.4

0.4

0.4

Electrical motors (PJ)

0.8

0.6

0.6

Ammonia synth. (PJ)

0.5

0.5

0.4

Chemical industry

0.4

0.6

0.5

Chlorine (PJ)

0.5

0.6

0.6

Iron & steel prod. (kton / yr.)

0.2

0.1

0.0

Iron & steel proc. (kton / yr.)

0.2

0.1

0.0

Other industry (PJ)

0.9

0.9

0.9

Lighting (PJ)

1.0

0.8

0.8

Space heating (PJ)

0.9

0.9

0.8

Warm water (PJ)

0.9

1.0

0.8


1.3

1.3

1.0

Space heating (PJ)

0.6

0.7

0.0

Warm water (PJ)

0.8

0.9

0.0


1.9

2.0

1.6

Private car use (billion veh. km.)

2.5

1.9

0.0

Rail (million veh. km.)

1.8

1.7

1.3

Freight (truck) (billion veh. km.)

1.6

1.4

1.1

373

Industrial sector

(steam + process heat) (PJ)

Service sector

Residential sector

Transportation sector

Table 14. Evolution of the demand for energy services – all scenarios (% average annual growth)

3.2.3

Resource availability and energy prices

The world’s oil, gas and coal resources are a key factor for future energy supply. Price levels are very uncertain and depend on a number of factors: cost of energy resource

374

Chapter 6

technologies, worldwide demand for fossil energy sources and strategic behaviour of exporting countries. As explained above, we assume in our scenarios that all fossil fuels will remain physically available over the next 50 years. For our assumptions regarding energy prices, we made use of two distinct (and rather extreme) prospects (Charpin et al. 2000)565. The average of these two prospects is taken as the standard value for energy prices in all scenarios (Table 15). Furthermore, a sensitivity analysis has been conducted by making use of the higher price estimates. All prices are net prices at the Belgian border, i.e. without additional taxes or duties. Price levels for 1990-2000 have been calibrated to the historic values. The price of gas is assumed to remain tied to oil product prices. The price of hard coal is supposed to remain relatively constant over the entire time horizon: coal resources are amply available and will certainly not be exhausted in the next century. Also, the uranium price is supposed to remain constant over the entire 50 year period. We have assumed a value of 3.3 MEuro/ton uranium, including the costs of mining the natural uranium and fabrication and delivery of fuel elements to power plants. Even assuming a rise in resource prices, the total nuclear fuel cycle costs would not rise significantly, since uranium is only a small fraction of the fuel cycle cost. In any case, this price effect will be very small compared to for instance uncertainties on the investment cost for new nuclear energy generation. The import price of electricity (again, net of taxes) corresponds to (in more familiar units) 4 Eurocent/kWh in 2000 and 10 Eurocent/kWh in 2050. No profit margin for the producer is included in this price. Cost (MEuro1995 / PJ)

1990

2000

2010

2030

2050

Crude oil High Sulphur Heavy Distillate Low Sulphur Heavy Distillate Gasoline Light Distilate Oil LPG Natural Gas Coal (non-residential use) Import of Electricity

3.12 2.21 2.53 4.88 4.21 3.07 2.48 1.34 10.33

2.42 1.72 1.96 3.59 3.20 2.76 1.69 1.16 11.81

3.11 2.52 2.88 4.86 4.55 3.07 2.47 1.19 14.76

4.75 3.86 4.40 7.28 6.92 3.93 3.41 1.24 20.66

6.40 5.19 5.92 9.69 9.28 4.78 4.35 1.28 26.56

Table 15. Resource prices for all scenarios

3.2.4

Discount rates

For the discount rates, we have distinguished two alternative ‘worldviews’ (inspired by the approach outlined in Lako et al. (1998)). Both worldviews should be considered to be 565

Resource prices for crude oil might seem to be rather low at first sight – e.g. crude oil is expected to cost about 32 $ per barrel in 2030 (in $ of 2000). However, according to the ‘International Energy Agency’ (IEA 2004), it is assumed that the oil prices will drop down to about 22 $ per barrel in 2006 (measured in $ of the year 2000). The price would then remain flat until about 2010 after which it would increase linearly till about 29 $ in 2030 (again in $ of 2000). This is only an assumption, but it shows that the energy analysts – ignoring temporary price fluctuations – expect that oil will remain quite affordable over the next 25 years.


375

‘technically’ possible – thus answering the demands of the ‘industrial commonwealth’ in Boltanski and Thévenot’s words. The ‘Rational Perspective’ worldview can be characterised as being more driven by ‘civic’ concerns; while in the ‘Market Driven’ perspective, these concerns take a backseat towards those of the ‘market commonwealth’. It is assumed that the process of global economic integration will lead to more collective public action. Strong penetration of new, more efficient demand and supply technologies (also at the end-use level) is facilitated. This strong penetration can for instance be achieved by setting efficiency standards, removing existing barriers for the introduction of efficient and decentralised technologies, and active energy service companies which carry out cost-effective efficiency improvements for third parties. The two distinctive storylines can be described as follows: In the ‘Market Drive’ worldview, the market mechanism is seen as the best way to generate prosperity and handle uncertainty. The penetration of new, more efficient demand and supply technologies is seen to depend more on market forces and the behaviour of actors. The environmental protection agenda is also more influenced by market actors than by public policy. Moreover, energy policy is driven by the desire to minimise government control and to maximise efficient operation of free markets. Governments will only intervene in issues of overriding importance: e.g. securing diversity of energy supplies or enforcing internationally binding agreements (e.g. under the UNFCCC). Barriers will persist in the uptake of efficient equipment. Efficiency gains will only be made for competitive reasons. The ‘Market Drive’ worldview assumes that in reality more stringent investment criteria apply for many energy related decisions and that hidden costs and market barriers do play a role. This scenario group assumes different discount rates per type of sector. The discount rates reflect representative hurdle rates applicable to that sector or kind of end-use (cf. Table 16). In the ‘Rational Perspective’ worldview, a single discount rate is applied. By applying a uniform rate, all technologies at the demand side of the energy system are allowed to compete with energy supply technologies like in a perfect market. RP assumes a market that works ‘rationally’ without barriers and with perfect information so that any difference in pay back opportunities will automatically be removed. The discount rate in RP amounts to 5% per year, which is typical of a low-risk investment climate. The above-mentioned general perspectives have been modelled (very schematically) by varying the discount rate across the scenario groups. A discount rate is required to annualise the capital cost in order to compare the costs of alternative technologies with different ratios of initial capital expenditure to annual running costs. The formulation of the MARKAL model allows choosing one uniform discount rate applied to all technologies or different discount rates applied to the different sectors.

376

Chapter 6 Sector

Discount rate

Power generation

8%

Industrial cogeneration, refineries, biomass conversion, all processes in industry

10%

Space and water heating in residential sector and commercial sector

15%

Trucks, vans and buses Electric appliances

20%

Passenger cars

25%

Table 16. Sector specific discount rates in ‚Market Drive’ scenario group

In addition to the differences between the worldviews relating to their perspective towards decision criteria on energy investments by individual actors (represented by different discount rates (Lako et al. 1998)), assumptions regarding the available technical potential of energy saving technologies and regarding cost and constraints on technologies for electricity generation were varied in line with the overall worldview under consideration (cf. Table 17) . Rational Perspective (RP)

Market Drive (MD)

Decision criteria

Uniform 5% discount rate for all energy decisions across all sectors

8 % discount rate for power generation, higher discount rates for end use

Technical potential energy saving

In some sectors higher than in MD

In some sectors lower than in RP

Technologies for electricity generation

Favourable towards decentralised electricity generation options

Favourable towards centralised electricity generation options

Table 17. Key differences between ‚Rational Perspective’ and ‚Market Drive’ scenario


3.3

3.3.1

377

Using the ‘building blocks’ and ‘framing assumptions’ to construct scenarios

Strategic goals

Overriding strategic goals (which apply to all scenarios) are: Security of energy supply • On a world-wide scale, it is assumed that sudden shortages of energy supply (energy crises) can be avoided – we do however take into account the possible consequences of a gradual depletion of fossil energy resources on energy prices; • On the level of the individual consumer (in Belgium), it is assumed that energy supply will be reliably, flexibly and qualitatively tailored to the demand for energy services as required by (at least) present-day demands for comfort (however, this does not imply that the same level of end-use energy demand will be sustained !); • On the level of the individual consumer (in Belgium), it is assumed that social energy provisions (social tariffs, minimal provision of energy, etc.) can be maintained. Economic criteria • It is assumed that worldwide ratio of economically proven (fossil or uranium) reserves (R [PJ]) over the production of primary energy carriers based on these reserves (P [PJ/a]) will allow for enough time (e.g. 20 years?) for a transition to an energy system based on virtually inexhaustible energy resources (e.g. wind, sun, fusion) when needed (beyond the time horizon under consideration); • The intensity of energy use (energy use per unit of GDP) should not increase in any scenario. Ecological criteria • SO2, NOx and NMVOC emission levels will have to be limited (the Göteborg Protocol emission reduction targets can be used as a guideline) ; • Greenhouse gas emission levels will be significantly reduced through follow-ups to the Kyoto Protocol. All ‘sustainable’ scenarios will have to comply with reduction goals of 7.5% in 2010 (Kyoto requirements), 15% in 2030 and 30% in 2050. These requirements are not very ambitious compared to long-term visions developed for other countries: the UK White Paper (2003) sets out a 60% reduction goal based on the principle of ‘contraction and convergence’ (in line with a potential global agreement which would set an upper limit of 550 ppm for the carbon dioxide concentration), while the Wuppertal Institute (Fischedick et al. 2002) even aims at an 80% emission reduction goal for Germany, both on a time scale of 50 years. However, we did investigate further reduction targets (-30% in 2030, -50% in 2050) in our sensitivity analyses. Furthermore, we only study energy-related emissions. These however account for 87% of the total Belgian GHG emission level in 1990 (Proost et al. 2000).

378

3.3.2

Chapter 6

Demand side analysis

3.3.2.1 Industry As explained above, the demand for energy services was kept identical for both scenario groups. However, the technical potential for energy savings in different sectors was sometimes varied in both scenario groups. We have drawn from a range of sources, including the original MARKAL database (described in CES (2001)), WEC (2000), and Fischedick et al. (2002)). We have chosen to limit ourselves to the proven existing technological potential, rather than simply assuming the continuation of certain trends in energy productivity (energy end use/energy service) over the entire time horizon (as is the case in the UK White Paper for instance). Technical potential


Market Drive

Cement

20%

20%

Construction materials

20%

20%

Glass flat

15%

15%

Glass hollow

37.5%

37.5%

Glass fibres

27.5%

27.5%

Lime

10%

10%

Industry - power

35%

20%

Industry electro thermal

20%

20%

Industry lighting

50%

20%

0-25%

0-25%

Steam (chemical industry)

20%

20%

Process heat (chemical industry)

10%

10%

Chlorine electrolysis

10%

10%

-

-

Iron & steel processing

20%

10%

Low temp. steam & heat

10%

10%

10%

10%

energy savings

Ammonia synthesis (gradual increase till 2050)

Steel production

(other ind.) Other industry

Table 18. Technical energy saving potential (industry)


379

In MARKAL, for each sector at least one conventional end use technology is modelled which represents the present mix of installations. In addition, one or more alternative processes or saving options are provided. These saving options can occur in addition to a change in end-use technology and thus contribute to the overall efficiency of the process. Only some of these saving options were reviewed for our exercise. 3.3.2.2 Residential and service sector For electricity saving potentials in the residential and service sector, we employed the following figures (cf. Table 19). For the heating demand and the potential for energy savings in the residential sector, a detailed bottom-up modelling philosophy is applied, starting from the basic unit: the number of houses. The house stock is characterised by the following variables: age, type (open, half-open, closed and flats), heating system (centralised or local) and insulation level. Considering the evolution of the population, the size of the households and the number of demolitions, the evolution of the number of houses in each category can be calculated. The demand for heating is then computed per type of house, and then corrected for losses and average temperature. Total heating demand takes into consideration the number of houses, the heating demand per house, and the effects of rising income on heat demand (for more details, see CES-VITO 2001). Technical potential


Market Drive

Residential electricity use

0-50%

0-10%

Small service sector elec. use

0-50%

0-10%

Large service sector elec. use

0-50%

0-10%

energy savings (gradual till 2050)

Table 19. Technical energy saving potential (residential and service sector)

The availability of insulation measures and their cost depends on the construction year of the building (3 insulation levels are considered for existing buildings). In both scenario groups, we have assumed that the K55 insulation level for new buildings applies in the residential sector since 1990. Furthermore, for new buildings, we have defined an additional new building standard with very low heating demand (the so-called ‘passivhaus’ (Fischedick et al. 2002), with specific heating demand of some 20 kWh/m2.a)566. Comparing this standard to the average heating requirement of a K55-open building of some 135 kWh/m2 shows this is very ambitious indeed! Both scenario groups then have built in different assumptions about the speed of retrofitting existing houses to the K55

566

The name derives from the fact that in this type of house the same level of heating comfort can be reached almost without making use of an ‘active’ heating unit. In a sense, our scenario assumptions are unrealistic (or rather optimistic), since we do not consider more detailed intermediary insulation standards over time (such as the K40 standard). In view of the goals of the scenario exercise, this is considered to be acceptable.

380

Chapter 6

isolation standard, and about the percentage of new houses built according to the ‘passivhaus’ standard567. 3.3.2.3 Transportation For the transportation sector, it is recognised that there are technological developments ongoing which are significantly improving engine and fuel efficiency, primarily through the introduction of a wide range of fuel saving technologies already developed (e.g. direct injection engines, variable transmission systems, and lightweight materials). There is also a range of emerging vehicle technologies which can further improve the energy efficiency of vehicles over the longer term (such as hybrid electric and fuel cell technology). Likewise, other measures such as reducing traffic congestion or the encouragement of modal shifts away from the car to more sustainable alternatives could have a large impact on the CO2 reduction potential in the transportation sector. No robust estimates of the long term potential for modal shift and their subsequent impact on carbon emissions are available, and given the wide range of potential measures in the transportation sector which could reduce carbon emissions, it is not possible at present to provide a comprehensive assessment of different long term options for this report. Therefore, we have only included technological options for CO2 reduction in the transportation sector. Vehicles are supposed to become much more fuel efficient, up to a factor 4 in 2050. 3.3.3

Supply side analysis

3.3.3.1 Brief description of key technological options The MARKAL database contains a great number of different technologies for power generation. Power generation options which are expected to play a key role in the future energy system are described in the following sections. Figures are taken from the original MARKAL database (described in CES-VITO 2001), the AMPERE report (Section I on the evaluation of the current and future costs of electricity production and external costs), the report by Lako et al. (1998) and the WEC (2000), and are given in Annex 3. The residual capacities of existing power plants were adapted to those proposed in the electricity sector equipment plan 1995-2005. Nuclear For existing nuclear power plants the average availability factor and the technical lifetime are very important parameters. The availability factor used in MARKAL is 85%, the Belgian historical average for this type of plants. It is assumed that the decided investment on existing nuclear plants will allow maintaining this availability factor in the future. The operation time is set equal to 40 years, in accordance with the Belgian phaseout law. In the scenarios with letter code P, no new investments in nuclear power are

567

The ‘Rational Perspective’ worldview assumes 2% of the existing housing stock will be retrofitted per year starting from the year 2000, resulting in a complete renovation in 2050. In the ‘Market Drive’ worldview, investment in isolation measures is solely driven by cost considerations.


381

allowed. However, in other scenarios we have left open the possibility of new investments. For this report, we have considered the use of two new reactor types: the EPR (European Pressurized Water Reactor) – an evolutionary design, or the MHTGR (Modular HighTemperature Gas-cooled Reactor) – a more innovative concept, described in more detail below. Future costs for these reactor types remain uncertain until one has actually been built and operated. High-temperature gas-cooled reactors (HTGRs) typically involve large numbers of uranium fuel pellets encased in layers of carbon, silica, or both (designed to contain the fission products from the reaction). Modern HTGRs are designed to be passively safe, offering the potential to avoid many of the complex, expensive safety systems used in LWRs. It is hoped that such features could lead to lower costs and improved safety. The key to enhanced safety for the so-called ‘pebble-bed modular reactor’ (PBMR) and the ‘gas turbine-modular helium reactor’ (GT-MHR) is a design that ensures that the highest temperature in the reactor core—under any conceivable operating or accident condition— never exceeds the operating limit of the fuel. This requirement limits the thermal output for a single module to 250 MWth and the electrical output to 100 MWe—a factor of 10 smaller than for a typical LWR – hence the modular characteristic of this technology. The spent fuel of the PBMR would be high-burn-up material in many tiny spheres, making it a comparatively unattractive source from which to recover weapons-usable material (WEC 2000). In our choice of nuclear technologies to be included in the MARKAL database, we only considered these two (in a sense) ‘extreme’ cases: very centralised (the EPR would have a typical installed capacity of 1400 MWe per reactor) vs. more decentralised (the MHGTR would have an installed capacity per module of 100 MWe). This does not imply that these reactor types represent the only nuclear technology options available for the future568. Fossil Fuel Based Power Plants Two types of coal power plants are considered in the model: an ultra super-critical coal power plant (USC) and integrated coal gasification combined cycle (IGCC). These types of power plants can become the main base load technology in the longer run when natural gas becomes more expensive. IGCC is a technology which is still at the demonstration level: important technical progress has been foreseen in the energy efficiency (from 42% in 1995 to 51% in 2030). Both technologies are, at least at the end of the horizon, in close competition with the cost and efficiency figures assumed here. Gas-fired plants are either gas turbines or steam and gas power plants (STAG). In addition, the option of CO2 capture and storage (CCS) is included in the database. Carbon dioxide capture and storage is essentially a process whereby CO2 is removed from the fossil fuel used to generate electricity (either pre- or post-combustion) and stored in

568

An ‘intermediate’ size new nuclear reactor type (600 MWe) would for instance be the advanced light water reactor developed by Westinghouse, known as the AP-600.

382

Chapter 6

natural underground reservoirs, preventing it reaching the atmosphere. It can achieve an 80% reduction in CO2 emissions to the atmosphere. Carbon dioxide storage will only be an effective way of avoiding climate change if the CO2 can be stored for several hundreds or thousands of years. The most promising storage options are depleted and producing oil and gas reservoirs (where CO2 injection for enhanced oil recovery is possible), deep saline reservoirs, unminable coal beds or aquifers. However the legal status and public or political acceptance of disposal in sub-sea strata is questionable. Carbon capture and storage technologies (CCS) are best suited to large-scale sources of CO2 such as power stations (coal or gas). However, the large capital costs involved in adapting existing generation plants to CCS and the resulting loss in efficiency, mean the technology is best applied to new plants, where it can be incorporated into initial construction. In addition, if hydrogen became established as a major fuel for cars, electricity production and heat and power generation; centralised, large-scale production of hydrogen from fossil fuels would be possible from precombustion capture of CO2 emissions. All the individual components of the technology exist and are commercially proven and could be deployed. However, there appears to have been no systematic probabilistic analysis of risks and environmental consequences, or systematic assessment of the available data on slow release (UK Energy White Paper 2003). In view of the uncertainties involved with this technological option, some scenarios will have constraints imposed on the amount of CO2 that can be stored in deep geological layers (in the scenarios with letter code LCS, a limit of 5 Mton CO2 /a has been applied). Cogeneration technologies In the industrial sector, gas turbines of 4MWe and of 35MWe and STAG of 30MWe are considered for high temperature steam and a backpressure turbine (20MWe) for low temperature steam. STAG units of 30 and 100 MWe and gas turbines and diesel engines of 30MWe for cogeneration in the residential and tertiary sector are modelled. A number of gas turbines on biomass, as well as fuel cells for cogeneration have been included in the model (fuel cells for residential use were added). Fuel cells either use hydrogen or employ a gas reformer. Small gas engines of 1MWe for decentralised cogeneration in this sector are also considered. Limits have been imposed on the penetration of cogeneration in the different sectors (see under constraints). Renewables Conventional renewables like hydro power, a large range of biomass-fuelled power plants and wind turbines (both onshore and offshore) are addressed in the database. Photovoltaic electricity generation has been added as an option for the long-term. The long-term potential of both wind energy (offshore) and photovoltaics is subject to uncertainty and thus constitutes a variable in different scenario groups. In scenarios with both a nuclear phase out and low potential for CCS (letter code P.LCS), cogeneration and renewables will play a comparatively more important role in energy provision.


383

Import of electricity Some scenarios (with the letter code I) rely on a significant import of foreign electricity to meet energy demands. In these scenarios, electricity imports of approximately one third of the total demand for electricity have been allowed in 2050 (linear interpolation from 1990 levels). To determine the impacts of this import of electricity, it is of course necessary to know the electricity generation mix of the imported electricity mix. We have taken as a reference the global energy perspectives developed by Nakicenovic et al. (1998). For the ‘Market Drive’ worldview, we have chosen the electricity generation mix of scenario A3 for Western Europe569. This mix amounts to (in 2020/2050): 20/35% gas fuelled power generation, 25/25% renewables (wind, solar, hydro, biomass and waste) and 25/40% nuclear. For the ‘Rational Perspective’ worldview, we have chosen the electricity generation mix of scenario C1570. This mix amounts to: 20/40% gas fuelled power generation, 25/45% renewables (wind, solar, hydro, biomass and waste) and 25/15% nuclear. 3.3.3.2 Constraints Some of the technologies incorporated in the MARKAL model are well-known, e.g. proven coal and gas-fired power technologies. A lot of technologies are in the stage of development, demonstration or early implementation. A first criterion to select technologies for uptake in the database is technical feasibility. Based on this criterion, more speculative options such as fusion power or a nuclear fuel cycle with advanced partitioning and transmutation can be excluded (at least during the time horizon under consideration). Constraints to the application of certain technologies can result from a number of factors. For instance, technical-economic constraints apply to certain renewable energy options. Wind energy has been split up in several categories (onshore at seaside, onshore polders, onshore inland, offshore) depending on wind regime and availability of space. Each category has an upper bound in a specific period. Photovoltaics suffer from an upper bound on capacity due to limiting factors such as availability of suitable surfaces and solar irradiation. However, as these bounds are rather uncertain on the long term, different assumptions have been used in both scenario groups (‘optimistic’ in scenario group ‘Rational Perspective’, ‘pessimistic’ in scenario group ‘Market Drive’). In the RP worldview, a higher potential has been assumed for offshore wind energy, but the

569

This scenario is described by the authors as “...an illustration of of a case where a ‘rich and clean’ energy future resolves some of the challenges of global warming without recourse to stringent environmental policy measures…” (p. 73). 570 This scenario is described by the authors as “...a challenging pathway of transition away from the current dominance of fossil sources to a dominance of renewable energy flows. Ambitious policy measures accelerate energy efficiency improvements and develop and promote environmentally benign, decentralised energy technologies…” (p. 74).

384

Chapter 6

additional 1 GWe (compared to the MD worldview) can only be realised at a 30% increase in investment costs571. For nuclear power generation, we assumed that only existing sites can be used for possible new reactor units. This limits the maximum installed nuclear capacity to 8 GWe (in all scenarios). If new nuclear power plants are allowed in the ‘Rational Perspective’ worldview, we assumed that new reactors would be of the MHTGR type (in line with the promotion of smaller-scale technologies). Of course for nuclear energy, public acceptance of the technology, and possible long lead times, high capital costs, waste management issues, uncertainty about back-end costs and, at present, lack of political readiness (in most OECD countries) to promote nuclear as an option, are the main barriers to new nuclear construction. All of these issues might lead to a political or a ‘de facto’ nuclear phase out (letter code P). Public acceptance might however also be an issue for other technologies, such as wind power (visual intrusion). This could well lead to some overestimation of the potential for wind power. However, one might argue that the resulting smaller potential for onshore wind energy could then be offset by a larger application of offshore technology. The only difference then is the total economic cost. Carbon capture and storage technologies might also suffer from a lack of public or political acceptance, or are even still subject to technological uncertainties. We considered two cases: either a constraint (of 5 Mton, letter code LCS) to the amount of CO2 which can be stored in an acceptable way has been applied, or the CCS potential was considered to be sufficiently large for (no constraint). Worldview RP

2000

2030

2050


1.0

1.6

2.0


0.9

1.2

1.4


-

10

10

1.0

2.7

3.9

0.2

0.5

0.6

0.4

1.2

1.6

0

8

8

Cogeneration (industry) – HT steam (GWe) Cogeneration (industry) – LT steam (GWe) Cogeneration (service + res) (GWe) MHTGR nuclear reactor (GWe)

571

A recent in-depth study (published after we developed our scenarios) has calculated an offshore economic potential of 2.1-4.2 GWe installed capacity (Van Hulle et al. 2004). Thus our ‘optimistic’ assumption might turn out to be too ‘pessimistic’ after all.

Criteria and scenarios as a support for sustainable energy governance Worldview MD

2000

2030

2050


1.0

1.0

1.0


0.75

0.75

0.75


-

10

10

1.0

1.8

2.0

0.2

0.3

0.3

0.4

0.8

0.8

0

8

8

Cogeneration (industry) –

385

HT steam (GWe) Cogeneration (industry) – LT steam (GWe) Cogeneration (service + res) (GWe) EPR nuclear reactor (Gwe)

Table 20. Constraints on technological options in both worldviews

Table 20 summarises the default constraints adopted in both worldviews. For cogeneration technologies, maximum capacities are taken from the AMPERE report (2000). However, as these capacities are largely uncertain on a long term, the possibility of a doubling of maximum capacity has been allowed under the ‘Rational Perspective’ worldview. 3.3.4

Policy measures

This kind of analysis does not deal explicitly with the policy measures needed to attain the strategic objectives. For instance, (post-)Kyoto targets can simply be ‘imposed’, without however considering the exact mix of policy measures needed to arrive at these targets. It is simply assumed that policy measures taken will be in line with the defined objectives and the major technological options chosen within different scenarios (e.g. policy measures to promote cogeneration in scenarios with nuclear phase out, continuation of nuclear related R&D in scenarios without a phase out, R&D for options with carbon capture & storage, etc.). 3.3.5

Visions beyond 2050

As already stated, none of the scenarios can claim to provide a definitive sustainable solution to energy provision in 2050. For fossil-fuel based power generation, carbon capture and storage can only be used as a transitional solution towards more sustainable energy supply options, as fossil fuels will inevitably run out (however, coal could still provide a resource base for some centuries to come). Proponents of the nuclear option propose (as an option for a very far future, e.g. 20402060) to recycle the majority of the fission products with long half-lives (mainly 99Tc and 129 I) and of the actinides (mainly Pu and Am, elements with long half-lives) to fission in fast-spectrum reactors. The remaining waste would then be mainly fission products with half-lives of decades or shorter, and after 300 years the total fission-product inventory would decline to a radio toxicity level lower than that of the original ore. Of course, it is very uncertain whether (and when) this option will ever become technically and

386

Chapter 6

economically feasible. Also, innovative fuel cycles based on thorium (largely available) are investigated in response to possible uranium resource limitations. Furthermore, fusion energy might become available after 2050. One could also envisage an energy system largely based on renewable energy sources (combined with a reduction in the demand for energy services) – see e.g. the report by the LTI research group (LTI 1998). Optimal integration of a large share of different renewable energy production options in the energy system is however only conceivable on a European scale.

3.4

Results

This chapter gives results for the 8 scenarios developed for this exercise. In section 3.4.1, key findings will be discussed for each option (under the ‘Market Drive’ and ‘Rational Perspective’ worldview). Then, some striking differences between the scenarios will be illustrated (Section 3.4.2). Finally, complete results for quantifiable parameters for all scenarios will be discussed (Section 3.4.3). The assumptions behind these scenarios have been described in section 3.2, but we recapitulate here for practical purposes: • M.K.P.CS – Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity; • M.K.LCS – Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), new nuclear build, limited possibilities for carbon capture and storage technology, limited import of electricity; • M.K.P.LCS: Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; 15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity; • M.K.P.LCS.I: Market Drive worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity; • R.K.P.CS – Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, introduction of carbon capture and storage technology for fossil fuel-based power generation, limited import of electricity; • R.K.LCS – Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), new nuclear build, limited possibilities for carbon capture and storage technology, limited import of electricity;


387

• R.K.P.LCS: Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology – and thus greater reliance on renewables and cogeneration, limited import of electricity; • R.K.P.LCS.I: Rational Perspective worldview complying with post-Kyoto requirements (-7.5% in 2010; -15% in 2030; -30% in 2050), phase out of existing nuclear power plants (40 yr lifetime), no new nuclear build, limited possibilities for carbon capture and storage technology, increasing reliance on import of foreign electricity.

In addition, two baseline scenarios (for the ‘Market Driven’ en ‘Rational Perspective’ worldview) have been calculated for cost comparison purposes. 3.4.1

Selected results per option

3.4.1.1 Nuclear phase out (scenario R.K.P.CS and M.K.P.CS) Primary energy use The total primary energy demand increases slightly for the M.K.P.CS scenario, while – after an initial increase - remaining almost equal to the present primary energy demand (some 2300 PJ) in the R.K.P.CS scenario (notwithstanding economic growth). Several changes in the fuel mix occur over time. Until the year 2020, oil consumption remains rather constant (compared to 1990 levels), gas consumption rises rapidly (more than a doubling!), and coal consumption decreases. Trends for oil, gas and coal are more or less the same for both worldviews (although the absolute rise in gas consumption is less pronounced in R.K.P.CS).


2500

PJ

2000


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050

388

Chapter 6


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 8. Primary energy demand – scenarios M.K.P.CS (top) and R.K.P.CS (bottom)

After 2020, appreciable changes occur. Nuclear energy is phased out, and coal use rises sharply. Gas use remains high (more or less levelling at 2020 demand) for the M.K.P.CS scenario, while in the R.K.P.CS scenario it decreases to some 700 PJ in 2050. Oil use decreases (due to the introduction of fuel efficient vehicles). Renewables gain a modest share of some 5-6% in 2050572. Electricity production The trends for electricity production are more pronounced than those of primary energy demand. In the M.K.P.CS scenario, coal shows a rather steep decline until 2015. After that, coal replaces the nuclear generation capacity and becomes the dominant fuel choice for electricity generation in 2050. Natural gas shows a steady increase until 2035 and declines a bit afterwards. The coal gasification combined cycle (IGCC) and steam and gas combined cycle (STAG), both equipped with carbon capture technology, become the dominant electricity generation options from 2025 onwards. Electricity demand is very high. This is logical: since electricity generation in this scenario is essentially carbon-free, switching to electricity as energy carrier presents a robust and economically attractive (relative to other options) solution to meeting CO2 reduction targets.

572

Primary energy use is calculated based on the net calorific values of fuels. For electricity from renewable sources, we applied the “substitution equivalence” method, by which primary energy equivalents are calculated assuming a conversion efficiency of 38.6 %. For nuclear electricity, we used a value of 33%.


389


160

140

120 Renewables

100 TWh


80

Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 9. Electricity production – scenarios M.K.P.CS (top) and R.K.P.CS (bottom)

In scenario R.K.P.CS, coal becomes even more important (in relative terms) than in M.K.P.CS, and completely replaces gas for electricity production (the remaining cogeneration capacity is still gas-fuelled however): since the discount rate is lower in the ‘Rational Perspective’ scenario, investment in power plants will take into account price effects on a longer timescale (and gas prices show a steady increase compared to coal prices). Gas is reserved for other uses (e.g. process heat for industry) where no electric alternative exists. Overall electricity demand is lower as a result of more energy-saving measures. Cogeneration (mainly for high-temperature steam in industry) gains a relatively

390

Chapter 6

large market share in the period 2010-2035 before declining. Renewables are more important in the R.K.P.CS case (gaining a share of 13% of total electricity production in 2050). CO2 emissions CO2 emissions in both scenarios are somewhat similar. A decrease is most noticeable in the residential & service sector, although the mechanism is somewhat different in both scenarios: in the M.K.P.CS scenario, electric heating appliances (various heat pumps and some electric accumulation heating) gain a large market share, while in the R.K.P.CS case, lower CO2 emissions result mostly from energy saving measures (e.g. better isolation). CO2 emissions [Mton]

2000

2010

2020

2030

2040

2050

Electricity

20

15

15

16

18

19

Industry

34

25

26

24

23

23

Resid. & service

34

28

25

23

21

15

Transportation

26

30

28

28

21

19

Other

4

4

3

3

2

1

Total

118

102

97

94

85

77

2020

2030

2040

2050


2000

2010

Electricity

19

13

15

20

17

18

Industry

34

27

27

28

29

26

Resid. & service

32

26

22

19

17

13

Transportation

26

30

29

25

21

19

Other

4

4

3

0

0

0

Total

115

100

96

92

84

76

Table 21. CO2 emissions – all sectors (scenario M.K.P.CS (top) and R.K.P.CS (bottom))

Electricity consumption in the residential & service sector amounts to 92 TWh in M.K.P.CS, compared to 47 TWh in R.K.P.CS. Also, the R.K.P.CS case relies heavily on centralised district heating (mainly for urban dwellings): 93 PJ is provided by this technology in 2050 (vs. 31 PJ for the M.K.P.CS scenario). In both scenarios, some of the heating demand in flats is met by combined heat-and-power (CHP): 23 PJ in M.K.P.CS vs. 28 PJ in R.K.P.CS. CO2 emissions in the electricity sector remain relatively low despite the nuclear phase out as a result of the availability of carbon capture and storage technology.


391

3.4.1.2 Low potential carbon capture and storage (scenario M.K.LCS and R.K.LCS) Primary energy use


2500

PJ


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


2500

PJ


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 10. Primary energy demand – scenario M.K.LCS (top) and R.K.LCS (bottom)

Again, the total primary energy demand increases some 20% over the entire time horizon for the M.K.LCS scenario, while the increase is less marked for the R.K.LCS scenario. In

392

Chapter 6

both cases, nuclear power plants are constructed up to the full capacity of 8 GWe starting from 2020 (after the existing nuclear power plants are retired). Coal only plays a minor role, while the economy becomes increasingly dependent on imported gas (less dependence in absolute terms in the R.K.LCS scenario: 750 PJ vs. 1075 PJ in 2050 for the M.K.LCS scenario). Renewables again play only a relatively small part: 6% in both cases. In these scenarios there is a much greater emphasis on gas than in the scenarios of the previous section (M.K.P.CS and R.K.P.CS) because coal is basically forbidden by post-Kyoto GHG emission limits and the non-existence of carbon storage, while at the same time nuclear is limited to 8 GWe. In the previous section coal with CS was allowed to expand unconstrained. Electricity production The trends for electricity production are shown in Figure 11. Some trends can be noted. In the M.K.LCS scenario, gas becomes the second choice fuel after the nuclear electricity generation capacity has been fully developed. From 2040 onwards, fuel cells become the preferred option for industrial cogeneration (generating some 13 TWh in 2050). The hydrogen for these fuel cells is produced starting from natural gas (in a steam reforming process); it is assumed that the CO2 generated during the process is captured and stored onsite, while the hydrogen is transported (e.g. by pipeline) to the power plant573. Renewables (mostly wind power and photovoltaics) represent a modest share of electricity generation: some 12% in 2050.


140

120

100 Renewables

TWh

Cogeneration

80

Gas Coal Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

573

For steam reforming, a production cost of 5.92 MEuro per PJ natural gas is assumed (in addition to the natural gas price of 1.69 – 4.35 MEuro/PJ). No detailed cost calculations for transport, storage and distribution of hydrogen have been included. This of course implies that the costs of the hydrogen option are underestimated.


393


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 11. Electricity production – scenario M.K.LCS (top) and R.K.LCS (bottom)

In the R.K.LCS scenario, the use of gas for electricity generation is restricted to cogeneration plants from 2035 onwards. Cogeneration is mainly used to produce hightemperature steam for industrial uses, but some combined heat-and-power is also installed in flats and large service sector buildings (providing 4.4 TWh of electricity in 2050). In this scenario, almost no use has to be made of the (relatively expensive) hydrogen to reach the post-Kyoto requirements. Renewables steadily increase over time, providing some 15% of electricity in 2050. Photovoltaics are not used up to their full potential. The existing nuclear production park is replaced by the more small-scale (100-300 MWe per unit) modular high-temperature gas-cooled reactor starting from 2020, up to the full potential of 8 GWe. CO2 emissions CO2 emissions [Mton]

2000

2010

2020

2030

2040

2050

Electricity

20

14

15

11

12

15

Industry

34

25

26

25

24

24

Resid. & service

34

29

26

27

25

17

Transportation

26

30

27

28

21

19

Other

4

4

3

3

3

3

Total

118

102

97

94

85

78

394

Chapter 6


2000

2010

2020

2030

2040

2050

Electricity

19

13

13

11

8

9

Industry

34

27

28

29

29

30

Resid. & service

32

26

22

24

23

15

Transportation

26

30

30

25

21

19

Other

4

4

3

3

3

3

Total

115

100

96

92

84

76

Table 22. CO2 emissions – all sectors (scenario M.K.LCS (top) and R.K.LCS (bottom))

Similar observations as in the previous case can be made. A decrease is most noticeable in the residential & service sector. More electricity is used in the M.K.LCS scenario in these sectors (although the effect is now less pronounced: 78 TWh vs. 49 TWh for the R.K.LCS scenario). Heating is provided for with a mixture of heating pumps and gas boilers in open and half open residences. Again, in the R.K.LCS scenario central district heating is used to meet some of the heating demand in urban dwellings (45 PJ in 2050), while this option does not become available in the M.K.LCS scenario. Decentral combined heat-and-power provides for another 21 PJ of heating demand in the M.K.LCS case, compared to 29 PJ in the R.K.LCS case. As a result of the lower electricity demand in the R.K.LCS scenario, CO2 emissions in the electricity sector remain very low, while an increasing reliance on gas in the M.K.LCS scenario results in higher CO2 emissions starting from 2030. 3.4.1.3 Phase out and low potential carbon capture and storage (scenario M.K.P.LCS and R.K.P.LCS) Primary energy use For the first time, an absolute decoupling between economic growth and primary energy use can be noticed for the R.K.P.LCS scenario, while in the M.K.P.LCS case, primary energy use more or less returns to the 1990 demand level. Both scenarios show somewhat similar results. Most noticeable is the increasing dependence on imported gas: in the M.K.P.LCS scenario, this dependence amounts to 1110 PJ in 2050, while in the R.K.P.LCS scenario, it is limited to 1034 PJ (due to the overall lower primary energy demand). Also, renewable energy sources gain a relatively large importance: 9% of primary energy demand in the R.K.P.LCS scenario and even 12% in the M.K.P.LCS scenario (this is because in the period 2040-2050 solar boilers are installed in order to meet CO2 reduction targets).


395


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 12. Primary energy demand – scenario M.K.P.LCS (top) and R.K.P.LCS (bottom)

Electricity production The results are more marked for electricity production (cf. Figure 13). In the M.K.P.LCS scenario, the existing nuclear power plants (phased out in 2015-2025) are almost exclusively replaced by STAG power plants (some equipped with carbon capture and storage technology, up to the maximum potential of 5 Mton CO2 stored per year). Renewables (mainly wind energy and photovoltaics) are developed up to their maximum potential, contributing some 14% of electricity demand in 2050. Cogeneration is, starting

396

Chapter 6

from 2025, largely based on fuel cell technology (to generate high pressure steam for industrial demand), contributing some 14% of electricity demand in 2050. Again, the hydrogen for these fuel cells is produced starting from natural gas (in a steam reforming process); it is assumed that the CO2 generated during the process is captured and stored onsite, while the hydrogen is transported (e.g. by pipeline) to the power plant. In the R.K.P.LCS scenario, the large dependence on cogeneration technology immediately stands out. When the existing nuclear power plants are phased out in 20152025, they are still largely replaced by gas-fuelled electricity generation (STAG power plants), although in later years (2030-2050), this technology loses some market share to cogeneration technologies. Renewables gain a market share of 20% in 2050 (all options are used up to their full potential), cogeneration even 37% (remember that the ‘Rational Perspective’ scenarios include ‘optimistic’ assumptions about constraints to both technological options). The dominant technology for cogeneration is the hydrogen-fuelled fuel cell for high-temperature steam in industry (23 TWh in 2050), although some applications can also be found in centralised and decentralised heating for the residential and service sectors.


120

100

Renewables

80 TWh


60

Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050


397


100

80 Renewables

TWh

Cogeneration

60

Gas Coal Nuclear

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 13. Electricity production - scenario M.K.P.LCS (top) and R.K.P.LCS (bottom)

CO2 emissions In the M.K.P.LCS scenario, the electricity production sector becomes the largest contributor to total CO2 emissions in 2050 due to the increasing reliance on natural gas as the energy vector of choice. CO2 emission reduction efforts are largely borne by the residential and service sector. Emission reductions are realised through the diffusion of solar boilers (providing for 108 PJ of space heating and warm water in 2050), efficient technologies (heat pumps and natural gas condensing boilers) and conservation measures. Also, there is a substantial decrease in CO2 emissions from the transportation sector in the period 2040-2050: this is because (imported) ethanol is introduced as a fuel for the transport fleet574.


574

2000

2010

2020

2030

2040

2050

Electricity

20

14

17

25

28

29

Industry

34

25

27

23

23

23

Resid. & service

34

33

29

21

10

9

Transportation

26

25

22

22

21

13

Other

4

4

3

3

3

3

Total

118

101

98

94

85

77

The ethanol fuel price is assumed to be 13.2 MEuro/ PJ (compare to the gasoline fuel price of 3.59 – 9.69 Euro/PJ).

398

Chapter 6


2000

2010

2020

2030

2040

2050

Electricity

19

13

16

20

18

16

Industry

34

27

27

25

26

24

Resid. & service

32

26

22

18

16

14

Transportation

26

30

28

25

21

19

Other

4

4

3

3

3

3

Total

115

100

96

91

84

76

Table 23. CO2 emissions – all sectors (scenario M.K.P.LCS (top) and R.K.P.LCS (bottom))

In the R.K.P.LCS scenario, CO2 emissions from the electricity sector are markedly lower (a combination of lower electricity demand and higher potential for renewables and cogeneration). This allows for less stringent (and costly!) reduction efforts in residential and transportation sectors.

3.4.1.4 Phase out, low potential carbon capture and storage and import of electricity (scenario M.K.P.LCS.I and R.K.P.LCS.I) Primary energy use


2500

PJ


1500

1000

500

0 1990

1995

2000

2005

2010

2015

2020 Year

2025

2030

2035

2040

2045

2050


399


2000

1500 PJ


1000

500

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 14. Primary energy demand – scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom)

Primary energy use remains rather flat for both scenarios575. As in the previous scenario cases, the increasing dependence on gaseous energy carriers (once existing nuclear power plants are phased out) is obvious. Renewable energy sources provide for 14% of primary energy demand in the R.K.P.LCS.I scenario. Electricity production In the M.K.P.LCS.I scenario, nuclear electricity generation is replaced by gas-fuelled electricity generation (in STAG power plants, up to 40% in 2050), imports of foreign electricity (up to 32% in 2050), electricity from renewable sources (up to 12% in 2050) and cogeneration (up to 13% in 2050). Cogeneration is (again) mainly used to produce high temperature steam for industry. Hydrogen-fuelled fuel cells become the dominant technological choice for this option. In the R.K.P.LCS.I scenario, electricity demand is generally much lower. The reliance on STAG technology is smaller compared to the previous case: nuclear electricity generation is replaced by a mix of cogeneration (up to 26% in 2050), renewables (up to 17% in 2050) and import of foreign electricity (up to 36% in 2050). Gas-fuelled electricity generation in STAG power plants makes up for the rest (up to 18% in 2050). The preferred cogeneration technology is the gas turbine (providing 11 TWh of electricity in 2050), although in the later years (2040-2050) it is increasingly being replaced by hydrogenfuelled fuel cell technology (providing 6 TWh of electricity in 2050).

575 The imported electricity has been subdivided according to origin (renewable, nuclear, gas, coal) and then added to the corresponding primary energy category.

400

Chapter 6


140

120

TWh

100

Import Renewables Cogeneration Gas Coal Nuclear

80

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year


100

TWh

80 Import Renewables Cogeneration Gas Coal Nuclear

60

40

20

0 1990

1995

2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Year

Figure 15. Electricity production - scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom)

CO2 emissions Table 24 represents the sectoral CO2 emissions for both scenarios. CO2 emissions (or other emissions) resulting from imports of foreign electricity are not included in this table; it is assumed that neighbouring countries also have to comply with a comparable CO2 emission reduction target in 2050.



2000

2010

2020

2030

2040

2050

Electricity

19

11

12

15

15

18

Industry

34

25

25

23

23

23

Resid. & service

34

31

26

25

23

13

Transportation

26

30

31

27

21

19

Other

4

4

3

3

3

3

Total

117

101

97

93

85

76

2000

2010

2020

2030

2040

2050

Electricity

18

12

13

16

15

13

Industry

34

27

27

28

28

27

Resid. & service

32

27

23

19

17

13

Transportation

26

30

29

25

21

19

Other

4

4

3

3

3

3

Total

114

100

95

91

84

75


401

Table 24. CO2 emissions – all sectors (scenario M.K.P.LCS.I (top) and R.K.P.LCS.I (bottom))

These imports notwithstanding, CO2 emissions in the electricity sector rise after 2020 in the M.K.P.LCS.I scenario due to the increasing reliance on (domestic) STAG electricity production. In the R.K.P.LCS.I scenario, emissions can be kept low due to the higher potential for cogeneration and renewables. Again, a large part of the necessary reductions can be achieved in the residential and service sectors. The R.K.P.LCS.I scenario relies more on centralised district heating (45 PJ in 2050) and decentralised combined heat-andpower (28 PJ in 2050) than the M.K.P.LCS.I scenario (0 PJ and 22 PJ respectively). 3.4.2

Comparison between scenarios

In this part, a summary of the most important results is given. Some graphs explaining differences in final energy demand and electricity demand in the above-mentioned scenarios will be explored. More complete results on quantifiable scenario indicators (including sensitivity analyses) are given in the next section. 3.4.2.1 Final energy demand The final energy demand projection for all scenarios is considerably lower than in the baseline ‘Market Driven’ case. For the ‘Rational Perspective’ scenarios, this difference amounts to some 30% in 2050. This is due to substantial increases in the efficiency of enduse technologies. In the RP scenarios, more efficient technologies are introduced than in the MD scenarios, due to the lower discount rate applied. In the base case, final energy demand would reach some 2100 PJ in 2050, while most MD scenarios level off at about 1800 PJ (a 10% increase compared to 2000). The M.K.P.LCS scenario is an exception: here, supply-side measures to reduce CO2 emissions are limited in potential, so that in this

402

Chapter 6

scenario reductions in energy demand comparable to most RP scenarios have to be achieved (notice however the much more drastic reduction effort in 2020-2030 as a result of the higher discount rates). In the later years (2040-2050), some supply side measures (ethanol fuelled cars for transport and solar boilers for heating) are again preferred. Most RP scenarios level off at some 1600 PJ (or even lower) – almost equal to final energy use in 2000. This would on average imply that the improvements in energy intensity of the national economy have to keep up with economic growth. 2,200.00

2,100.00

2,000.00 BASE MKLCS MKP

1,900.00 PJ

MKPLCS MKPLCSI RKLCS 1,800.00

RKP RKPLCS RKPLCSI

1,700.00

1,600.00

1,500.00 2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Figure 16. Final energy use (all scenarios)

3.4.2.2 Electricity demand Electricity demand varies considerably between the different scenarios. The increase in electricity demand until 2050 ranges between 30% and 100% (a doubling) compared to the electricity demand in 2000. This means that, in all scenarios, electricity will increase its share in meeting the final energy demand. A shift from other energy vectors towards electricity thus appears to be a relatively robust option for achieving CO2 emission reductions under a variety of assumptions.


403

160.0

150.0

140.0

130.0 MKLCS 120.0

MKP

TWh

MKPLCS MKPLCSI

110.0

RKLCS RKP RKPLCS

100.0

RKPLCSI 90.0

80.0

70.0

60.0 2000

2005

2010

2015

2020

2025

2030

2035

2040

2045

2050

Figure 17. Electricity production (all scenarios)

The choice of technological options for the electricity production also clearly influences the level of electricity demand in the MD scenarios. For instance in the M.K.P.CS scenario, a comparatively cheap source of carbon-free energy is available (electricity produced by fossil fuel-based power generation with carbon capture and storage), and MARKAL will correspondingly use this energy carrier to a maximum degree576. This is also true for the M.K.LCS and M.K.P.LCS.I scenarios, but to a lesser extent: the availability of carbon-free electricity is constrained by the 8 GWe limit for domestic nuclear electricity production, and roughly one third of total electricity demand for import of electricity. In the scenario variants with high electricity demand, electricity increasingly replaces fossil fuels, for instance for water and space heating (heat pumps gain a large market share). Impacts of electricity generation options on electricity demand are less outspoken in the RP scenarios, because carbon-free electricity sources have to compete with demand-side measures (e.g. better isolation) for a market share.

576

A sensitivity analysis with 30% higher investment costs for all electricity generation option equipped with CCS showed no significant impact on the level of output of these power plants.

404

3.4.3

Chapter 6

Detailed results (including sensitivity analysis)

This section gives some detailed results on the quantifiable criteria for the different scenarios. For each scenario, two sensitivity analyses were performed: one with higher costs for fossil fuel imports (gas and oil) (cf. Table 25), the other with a more stringent CO2 reduction target (-7.5% in 2010, -30% in 2030, -50% in 2050) and lower demands for energy services in the industrial and transportation sector (-30% for both sectors in 2050 compared to the base scenarios, with linear interpolation from 1990 demand levels). This second sensitivity analysis might for instance represent an economic development with structural changes towards less energy intensive industries (e.g. more service-oriented), less energy-intensive production processes in industrial sectors (e.g. use of new materials to replace steel, use of membrane technology, etc.), higher efficiency gains than presumed in the base scenarios, or the emergence of new consumer values (e.g. less materialism, ecofriendly consumption, etc.). Results are presented on a comparative basis (the exact magnitude of the parameters involved is of less importance), with ‘optimistic’ and ‘pessimistic’ scores for each scenario (i.e. the ‘best’ and ‘worst’ result for the scenario in question, including both sensitivity analyses). Low scores for a certain criterion always represent a ‘bad’ score, even though the primary indicator on which the criterion is based might be higher for the ‘bad’ result. For instance, if the cost for scenario A would be in the range of 80 - 100 MEuro, and for scenario B in the range of 30-50 MEuro, scenario A would have a score of 0 - 20 and scenario B a score of 50 - 70 on a scoring scale where 0 = 100 MEuro and 100 = 0 MEuro.

Cost (MEuro/PJ)

1990

2000

2010

2030

2050

Import of High Sulphur Heavy Distillate

2.20

1.71

2.97

4.95

6.91

Import of Low Sulphur Heavy Distillate

2.52

1.96

3.39

5.64

7.87

Import of Gasoline

4.88

3.59

5.54

9.23

12.90

Import of Light Distillate Oil

4.21

3.19

5.30

8.84

12.35

Import of LPG

3.07

2.75

3.27

4.75

6.24

Import of Natural Gas

2.48

1.68

2.97

4.33

5.67

Table 25. Resource prices (sensitivity analysis)


405

3.4.3.1 Health and environment Impacts on public health caused by air pollution (during power generation stage) Impacts on public health caused by air pollution (SO2, NOx and particulate matter (PM)) mainly arise during the power generation stage. In order to estimate future emissions of these pollutants, emission coefficients of SO2, NOx and PM associated with the different electricity producing technologies have been added to the Belgian MARKAL database on an average basis (therefore they can only be considered as indicative). For our purposes, the impacts of air pollutants on public health have been weighted according to the monetary valuation of the damages they cause. Results are taken from the ExternE project of the European Commission (cf. Chapter 3). The figures used here are based on the framework developed in this project, although at a more aggregated level (cf. Table 26).

Power generation All fuels (Euro/ton) SO2

5,998

NOx

5,000

PM

11,989

Table 26. External costs of air pollutants (Source: AMPERE Principal Report, Section I, p. 6)

Briefly, the ExternE methodology involves the estimation of external costs (based on the willingness-to-pay or willingness-to-accept concept) based on the impact pathway (starting from primary emissions, transportation and atmospheric chemistry, deposition processes and different burdens are taken into account). The damage categories considered in Table 26 are acute morbidity (e.g. emergency room visits, restricted activity days, etc.), mortality and chronic morbidity (e.g. chronic bronchitis, asthma, non-fatal cancers, etc.), but no occupational health effects (these are treated under a separate heading, because they mainly arise during other stages of the fuel cycle, mostly outside Belgium). Also, damages to materials and agricultural crops are included, although these damage categories only contribute to a minor extent to the total cost figure. It is clear that measuring environmental costs at such a global level as in this model raises different problems (which have been discussed in more detail in chapter 3). However, despite all uncertainties involved, external cost calculations can still provide an informative, comparative and quantified indicator of the health impacts of the different scenarios. Other (more precise) possible indicators (such as the absolute emission levels of the pollutants under consideration) give no information on the actual health impacts associated with these emission levels. Results are shown in Table 27. For each scenario, cumulative

406

Chapter 6

(undiscounted) health damages have been calculated for the time periods 1990-2020 and 2020-2050. The differences between the scenarios over the period 1990-2020 are relatively small. However, in 2020-2050, scenarios involving a large reliance on coal technologies generally score lower (M.K.P.CS and R.K.P.CS). The best performing scenarios are those which allow for a continuation of nuclear power (M.K.LCS and R.K.LCS) and those which rely on imports of electricity (M.K.P.LCS.I and R.K.P.LCS.I). Score (0-100): 0=25000 ; 100=2500 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI


30

29

25

30

26

25

25

25

Worst score

26

24

21

27

23

22

23

23

Best score

96

51

86

94

98

60

91

89

Worst score

93

22

81

91

91

26

87

83

2020-2050

Table 27. Impacts on public health caused by air pollution

Impacts on occupational health (coal and gas fuel cycle) Burdens on the workforce employed in the gas and coal fuel cycle arise mostly in other stages than the power generation stage. For the gas and coal fuel cycle, monetary values for these impacts have also been calculated (De Nocker et al., 1998). For Belgium, values of 0.69 mEuro/kWh are used for the coal fuel cycle and 0.07 mEuro/kWh for the gas fuel cycle. Impacts on occupational health mainly arise due to accidents (causing minor injuries, major injuries or death) during coal mining and transport of coal or waste materials, or due to radon exposure (causing lung cancer) or diseases caused by coal dust (chronic bronchitis, chronic cough, etc.). For the gas fuel cycle, impacts are mainly accident-related and generally lower than for the coal fuel cycle. Impacts arising during the nuclear fuel cycle are described under a separate section. Table 28 summarises the cumulative external cost implications (1990-2050) for the different scenarios. Clearly, scenarios where coal is used intensively score rather low on this criterion. Score (0-100): 0=2500 ; 100=0 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

84

53

78

82

85

51

81

84

Worst score

82

8

72

76

82

18

80

81

Table 28. Impacts on occupational health (coal and gas fuel cycle)


407

Radiological health impacts of nuclear fuel cycle Radiological impacts of the nuclear fuel cycle have also been quantified according to the ExternE methodology (De Nocker et al. 1998; AMPERE Principal Report, Section I, p. 6). Due to the long half-lives of some radionuclides, low-level doses will exist into the far future. These low-level doses can add up to larger values when the total population dose is considered for thousands of years. Of course, the uncertainty of the models used to evaluate damages increases over this time span, while the level of doses also falls into the range where there exists scientific uncertainty of resulting radiological health effects. Issues of discounting and the spatial boundaries used in the analysis thus become very important. Also, health impacts include hereditary effects, a category not encountered in health impacts in other fuel cycles. Thus, participants in the multi-criteria scoring exercise are given the opportunity to reflect their judgement on these issues by providing the opportunity to score impacts of the nuclear fuel cycle on a separate basis. Table 29 gives results for the different scenarios based on a central estimate of 0.69 mEuro/kWh (0% discounting, impacts studied over 10000 years). Score (0-100): 0=2500 ; 100=500 (MEuro) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

12

79

79

63

13

78

78

73

Worst score

12

79

79

63

13

78

78

73

Table 29. Radiological health impacts of nuclear fuel cycle

Need for very long term management of high-level waste The output from fission power includes chemical and low-level radioactive wastes, as well as used spent fuel, which is the main disposition challenge. Spent fuel from nuclear power reactors contains approximately 5% fission products (atoms produced by splitting another atom or by radioactive decay of another fission product), of which approximately 10% are long-lived fission products (LLFP); and 1% plutonium and other minor actinides (MA)577. Together, these products are responsible for the radiotoxicity on a very long timescale. Table 30 summarises the cumulative production of MA and LLFP (1990-2050) in the different scenarios. MA and LLFP resulting from the import of nuclear electricity from foreign countries are taken into account. Scenarios which incorporate a nuclear phase out score equal. The differences between the M.K.LCS and the R.K.LCS scenario relate to the different technologies for nuclear electricity production.

577

The exact figures used were, for the LWR: 22 kg Pu/TWh, 4 kg MA/TWh and 12 kg LLFP/TWh; and for the MHTGR: 15 kg Pu/TWh, 2 kg MA/TWh and 9 kg LLFP/TWh (Charpin et al., 2000).

408

Chapter 6

Score(0-100): 0=125 ; 100=0 (ton)

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

1

59

59

46

19

59

59

55

Worst score

1

59

59

46

19

59

59

55

Table 30. Need for very long-term management of high-level waste (HLW)

Impacts on natural ecosystems caused by air pollution Effects of electricity generation on natural ecosystems are very difficult to quantify in monetary values. Ecosystems may be damaged by fuel cycles in a number of ways, but the most serious and widespread effects are caused by acidic and nitrogenous deposition and photo-oxidants. Ideally, one should identify various types of ecosystems in the neighbourhood of the power plant(s) under consideration, and estimate the exceedance of critical loads or levels of pollutants in these ecosystems caused by the emissions of the power plant. Such a detailed analysis could not be performed for this exercise. As a proxy for the acidification damages caused by the power generation stage, we have calculated a composite indicator based on the SO2 and NOx emission levels. SO2 emissions have been given a weight of 2 compared to NOx emissions, since the former provide for two acid equivalents in the acidification process, while the latter provide maximum one (so in fact, we calculated a conservative estimate). Results for the different scenarios are shown in Table 31 (cumulative emissions over the period 1990-2020 and 2020-2050). Score (0-100): 0=5000 ; 100=500 (kton acid eq.) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI


14

13

11

14

10

9

10

10

Worst score

11

9

10

12

9

6

8

8

Best score

98

43

85

95

100

49

93

90

Worst score

95

0

80

92

93

5

88

84

2020-2050

Table 31. Impacts on ecosystems caused by air pollution

Environmental impacts caused by solid waste from coal fuel cycle The treatment of solid waste matter and by-products from the coal cycle deserves special mention, since these impacts have not been captured in external cost calculations. The following types of wastes can be produced: mine wastes (accumulated at surface in slag heaps), furnace bottom ash (FBA), pulverised fuel ash (PFA), flue gas desulphurisation


409

sludge, and gypsum. However, impacts associated with waste utilised elsewhere (which should rather be referred to as by-products) should be considered as part of the system to which they are transferred from the moment that they are removed from the boundaries of the fuel chain (e.g. FBA and PFA can be used by construction industry). It is of course important to be sure that a market exists for any such by-products. The capacity of, for example, the building industry to utilise gypsum from flue gas desulphurisation systems for coal power plants is clearly finite. If it is probable that markets for particular by-products are already saturated, the ‘by-product’ must be considered as waste instead. A further difficulty lies in the uncertainties about future management of waste storage sites. For example, if solid residues from a power plant are disposed in a well engineered and managed landfill there is no impact (other than land use) as long as the landfill is correctly managed; however, for the more distant future such management is not certain. Also, the impact of liquid effluents on the receiving surface waters has not been quantified, but can play an important role in the mining and production stage. As a rough comparative proxy for all these impacts, we have calculated the cumulative use of coal578 (1990-2050) in the different scenarios (cf. Table 32). Score (0-100): 0=35000 ; 100=10000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

88

55

88

87

83

47

86

84

Worst score

83

3

84

84

73

5

80

75

Table 32. Environmental impacts caused by solid waste from coal fuel cycle

Catastrophic risk: nuclear No attempt was made to quantify the catastrophic risk associated with nuclear power use, neither in terms of probability of occurrence multiplied by expected number of casualties, nor in monetary terms (expected damages). Rather, we have simply calculated the total amount of TWh produced by nuclear power plants in the different scenarios as a (rough) comparative measure of the risk of actually experiencing a catastrophic accident (all else being equal – e.g. we do not take into account the reported lower probability of occurrence for new reactor types such as the modular high-temperature gas reactor). It is left up to the participants in the multi-criteria scoring to weight this nuclear risk according to their own framing of the issue (e.g. which might include other dimensions such as irreversibility of the risk, gravity of single events, issues of intragenerational equity, control, trust in the institutions responsible for managing the risk, etc.). Electricity generated by foreign nuclear power plants is included in Table 33. Fatalities in other fuel cycles are covered by the ‘occupational health’ criterion. This way, we were able to make

578

1 PJ of coal equals some 40000 tons of coal approximately.

410

Chapter 6

a distinction between effects on individuals and effects that potentially threaten society as a whole. Score (0-100): 0=3500 ; 100=0 (TWh)

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

7

61

61

49

8

61

61

57

Worst score

7

61

61

49

8

61

61

57

Table 33. Catastrophic risk (nuclear)

3.4.3.2 Economy Intensity of energy use Table 34 summarises the intensity of energy use for the different scenarios. Since it is assumed that all scenarios attain equal economic growth, final energy use (cumulative over the period 1990-2050) is a suitable indicator for the intensity of energy use (defined as final or primary energy per unit of GDP). Intensity of energy use can be seen as a criterion to measure the general ‘rationality’ (in the logic of the industrial/market commonwealth) of the ways in which society makes use of energy resources, e.g. avoiding unnecessary squandering. Score (0-100): 0 = 120000 ; 100 = 90000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

57

59

67

63

76

83

91

84

Worst score

17

18

33

24

38

46

53

47

Table 34. Intensity of energy use

Security of energy supply Imports of gas and oil can be used as a rough indicator for security of energy supply. Security of supply is most commonly defined as the robustness of the Belgian energy system to ensure that Belgian citizens will not be exposed to shortages of energy and that Belgium is less vulnerable to international policy or conflicts in this area. Therefore, besides keeping import of energy resources to a minimum, this criterion also reflects a concern for regional diversification of import sources (especially with regard to regions which might be vulnerable to risks of erupting wars or political upheavals). It is generally accepted that both the coal and the nuclear fuel cycle are less vulnerable to this aspect.


411

Score (0-100): 0=110000 ; 100=85000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

68

40

20

46

95

88

70

73

Worst score

21

12

0

0

58

52

27

35

Table 35. Security of energy supply

Table 35 summarises the cumulative reliance (1990-2050) on imports of gas and oil for all scenarios. Imports of electricity produced by gas are taken into account. ‘Rational Perspective’ scenarios score better than their ‘Market Driven’ counterparts. Scenarios which allow for a nuclear continuation score better on this criterion. Total system costs Table 36 summarises the total annualised system costs for the different scenarios (in 2010, 2030 and 2050), expressed as a percentage of the Belgian GDP in 2000. Total system costs include all cost parameters modelled in MARKAL, i.e. investment costs for demand and supply technologies, fuel costs, delivery costs for fuels, operating & maintenance costs, variable costs for power plant operation, etc. – for all sectors of the energy system. Score (0-100): 0=5% ; 100=0% of GDP 2000 579 MKLCS MKPCS MKPLCS MKPLCSI RKLCS RK.PCS RKPLCS RKPLCSI

2010

91

91

91

91

96

96

96

95

2030

88

73

65

66

94

83

81

79

2050

62

55

0

24

88

78

69

62

Table 36. Total system costs580

The reported costs are relative with regard to the respective baseline scenarios (‘Market Drive’ or ‘Rational Perspective’) and thus represent the additional cost in order to meet GHG emission requirements. Therefore, ‘Rational Perspective’ and ‘Market Driven’

579

GDP2000 equals 247.7 billion Euro. Actually, the total system costs for M.K.P.LCS in 2050 are 8.6% of GDP2000, but because all other costs fell in the range of 0-5% we have chosen to treat this result as an (unrealistic) outlier and have attributed it with a score of 0. Compared to Proost and Van Regemorter (2000), these costs are rather low. These authors foresee a cost of 2.7% of GDP2000 for a post-Kyoto scenario (-15% GHG emissions in 2030) with a nuclear phase out (as planned), while our M.K.P.LCS scenario (which comes closest to the assumptions used by Proost and Van Regemorter) this cost would be 1.7% of GDP2000 in 2030. However, costs in our case rise rapidly after 2030. 580

412

Chapter 6

scenarios are not directly comparable, since both scenario groups differ in the discount rates used. The lower discount rate of the ‘Rational Perspective’ scenario group implies better anticipation of the future conditions facing the energy system, and hence a different investment behaviour. Scenarios where nuclear energy is phased out, with a low potential for carbon capture and storage and no import of foreign electricity (P.LCS) generally show higher economic costs to meet the CO2 emission reduction demands in 2050. Also, generally higher costs are incurred in the period 2030-2050, since much of the existing energy infrastructure will have to be replaced by then. Marginal cost of electricity The MARKAL model also calculates the marginal costs for different energy carriers, i.e. the delivery cost for one extra unit of the energy carrier under consideration. Also for electricity, this cost is calculated (without consideration of cost of distribution or a margin of profit for the producer). Results for the average yearly marginal costs of electricity for the periods 1990-2020 and 2020-2050 are given in Table 37 for the different scenarios. The rapid increase in marginal costs of electricity in the period 2020-2050 is explained by the absolute limits imposed on CO2 emissions: every additional kWh must be produced without adding significantly to the global carbon budget (especially in the ‘Market Driven’ scenarios, were demand-side measures are less readily available than in the ‘Rational Perspective’ scenarios), which implies some rather radical changes in the technologies used for electricity generation and the use of existing installations. Actually, in the MPLCS scenario marginal costs were so prohibitively high that we have chosen to attribute a score of 0 for this scenario. Score (0-100): 0=0.15;100=0.03 (Eurocent/kWh) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI


88

88

82

92

93

93

93

96

Worst score

77

75

68

79

86

86

86

92

Best score

50

65

0

27

83

76

56

67

Worst score

44

61

0

15

80

74

50

62

2020-2050

Table 37. Marginal costs of electricity

3.4.3.3 Social, political, cultural, ethical needs Need for intermediary storage of spent fuel Spent fuel from nuclear power plants will have to be stored for several decades in order to cool down, also implying that appropriate institutions for control have to be kept in


413

place. Table 38 summarises the total amount of spent fuel produced in the different scenarios. Score (0-100): 0=7500 ; 100=3000 (ton)

MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

5

87

87

87

35

87

87

87

Worst score

5

87

87

87

35

87

87

87

Table 38. The need for intermediary storage of spent fuel

The R.K.LCS scenario scores better than the M.K.LCS scenario because the hightemperature gas reactor (deployed in this scenario) operates at a higher burn-up rate than the standard light-water reactor technology. All scenarios with a nuclear phase out have an equal score on this criterion. Use of non-renewable resources The non-use of non-renewable resources (oil, gas, coil and uranium) can be regarded as a measure of leaving development opportunities open for other (Third World) countries. Table 39 summarises the cumulative use of non-renewable resources (1990-2050) in the different scenarios. Score (0-100): 0 = 135000 ; 100 = 100000 (PJ) MKLCS MKPCS MKPLCS MKPLCSI RKLCS RKPCS RKPLCS RKPLCSI

Best score

81

38

49

40

97

66

81

64

Worst score

45

0

30

6

67

33

48

33

Table 39. Use of non-renewable resources

Degree of decentralisation of electricity sector Table 40 summarises the installed capacity of decentralised electricity production technology (mainly cogeneration and photovoltaics) in 2050 as a percentage of the total installed capacity in Belgium. Decentralised electricity production is defined as production that is fed directly into local power grids. Decentralisation of the electricity sector might be considered to be important for a number of reasons – e.g. making energy production more ‘visible’ to the users, enabling local participation in energy projects, limiting the powers of large energy companies, etc. Thus, this criterion will tend to overlap with other criteria taken from the third category, although it is not exactly identical (e.g. a large energy company might still be the owner of most of the decentralised power units). Offshore wind power is not counted as a decentral technology. Also, the scores reported in

414

Chapter 6

Table 40 relate to the installed capacity and thus not necessarily to the amount of electricity effectively produced by these decentralised technologies. ‘Rational Perspective’ scenarios generally allow for a higher percentage of decentralised electricity production technologies. Score (0-100): 0 = 0% ; 100=100%

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

38

40

39

44

57

58

58

65

Worst score

38

40

39

44

57

58

58

65

Table 40. Degree of decentralisation of electricity production

3.4.3.4 Diversity of the electricity production park Diversity is generally seen as a good strategy to deal with the many incertitudes facing an energy system. Discussion of diversity in the energy debate has tended to remain confined to the issue of uncertainty concerning the security of future energy supplies. But new insights (Stirling 1994) also relate diversity to a more general protection against ignorance, for instance in financial or environmental performance, and even as a potential way to reconcile different values and interests related to controversial issues. As a measure of diversity, Stirling proposes to use the so-called Shannon-Wiener function581. Table 41 summarises the value of this diversity function (in 2050) for the different electricity production parks (as a result of the different choices made in the scenarios). For the calculation of this function, we used the following broad categorisation of electricity generation options: nuclear, gas, coal, cogeneration, renewables and import. Score (0-100): 0=0.7 ;100=1.5

MKLCS

MKPCS

MKPLCS

MKPLCSI

RKLCS RKPCS

RKPLCS

RKPLCSI

Best score

82

63

40

85

67

22

63

96

Worst score

79

31

24

81

46

8

59

91

Table 41. Diversity of electricity production park

581

H = -∑ pi ln pi , with H the value taken by the diversity index for a mix of options, pi the proportional reliance on option i and ln the natural logarithm. This index captures both the variety (the number of options) and balance (the relative importance of the different options in the mix).


415

4 Summary and conclusions In this chapter a first attempt has been made to achieve some tractability in the otherwise ‘unstructured’ debate on which direction a sustainable development of the energy system should take. The first necessary step in this process entails arriving at an agreement on the list of ‘actants’ to be taken into account in deliberations. On this point, while we concede we cannot give absolutely watertight guarantees that we have fully covered all possible points of view, we are still confident that our combined value tree is reasonably comprehensive, as it is based on the inputs of the different perspectives which have been most ‘vocal’ in the general energy debate. In addition, the process by which the tree was constructed (i.e. a bottom-up interview approach) is likely to provide it with some political legitimacy. However, in order to function as a ‘boundary object’, the combined value tree should also meet scientific quality criteria. Assuring scientific legitimacy depends upon requirements such as the logical coherence and non-redundancy of the value tree structure, and the general pertinence, fidelity, specificity and sensibility of the associated indicators. While we have tried to meet criteria of logical coherence through the use of Boltanski and Thévenot’s commonwealth model as a structuring framework, (quantitative) indicators have at this stage only been proposed for some of the most ‘uncontroversial’ criteria. Further work should thus include a discussion about which of the criteria are most pertinent for being measured; how they could be measured and perhaps how different measurement should be aggregated (Chapter 7 will provide some further indications on this topic). In a second step, we constructed a number of scenarios covering a broad range of strategic options for the future of the Belgian energy system. This scenario approach allowed to give an idea of the magnitude of the impacts associated with these strategic options under two different worldviews (one stressing ‘market’ values, the other ‘civic’ values). Again, here we took care to tailor our scenario approach to the needs of different parties involved in the discussion. For instance, we relied on the most simple form of engineering-economic modelling without making use of increasingly complex macroeconomic modelling tools. Choosing this approach might perhaps make the exercise less credible in the eyes of experts in econometric modelling, but, in return is likely to increase its overall transparency. While both the value tree and the scenarios were discussed in a workshop, we chose to further test their capacity for functioning as boundary objects in the ‘unstructured’ debate on sustainable energy policy in a subsequent individual multi-criteria exercise. The next chapter gives the details and conclusions following from this exercise, as well as suggesting areas where common ground can be found.


I.

High-level criterion

Intermediate level criterion

Low-level criterion

Environmental & human health and safety

Air pollution

Impacts of air pollution on human health: mid-term Impacts of air pollution on human health: long-term Impacts on occupational health (gas+coal) Radiological health impacts (nuclear) Need for long-term management of HLW Visual impact on landscape Noise amenity Impact on natural ecosystems (air pollution): mid-term Impact on natural ecosystems (air pollution): long-term Environmental impact from solid waste (coal) Land use Water use Catastrophic risk: nuclear Geographical distribution risks / benefits

Occupational health Radiological health impacts Aesthetic impacts Other environmental impacts

Resource use Other energy related pressures

II.

Economic welfare

Overall economic benefit

Producer need/benefit

Consumer need/benefit

Intensity of energy use Security of energy supply Distribution of economic benefits / burdens Economic risks Overall cost energy system: mid-term Overall cost energy system: long-term Ability to provide specialist market Required investment in supply technology: mid-term Required investment in supply technology: long-term Net expenditure on fuels: mid-term


International co-operation Need for government intervention III.

Social, political, cultural and ethical needs

Individual/consumer choice/benefit

Consumer choice Citizen participation Contribution to rational energy use

Institutional needs/benefits

Degree of decentralisation Need for intermediary storage of spent fuel Control and concentration of power Influence on political decision-making Need for socio-political stability Need for direct political intervention Reversibility of technology choice Knowledge specialisation Need for institutional non-proliferation measures Potential for technology transfer Leaving resources for development Equity (general) Job opportunities in the energy system

Development opportunities

Jobs IV.

Diversification

Net expenditure on fuels: long-term Marginal cost electricity: mid-term Marginal cost electricity: long-term Strategic factors for export Compatibility with international R&D agenda Amount of direct or indirect subsidies needed

Diversification

417

CHAPTER 7 LONG-TERM

OPTIONS FOR

BELGIAN

ENERGY POLICY

– A

MULTI-

CRITERIA MAPPING EXERCISE In this chapter the functioning of the combined value tree and long-term energy scenarios (developed in chapter 6) as ‘boundary objects’ in the sustainable energy debate, with different lines of accountability and legitimacy to the actors involved, is put to the test. For this, we have relied on a technique called ‘multi-criteria mapping’ (MCM). This technique relies on some of the tools developed in the broad research tradition of multi-attribute utility theory, but it nevertheless applies these tools in a fundamentally different way. The approach is called ‘multi-criteria mapping’ precisely because the intention is to demonstrate how a debate on a controversial technological development can be ‘mapped’ – i.e. by establishing the main contours of the debate and identifying key areas of convergence and divergence – rather than arriving at a ‘consensus’ solution. The MCM exercise, as applied here, consisted of four main stages: attribute selection (in which actors identified around ten main areas of concern for sustainable energy policy); attribute rating (in which an assessment is made of the relative importance of the different concerns); option evaluation (in which the actors are asked to score the different energy scenarios for each of the attributes); and experimentation (in which actors are encouraged to explore and modify their point of view based on a graphical and tabular representation of the results). This approach allows us not so much to draw conclusions on the general desirability of the different long-term options for Belgian energy policy, but rather on the suitability of the proposed ‘tools’ as a support for long-term sustainable energy governance.

1 Introduction By way of a catchphrase, constructivism is often said to be all about establishing the links between different hybrid ‘objects’ (called ‘actants’) forming a socio-technical network. However, this (and similar) catchphrase(s) tend to omit that the constructivist vantage point equally allows an analyst to draw the attention towards the (intentionally or unintentionally) non-realised links. Taking into account both faces of the constructivist coin has been our main concern so far, most notably in chapter 3 (where we described the necessary framing operations to arrive at an easily comparable external cost figure for different electricity production technologies) and chapter 4 (where we investigated the network drawn up in the ‘official’ policy discourse and the perspectives of different stakeholders in the debate). Based on insights gained from these investigations, we have argued (in chapter 5) in favour of setting up a governance structure explicitly aimed at collective experimentation and learning. As we explained, this does not mean that we have to (or, for that matter, can) do away with all forms of societal struggle on the desired sociotechnical changes. On the contrary, resistance might sometimes be useful and productive, because it forces actors to develop new strategies and articulate choices. One might even

420

Chapter 7

speculate that if economic development would have been allowed to proceed ‘smoothly’, without any resistance by the environmental movement, perhaps we would not have been talking of sustainable development now in the first place. But there are also instances where ‘antagonistic’ interactions and decision-making structures lead to entrenched positions hampering the potential for learning to occur. Arguably, this has been the case for the (nuclear) energy debate in Belgium (and many other countries)582. Nevertheless, distinguishing between ‘productive’ and ‘unproductive’ forms of learning is a difficult (if not impossible) assignment, especially since according to the central constructivist tenets there can be no a priori criteria for deciding in which direction learning should occur, or whether it has gone far enough in that direction. Arguing in favour of facilitating learning through the alignment of important social actors (as we propose to do) thus inevitably represents a value-laden choice. In our case, this choice (grounded on political philosophy – most notably Habermas – and constructivist insights – most notably Latour’s “Politics of Nature”) entails the attitude of being as independent as possible from specific values, interests or rigid frameworks when structuring policy problems, anticipating possible future impacts, designing long-term strategies for sustainable energy development, and/or implementing concrete policy measures (cf. Chapter 5). Constructivism (as we propose it here) does not so much represent a thesis, a belief or a theory, but rather an attitude towards the role of beliefs, theories, and theses in science and technology studies (cf. Van Brakel 1998, p. 58). Actor alignment (as a precondition for ‘productive’ learning) can of course not be simply assumed to happen; indeed, a lot of ‘construction’ work has to be put into operation. Framing the question of nuclear energy’s possible future role in the wider context of the sustainable development principle (and subsidiary principles such as precaution) is a first and necessary – albeit largely insufficient – step in the direction of alignment. Since the sustainability concept enjoys a broad moral support (admittedly, because of its diffuse nature), one can simply assume its validity and commence negotiations on this ground. The picture of course becomes different when more concrete long-term strategies for sustainable energy provision have to be drawn up. In chapter 5, we have argued that for this task, learning processes could be facilitated by the creation of ‘boundary objects’ (i.e. common nodes in the networks of the actors involved in the negotiations) and specific loci for discussion (e.g. a new ‘energy agency’, intermediary discussion platforms on key technologies, etc.). The present chapter is concerned with an attempt to evaluate to which degree the combined value tree and scenarios developed in chapter 6 are able to meet this goal. We will do so by setting out a ‘similarity measure’ aimed at measuring the ‘distance’ that remains to be covered in order to arrive at a productive alignment. As an introduction to the analysis contained in the following sections, it might be a good idea to revisit some of the constructivist basics. When proposing a strategic orientation for long-term sustainability in the energy field, social actors will essentially rely on the

582

The argument has been developed in more depth in Laes et al. (2004c), but the examples given in chapter 2 already give a general taste of the degree of polarisation as a result of historical interactions.

Long-term options for Belgian energy policy – A multi-criteria mapping exercise

421

recruitment of a number of ‘allies’ able to support their vision. These ‘allies’ can take numerous forms: a technological artefact (e.g. a power plant), a moral principle, existing legislation, a scientific study, etc. In the collective negotiations, each actor will thus exert force by positioning himself as the representative of a network of hybrid ‘actants’. Bringing different actors (and the associated networks) together thus introduces a dynamic which can be analysed in terms of convergence or divergence. A network is convergent if the following conditions hold (slightly modified from Callon (1995, p. 315)): 1. there is agreement on the list of ‘actants’ to be taken into account during the negotiations; 2. there is agreement on the way these ‘actants’ are represented (e.g. a spokesperson from a major energy company might likely have a different representation of consumer habits than a spokesperson of a consumer organisation or an environmental NGO); 3. there is agreement on the future direction(s) the socio-technical network being conceived will or should take. Of course, convergence always remains a matter of degree and can be applied in different ways. For instance, when applied to the choice of actors represented at the negotiation table, recruitment of FRDO members for the present exercise (and possibly also in the new ‘energy agency’) can be defended on the grounds that these actors, through the force of earlier conventions, have acquired a certain ‘legitimate voice’ in the debate. Deciding on what will be negotiated is already a more difficult endeavour. Although it is based on the interviews held with the (representatives) of FRDO members, the combined value tree proposed in chapter 6 still is a new ‘object’ the introduction of which in the networks represented by the different actors could possibly be resisted. This danger is even greater for the scenarios we developed, since these rely on simplifications and necessary interventions to make them more tractable for formal modelling. Another facet in the use of decision-analytic models (such as value trees and scenarios) is that the models feed into the power plays that are simply pervasive in policy making. To acknowledge the diversity of (possibly conflicting) perspectives at stake, we thought it would be more pertinent to carry out separate analyses for each of the stakeholders, especially since the interviews (cf. Chapter 4) already revealed us that there is little chance of developing a commonly accepted framework583. The key benefit of these separate analyses is that they: 1) explicate those concerns that matter most to the different stakeholders, and 2) expose the issues where disagreements are largest. At best, such exploratory uses of decision analysis clarify how problems are perceived and suggest directions where common ground can be gained (Salo 1999). 583

Although the value tree and the scenarios were discussed in a workshop, this workshop was oriented more towards developing an understanding of the approach rather than eliciting possibly conflicting personal statements on its general usefulness.

422

Chapter 7

Schematically, four different cases of agreement/disagreement can be identified (cf. Figure 18). Case 1 corresponds to a situation of perfect agreement. The consensus applies both to the expected (or desired) socio-technical networks (the scenarios) and the list of ‘actants’ present at the negotiation table. The ‘strategic’ network in this case is ‘convergent’, and can be taken up for consideration by policy makers when designing policy measures. Case 4 is that of total disagreement concerning both the definition and analysis of the present situation and the expected (or desired) socio-technical networks related to them. Cases 2 and 3 represent cases of partial disagreement which give rise to different points of view.

Agreement on future actants’ association to worlds (scenarios)

Agreement on the list of ‘actants’

Non-agreement on future actants’ association to worlds (scenarios)

1.

2.

3.

4.

Non-agreement on the list of ‘actants’

Figure 18: The possible dynamics resulting from the confrontation of different networks (Source: Callon 1995, p. 316)

Figure 18 makes clear that the degree of alignment is always the (provisional) result of a process which leads to an agreement or disagreement on the description of the (anticipated) socio-technical network and the role and duties of the involved ‘actants’. If the starting point is complete disagreement (as seems to be the case for us), agreement can only be reached after a (lengthy) negotiation process during which both the definition of the sociotechnical networks and the list of actants have to be adjusted and modified. The dynamics of these adjustments crucially depend on the co-ordination rules (i.e. who negotiates with whom, in what order, for which purpose, according to which sequences, etc.). As Callon (1995, p. 316) points out, the distribution of competence is also crucial: who are the ‘spokespersons’ competent to speak for the ‘actants’, and can they be substituted (and to what degree?) by other actors in the debate. With regard to these issues we immediately have to put in a caveat against unrealistic expectations for the present exercise. Indeed, one of the basic limitations of our


423

‘experimental’ effort is that it cannot take into account the crucial importance of government as an initiating actor, in the sense that (as advocated in chapter 5) we consider government to be a privileged actor for setting up a (tentative) institutional frame and trying to formulate an overall orientation based on stakeholders’ heterogeneous views. How this impacts on the validity of the exercise will be discussed under section 4. For the moment, we propose to suspend any objections of this kind and to proceed with the development of the ‘similarity measure’. Section 2 discusses the particular methodology adopted for this purpose (MCM). In sections 3 and 4, the results of the exercise will be analysed with regard to content and process. Section 5 summarises the lessons learnt from this exercise.

2 Methodology A wide variety of decision-analytic techniques exists and has been used to some extent in this dissertation. The most important general approaches include policy analysis (cf. Chapter 2 and 4), life cycle analysis and environmental (or social) impact assessment – whether or not coupled to economic valuation (cf. Chapter 3), probabilistic risk assessment, and multi-criteria evaluation584. From our previous discussions, it is clear that each of these approaches tends to have its own strengths and weaknesses concerning issues such as representativeness, feasibility and accountability; as well as raising concerns about ‘manipulation’ or capture of the results of such techniques by special interests. As we have tried to argue in the preceding chapters, our aim is not an outright rejection of such formal approaches to decision analysis, but rather to find an appropriate ‘negotiated’ niche of application for some of them. The ExternE method (cf. Chapter 3) for instance has been shown to offer a very coherent and consistent framework for approaching energy policy questions, albeit that it remains relatively inflexible and narrow in its scope. Therefore, its niche of application seems to be limited to more specific questions regarding policy measures in specific fields (e.g. energy taxation, standard-setting, etc.) rather than to the broad strategic deliberations on sustainable energy system development. For this specific niche of more interest to us, we propose a deeper investigation of the ‘market potential’ presented by multi-criteria evaluation methods. The common ‘multi-criteria evaluation’ denominator actually captures a range of different methods, developed as a result of a long history of experimentation and theory building585. The main drive behind this research field has been the observation that the simultaneous consideration of quantitative, qualitative and often even purely intuitive aspects of a decision problem usually presents a difficult (if not impossible) task for an unaided human mind.

584

Setting aside the various formal social scientific approaches such as consensus conferences, citizens’ juries, focus groups, etc. which have been increasingly used in the context of complex socio-technical decisions (see e.g. Joss and Bellucci (2002) and Slocum (2003)). 585 Multi-criteria appraisal has its roots in military logistics and operations research during the Second World War. Keeney et al. (1976) is generally considered to be one of the first ‘handbooks’ providing multi-criteria appraisal with a solid theoretical background (multi-attribute utility theory).

424

Chapter 7

Individuals (including experts), when confronted with complex problems, become liable to a number of cognitive biases586. Multi-criteria evaluation aims to avoid these biases through a reliance on a variety of carefully designed formal decision-making tools. As such, it is often informed by a highly technical literature on rational choice or utility theory (see e.g. Hobbs and Meier 2000; Colorni et al. 2001), and the tendency has been for such techniques to become increasingly complex over the decades of theory building. Nevertheless, no single technique or method has been universally accepted. Rather, techniques are adapted to the requirements of the decision context at hand. For instance, various forms of decision analytic models based on multi-attribute utility theory have been used to study a variety of energy policy issues in the last decade, including the meaning of sustainability in the energy sector (see e.g. Jones et al. 1990; Hämäläinen 1990; Hämäläinen and Karjalainen 1992; Afgan et al. 2000; Haldi et al. 2002). It is not our intention here to go into the details of all of these models – they are reviewed in great detail in a number of publications (e.g. Salo 1999; Hobbs and Meier 2000; DTLR 2001). Put (very) briefly, all forms of multi-criteria appraisal involve multiplying a performance score under an individual appraisal criterion with an importance weighting assigned to that criterion. The overall rank derived for each option is then derived as a result of a weighting procedure applied to the scores. The main point of concern for our present purposes is however how such multi-criteria assessment procedures are used in a decision-making context. Firstly, multi-criteria assessment is often used as a means to derive a single ‘best’ solution to a complex problem (in a quite similar fashion to ‘traditional’ cost-benefit or risk analysis). However, it can be shown on ethical grounds that multi-criteria methods, when applied as a decision-making method, necessarily rely on utilitarian presuppositions and therefore cannot provide us with a standard which enables us to reach a decision in a situation of profound ethical disagreement (Rauschmayer 2000). A second point of concern is that, in practice, the criteria in a multi-criteria appraisal exercise are established according to the requirements of ‘quantitative’ sciences (e.g. ecology or economics), and not for instance to the requirements of ethical reasoning (e.g. what is the meaning of inter- and intragenerational ethics?). This approach often seems to be motivated by a concern for avoiding deep-seated value conflicts. For instance, multi-criteria discussions on sustainable development are often framed in terms of ‘technical’ criteria such as ‘environmental impacts’, ‘social impacts’, ‘economic impacts’, etc. (see e.g. Afgan et al. 2000; Haldi et al. 2002). However, as demonstrated by Boltanski and Thévenot’s commonwealth model, 586

Some of the most well-documented biases include (Kahneman et al.1982): ‘anchoring’ (in which undue weight is given to a conventional value, first value given, or information from previous assessments); ‘availability’ (the tendency to give too much weight to readily available data or recent experience); ‘coherence’ (the tendency to assign probabilities to events on the basis of the ease with which coherent accounts can be constructed of how those events could arise); ‘motivational’ (arising from various sources including a desire to influence a decision, a desire to appear knowledgeable or authoritative); ‘overconfidence’ (over-estimation by an individual of their ability to make quantitative judgements), ‘representativeness’ (the tendency to place more confidence in a single piece of information that is considered to relate directly to the issue under consideration than to a larger body of more generalised information); ‘satisficing’ (sacrificing comprehensiveness to expediency by considering only a limited number of options); ‘unstated assumptions’ (if the question answered is not identical to the question asked or assumed to have been asked).


425

establishing comparisons on a technical basis reflects in itself a deep link to a value system concerned with efficacy, performance, and functional exigencies (i.e. the ‘industrial commonwealth’). Finally, multi-criteria exercises are often conceived and concretely presented as individual bounded projects (i.e. arriving at a definite decision at one point in time), rather than as permanent running themes in a social learning process (as we envisage) (Stirling 1997, p. 192). However, as Stirling and Mayer (1999) remind us, there is nothing intrinsic about the multi-criteria methodology which would direct its use to impose an ‘analytical fix’ on a complex political controversy, as a way of determining a single ‘objectively’ best solution to a decision-making problem. In reviewing the literature on multi-criteria appraisal, it becomes clear that, when reduced to a more ‘straightforward’ form (cf. infra), these methods offer the potential for developing a more flexible decision-aiding tool. An example of such more ‘straightforward’ approach is provided by the work by Stirling (1997)587 in an embryonic form, and more fully developed in Stirling and Mayer (1999, 2000)588. This work shows how multi-criteria techniques can be used as a ‘heuristic’ (rather than an ‘analytical fix’) – i.e. a way of exploring the key dimensions of a controversial issue and establishing their characters, relationships and relative importance. The particularities of the MCM method adopted in this chapter are discussed more fully in Annex 4 and in the following sections of the present chapter. In short, the basic philosophy behind MCM is that multi-criteria methods – through the explicit separation of the concepts of relatively ‘technical’ scores and more openly ‘subjective’ weightings – establishes an framework for the treatment of other appraisal dimensions, such as the scope of analysis, the framing of crucial assumptions in appraisal and the treatment of uncertainties. In this way, multi-criteria analysis tools can be harnessed for the purpose of establishing the main contours in a debate and clarifying key areas of dissent and/or convergence (hence, it is called ‘mapping’). Before discussing the MCM methodology in more detail, it is perhaps useful to review the basic qualities which Stirling (1997, p. 195) claims to be upheld by MCM in comparison with other appraisal methodologies (and which have also drawn us towards adopting this approach for our purposes): • Flexibility and breadth of scope – in the sense that MCM does not require the ‘artificial’ constraints on the type of issue or measurement that can be taken into account into the appraisal, thus permitting the application by different participants of whatever are judged to be the most effective or appropriate techniques for each ‘matter of concern’; • Pluralism – in the sense that MCM is not intended as an ‘analytical fix’ to complex socio-technical questions, but rather as a ‘heuristic’ help in systematically ‘mapping out’ 587

Concerned with a broad evaluation of energy policy options. Concerned with agricultural strategies, and more specifically the use of genetically-modified herbicidetolerant crops compared with other strategies for the cultivation of oilseed rape. 588

426

Chapter 7

the links between different framing assumptions and the associated points of view (or decisions); • Transparency – in the sense that the MCM method allows for an ‘audit trail’ clearly linking the results with the various inputs, assumptions and parameters used in the analysis; • Analytic discipline and rigour – in the sense that MCM is founded in the wellestablished disciplines of rational choice and utility theory, with an extensive literature developing ‘rules of good practices’; hence, the method is ‘scientifically credible’; • Openness to participation – in the sense that MCM can handle as inputs both technical information and ‘subjective’ framing assumptions, and therefore necessarily requires both technical expertise and wider deliberation; • Feasibility and efficiency as part of a governance process – in the sense that, if embedded in an appropriate institutional context, the above-mentioned qualities could allow MCM to play a role in a more open-ended learning approach to energy policy (and other controversial technological choices). Furthermore, the MCM method is not too demanding (in terms of the theory behind it), expensive to implement, unduly protracted, etc. Of course, our main goal for this chapter is to investigate to what extent application of MCM is actually able to uphold these high expectations. From a methodological point of view, MCM (as any other multi-criteria appraisal exercise) has to work through six steps (see e.g. DTLR 2001): identification of the stakeholders to take part in the exercise (Section 2.1.1); identification of options for action (i.e. establishing the decision context) (Section 2.1.2); identification of decision attributes (i.e. matters of concern) (Section 2.1.3); identification of empirical indicators for these attributes; assigning relative weights to the attributes; and analysis of the results (quantitative and qualitative). In practice, the process logically falls apart into two separate phases: model development (in which the multi-criteria model and the options are actually constructed) and model operation (in which the multi-criteria model is actually used to structure decisions in an interactive way). For the largest part, the ‘model development phase’ has been described in chapter 6 (where we have discussed the combined value tree and long-term energy scenarios constructed for the purpose of the MCM exercise); we will only come back briefly to some of the key issues here. The discussion in this chapter will be focussed on the actual operation of the MCM model, with an emphasis on the points of divergence with Stirling and Mayer’s (1999, 2000) way of applying the model. These authors originally developed the MCM method with the objective of being entirely situation-based, i.e. its application did not employ any knowledge except the subjective assessments of the user. This extends to the definition of options, the definition and selection of criteria as well as the scoring of criteria. As we will point out in the following sections, adopting this point of view shows some significant disadvantages, which have led us to consider alternatives.


2.1

2.1.1

427

Development of the MCM model

Identification of stakeholders

For the operation of the MCM model, we have chosen to continue our participation with selected members of the FRDO. In a way, the MCM exercise can be regarded as a more mathematically-formalised sequel to the interviews discussed in chapter 4, with however a shift in focus towards future-oriented scenarios (whereas the interviews mainly discussed present problems). In other (Latour’s) words, the interviews were mainly concerned with making an inventory of ‘matters of concern’, whereas the MCM exercise should be regarded as a first attempt at ‘putting them into order’. As explained in Annex 1, these individuals were approached on the basis of their wider interest in both sustainable development and energy (governance) issues. Often, this meant that they were representatives of protagonists in the current energy debate with large stakes in its outcomes. However, this was not always the case, as we also explicitly strived to involve other actors than the ‘traditional’ interest groups in the discussions. As such, each participant in the MCM exercise could be expected to hold a general knowledge of the issues raised in contemplating energy options and their general implications, whilst also sometimes holding specialist knowledge on particular issues. As a group, it was important to include a sufficient number of perspectives, so that no point of view would be excluded a priori. As becomes evident from Table A-1 included in Annex 1, at least one representative of the major stakeholder categories (environmental NGO’s, labour unions, employers’ organisations, electricity generators, academia and advisory bodies) has participated in the MCM exercise, with the exception of development NGO’s589. However, this table also tells us that, due to busy schedules or other exigencies, most organisations did not participate in all research steps (interviews, scenario workshop and MCM). Also, in some cases different representatives from the same organisation participated in the different research steps. Maintaining full participation by the same representatives would have been more desirable though, in view of ensuring participants’ understanding of the logic behind each of the steps. Each participant was supplied with written information about the MCM exercise. In order to respect conditions of anonymity, we have chosen to assign each participant with a letter code. These letter codes are used throughout this chapter in the analysis and presentation of the results. As repeated before, this particular selection of participants for the present limited pilot exercise should not be seen as a definite choice in favour of these groups for representing ‘society’s views on sustainable energy’ (e.g. in the new ‘energy agency’). Deciding which groups should be involved in a concrete governance initiative would be a matter of further research and negotiation.

589

The representative of the development NGO made it clear after the first interview session that he really did not consider his organisation to be involved in the questions that were of interest to us (i.e. the role of nuclear power from a sustainable development perspective in Belgium), so he declined further participation.

428

2.1.2

Chapter 7

Identification of options for action

One crucial part of decision-analytic methods is how the decision question under scrutiny is constructed (cf. our reflections in chapter 2). In the context of a long-term policy for (sustainable) energy development, however, it is clear that there is no ‘single’ decision involved, but rather a set of interlinked decisions, none of which taken on its own constitutes the policy, but which in combination produces a process which we could describe as a ‘strategy’. Furthermore, the (estimated or desired) success or failure of these strategies will crucially depend on the associated ‘worlds’ (in Boltanski and Thévenot’s terms), e.g. concerning public acceptance of technological innovations, expectations of energy-saving behaviour, etc. Nevertheless, in order to use a decision-analytic procedure, we need to represent clearly distinctive ‘options for action’ in a way that would allow participants in the exercise to choose between them. Hence, a possible conflict emerges between the ‘complexity of the real world’ and the ‘simplicity’ required for the purposes of decision-analytic modelling. In principle, there is no ‘right’ solution to this dilemma; one can only try to propose an ‘acceptable’ pragmatic solution. For instance, in an application of MCM to energy policy, Stirling (1997) proposes to limit the selection of ‘options for action’ to a ‘conventionally recognised and highly aggregated set of options’ (fossil fuels, nuclear power and renewable energy), whilst leaving the ‘framing assumptions’ for assessing these options (in terms of the future ‘worlds’ associated with these options) open to the participants in the MCM exercise590. Stirling’s view appears to be motivated by a concern that the multi-criteria analysis should not be unduly constrained or biased by an externally imposed framework. While this may be true, it is also clear that leaving the framing assumptions entirely open to the participants’ insights leaves the door wide open to strategic behaviour – i.e. a participant simply assumes that ‘the world’ functions in accordance with the requirements for his/her preferred option performing optimally. While Stirling would of course contend that the purpose of MCM is precisely to make such framings more transparent (and hence also possibly open to discussion at a later stage), we nevertheless see two fundamental objections. The first is that without at least proposing some scenarios as a common framework for communication and discussion, the MCM exercise is likely to simply reproduce existing positions and statements. Hence, it is unclear to us what the precise added value of a multi-criteria exercise might be in this case. Secondly, simply accepting these framing assumptions at face value implies that there is no possibility to check whether these assumptions are applied consistently and coherently to all options under scrutiny – an important advantage offered by a reliance on formal modelling. Jones et al. (1990) offer another interesting solution to the dilemma. In their decisionanalytic model, these authors propose the use of five contrasting energy-policy scenarios 590

The same general strategy is applied by Stirling and Mayer (1999, 2000) in an MCM appraisal of agricultural options (including or not including genetically modified oilseed rape). Here, six basic policy options (combinations of technological options and flanking regulatory measures) were defined, but participants in the exercise were free to define any new options they thought to be appropriate.


429

(in their case developed for the UK), drawn from the publications of a variety of different organisations engaged in energy policy. Using existing scenarios has the advantage that participants in the exercise will likely already be familiar with these scenarios, thus greatly facilitating communication and discussion. However, as mentioned before, at the time we organised our MCM interviews (between October and December 2003) long-term energy scenarios for the Belgian context were simply not yet available, so we had to develop our own scenarios591. As explained in chapter 6, our solution has been to develop four broadly conceived technological options592 and ‘test’ them against a background of two different (summarily narrated) ‘worlds’ (the ‘market’ and ‘rational’ worldview). Of course, the danger then exists that the model used for calculation (MARKAL) subsequently becomes the main focus of discussion, and diverts the attention from the aim of the MCM exercise as a means of exploring and representing different viewpoints on sustainable energy policy593. However, in the individual MCM exercise performed on the basis of these scenarios, ample time was foreseen for discussing the model, including assumptions and hypotheses. A further characteristic of the scenarios under scrutiny is that although the principal focus concerned the relative merits of nuclear power, this option was nevertheless put in the context of alternative options for meeting the energy (and not only the electricity) needs of the future in a sustainable way594. Our intention was thus not to make a specific pronouncement on the ‘sustainability’ of the nuclear option as such, but rather to evaluate its relative (i.e. in comparison with other possible long-term options) performance under a number of different perspectives. Admittedly, the options used for this exercise were a bit ‘stylised’. For instance, scenarios which included the ‘nuclear’ option relied on either the EPR option (a large centralised power plant) in the ‘market’ scenarios or the smaller PBMR in the ‘rational’ scenarios, while of course the actual variety of existing and future reactor types is much greater595. Similarly, other ‘nuclear’ options – e.g. a lifetime extension for the current power plants or deliberately limiting the installed nuclear power production capacity below the 8 GWe assumed now in the scenarios – have not been investigated. Furthermore, some of the other chosen options are actually more hypothetical than others – e.g. currently, carbon capture and storage for electricity production is not yet applied on an industrial scale anywhere in the world. All in all, our basic concern for this MCM exercise has been to encompass a wide range of possible strategies – without (at this stage) necessarily going into the details of each of them – providing a potentially useful

591

With hindsight (in view of the comments received on our scenarios – cf. Section 4.1.1), the approach taken by Econotec and Vito (2005) in a recent advisory report for the Belgian ‘Federal public service of public health, food chain safety and environment’ (DG Environment) would perhaps have been more appropriate. In this report, scenarios for the Belgian context were defined starting from the IPCC global scenarios, which were subsequently ‘translated’ to the EU and Belgian level. 592 To recapitulate: a continued reliance on nuclear power; development of carbon capture & storage technology; import of electricity; or more energy conservation combined with renewables and/or cogeneration technology. 593 A fear that – in view of the results of the exercise – proved not to be entirely unjustified… 594 In fact, out of the eight scenarios, six included a nuclear phase out. 595 Possibly even including cogeneration of heat and power or ‘tri’generation of hydrogen, heat and power in the so-called ‘high-temperature reactors’ (HTR).

430

Chapter 7

frame of reference in the current debate over sustainable energy futures. However, whether we have succeeded in doing so remains to be seen (cf. Section 4). 2.1.3

Identification of decision attributes

As mentioned in chapter 6, decision attributes were derived from previous interviews with members of the FRDO and a range of publications and policy documents in the field of (sustainable) energy policy. The decision attributes were subsequently structured into a ‘combined value tree’. Based on Boltanski and Thévenot’s ‘commonwealth model’, three hierarchical levels were defined, starting from a top level with the three commonwealths most implied in the debate (the industrial, market and civic commonwealth) going down to the 44 bottom-level criteria. In any case, this classification of the criteria does not affect the final results of the MCM exercise, but is simply a matter of convenience: the possibility was left open to participants to choose between a smaller selection of attributes at a higher level of abstraction at any time in the exercise. For the purposes of a multi-criteria appraisal exercise, it is important that individual criteria are independent in the sense that, although different criteria might be related in various ways (e.g. policies aimed at reducing carbon dioxide emissions generally also lower emission of other air pollutants), the associated assessments of performance do not depend on judgments of performance under other criteria (e.g. measuring carbon dioxide emissions can be done entirely independently of the measurement of any other air pollutant)596. We have tried to structure the ‘combined value tree’ in such way that this requirement was met. However, since the 44 bottom-level criteria were still phrased in a rather general way (particularly those relating to the ‘Social, political, cultural and ethical needs’), some degree of overlap could probably (at this stage) not be avoided597. As also mentioned in chapter 6, we already provided technical ‘measurements’ for some of the criteria. Because working with 44 criteria at the same time would be generally unfeasible, we asked participants in the exercise to select from this list about 10 to 15 criteria which seemed to be most important to them. During the exercise, participants could also add new criteria or criticise chosen measurements for some of the criteria (and possibly even suggest others).

2.2

Operation of the MCM model

The actual operation of the MCM model was framed in the context of individual interviews, which took place between October and December 2003. Interviews usually lasted between 1 and 2 hours. During the interview, an iterative process was undertaken, comprising: 1) a discussion of the scenarios developed for the MCM exercise; 2) a discussion of the combined value tree developed for the MCM exercise (with possibilities for clarification and specification of new criteria); 3) the scoring of the performance of each scenario under a selection of criteria; 4) the weighting of the criteria in terms of their

596

In Latour’s terms: each ‘matter of concern’ should have a different and independent ‘spokesperson’. Also as a result of the different framings of the same criterion adopted by participants in the exercise (as became apparent when questioned what a criterion precisely meant for them). 597


431

relative importance as ‘matters of concern’ to the interviewee. The entire interview was organised in an iterative and reflexive way, so that participants were for instance able to add further comments on the scenarios while they were scoring criteria, or add new criteria along the way. 2.2.1

Identification of empirical indicators for the attributes (‘scoring’)

Having selected a maximum of up to fifteen criteria from the combined value tree, participants were then asked to evaluate the performance of each of the scenarios on each of the selected criteria. In case no ‘technical’ score was provided for a selected criterion, participants were asked to justify the ‘scores’ which they assigned under their various criteria by reference to what might be called their broader expertise and experience as established professionals in the field of energy and/or sustainable development and/or as representatives of important institutional ‘protagonists’ in the wider debate on energy options for the future. In practice, participants could express a relatively high, medium or low performance under a particular assessment criterion simply by adopting an arbitrary cardinal scoring scale on a scale of 1 to 100. In such cases, participants were asked to provide explicit ‘anchor points’ for the assigning of scores, for instance by reference to a ‘zero risk’ as a maximum score. In all cases, a high numerical score had to correspond with a high performance and vice versa. During the individual MCM exercises, scoring was performed by comparing each individual scenario’s performance under each individual chosen criterion. In practice, ‘scoring’ often turned out to be the most difficult part of the entire exercise. Participants often engaged in a demanding deliberation process of comparing one option to all the others under a particular criterion. Typically, reference would be made to a number of framing assumptions before arriving at a certain score. Even then, participants often expressed a certain feeling of unease about scoring the different scenarios based on ‘intuitive’ understandings about qualitative criteria. It is at this stage of intensive deliberation that we also checked whether the different criteria (as defined by each participant) were mutually independent for practical purposes. The procedure adopted also allowed for participants to express uncertainties when scoring individual scenarios, by leaving open the possibility to assign both high (‘optimistic’) and low (‘pessimistic’) scores for each scenario under each criterion. Participants were also asked for the reasons why they chose a particular ‘optimistic’ or ‘pessimistic’ score, which allowed us to reveal e.g. the perception of technical uncertainties or an expression of context-dependent variability. Of course, where neither uncertainty nor variability was felt to be a factor, the ‘optimistic’ and ‘pessimistic’ scores could be identical. Using the scores as a basis, a lap-top computer running a simple spreadsheet software was used to perform a relatively straightforward ‘linear additive weighting’ multicriteria procedure (cf. Annex 4)598. Essentially, this involves simply taking the

598

Microsoft ® Excel 2002.

432

Chapter 7

performance scores assigned by the participants and multiplying them by the importance weightings which were assigned separately in the next stage of the exercise. The result of this procedure is a ‘ranking’, reflecting the overall performance of each scenario under all the criteria taken together, taking into account the relative importance of these criteria. These results were displayed graphically in real time for each participant as a ‘bar chart’, and numerically as a set of tables. The bar charts clearly indicated the relative performance of all scenarios, taking into account the ranking under both ‘optimistic’ and ‘pessimistic’ performance scores. The tables comprised a listing of all criteria selected (with the associated performance scores, including ‘best’ and ‘worst’ values); the chosen weights for each attribute on each of the three levels; and the total weighted scores for each option. Since one of the primary aims of a MCM exercise is to enable users to explore their perspectives on (sustainable) energy policy, the presentation of results in clear tables and graphs in real time was of great importance. This allowed participants to ‘experiment’ with the results, e.g. by examining what would happen if any scores or weightings would be changed. When any of these values were changed, the results were immediately recalculated so that the effect of the changes could be explored interactively. 2.2.2

Assigning weights to the attributes (‘weighting’)

The final step in the interview process was the assigning of numerical weights by each participant, reflecting the relative importance of each of their chosen appraisal criteria. These weightings should reflect how much participants care about the differences in scenario performance under each criterion. During the exercise, we explained to participants that the weightings they assigned to the selected criteria should not be chosen in an abstract sense. Indeed, multi-attribute utility theory requires that weights should be assigned based on a comparison between attributes of what a change from the ‘best’ to the ‘worst’ score for each attribute would imply. For example, the relative importance of ‘air pollution’ cannot be compared to the ‘overall cost of the energy system’ unless it is clear how much ‘air pollution’ and how much ‘overall costs of the energy system’ are involved. Using the same example, suppose that ‘air pollution’ was chosen as an important criterion (with in the ‘worst’ case a health impact of 25000 MEuro and in the ‘best’ case a health impact of 2500 MEuro) and given a weight of 100. If it was then subsequently considered that changing the overall cost of the energy system (from the ‘worst’ case of 8% of GDP2000 to the ‘best’ case of 0% of GDP-2000) was about half as important, then this attribute should be given a weight of 50599. And another attribute considered to be only one quarter as important should be given a weight of 25, and so on. In order to allow participants to consider the practical implications of their weighting judgements, the scoring tables used in our exercise automatically represented the scores determined for the ‘best’ and the ‘worst’ scenario under each criterion. In the end, the final weighting scheme is a set of numbers

599

Notice that, even though both the ‘impacts of air pollution on health’ and ‘overall costs of the energy system’ attributes are scored in terms of money, we did not use these monetary values for the calculation of the relative weights. This was left entirely to the discretion of the participants in the exercise.


433

whose ratios reflect the relative importance of scoring differences under the various criteria. In contrast with most other multi-criteria appraisal exercises (cf. Annex 4), participants in the MCM interview sessions were left relatively free to undertake the weighting process in whatever way they felt most comfortable. Participants usually began by selecting one criterion (often the most important one) and then addressed the other criteria by comparing it with the first one. The intensity of the differences in importance was addressed by altering the ratios of weightings one by one. This continued in an iterative fashion until a final set of numbers was determined which the participant felt comfortable with. The MCM model developed for the present exercise allowed the participants to enter arbitrary weighting numbers (e.g. on a scale of 1 to 10) as the chosen weights were recalculated by the spreadsheet software in percentage terms. This procedure corresponds with an intuitive model of importance weighting in terms of 100 ‘importance points’ which had to be distributed over the selected criteria. Also, we had foreseen the possibility to work on the three different levels of generality (corresponding to the branches of the ‘combined value tree’), so that weights assigned to lower levels could be corrected (if necessary) at higher levels, to better reflect an overall value judgment. For instance, if a participant had initially chosen 5 out of 10 criteria from the subset of ‘economic welfare’ criteria, this would of course be reflected in a higher relative weight for this subset compared to ‘environmental and human health & safety’ and ‘social, political, cultural and ethical needs’. Thus, in order to be able to correct for any inadvertent biases arising from the first selection (rather than an explicit consideration of the relative importance of different ‘matters of concern’), participants were offered the possibility to correct the weights initially assigned to the selection of criteria either at the intermediate (16 categories) or highest level (4 categories). Again, results at these higher levels of generalisation were recalculated in percentage terms and ‘cascaded’ down towards the lower level. As mentioned in the previous section, all participants could freely experiment themselves with assigned weightings in order to explore sensitivities. The weighting procedure was not concluded until each individual participant expressed satisfaction that he/she had arrived at a meaningful expression of his/her position in the debate (although not all participants made use of this opportunity). 2.2.3

Analysis of results

Following the round of interview sessions, results were analysed based on quantitative (i.e. the set of weights and scores identified during each MCM exercise) and qualitative (i.e. the comments received during each MCM exercise) givens. Furthermore, a short period of time after the individual interviews, participants were again contacted (by telephone) by a researcher from the ‘Study centre for Technology, Economics and the Environment’ (STEM – University of Antwerp). In these telephone interviews, participants were asked to give their overall evaluation of the entire process, including reflections on its usefulness as a tool for governance. The telephone interviews were conducted on an independent basis by a STEM researcher, and indeed succeeded in eliciting some critical remarks. We will reflect on some of these remarks in section 4.2 below; a more thorough account (including a personal appreciation) is given in Keune et al. (2004).

434

Chapter 7

In the following section, we will mainly concentrate on the results of the MCM exercises in terms of comments received regarding the scenarios (Section 3.1); attributes (Section 3.2), scores (Section 3.3), weightings (Section 3.4) identified in the evaluation; the overall results of the exercise in terms of rankings of scenarios (Section 3.5); and sensitivity analyses performed by us on the results (Section 3.6).

3 Results The model described above has been used with representatives from a variety of organisations on an individual basis. As mentioned before, the main purpose of these individual exercises was to obtain information on the value and ease of use of the model as part of a governance process. Therefore, the following discussions will focus on criticisms received from ‘users’ of the model, as such criticisms provide insights into the users’ understanding of formal scenario modelling and the application of decision-making techniques to ‘unstructured’ problems such as sustainable energy policy.

3.1

The scenarios

The initial stage of the individual interviews was devoted to an open discussion of the scenarios developed for the purpose of the MCM exercise. It became clear that participants in the exercise generally fell into one of the following categories. From the comments made by some participants (most notably A1, C1, E1, G1) it became clear that these people already had accumulated some experience in working with mathematical energy models, and hence understood its principal advantages and limitations600. These people either accepted the scenario results as a good representation of a possible spectrum of outcomes or were able to give very pointed comments on certain aspects of the model assumptions (cf. infra). Other participants however were less familiar with the approach. Some of them also made it clear that they had been unable to devote a large amount of time to study these scenarios (with all the embedded hypotheses) in great detail. In these cases, the initial stage of the interview was generally limited to an explanation of what was meant exactly by the short labels given to each of the scenarios. However, it was of course impossible to ascertain in such a brief period of time whether these participants actually understood the formal mathematical modelling used to develop the scenarios. Although we organised a ‘scenario workshop’ (28 April 2003) with the specific purpose of making participants in the exercise more familiar with the approach, it was clear that the concept of formal mathematical energy modelling could not be communicated adequately either during the scenario workshop or at the beginning of the interview sessions. For instance, doubts were expressed on which factors were taken into account by the MARKAL model. Some

600

These codes refer to table A-1 (Annex 1).


435

participants in the MCM exercise had difficulties with treating the ‘options for action’ (the scenarios) as representing policy choices, and in deciding how to interpret them. On the first issue, one participant (A2) commented that the scenarios used for our exercise should give a clear picture of policy options as means by which supply and demand patterns might be altered, linking them to clearly defined outcomes – whereas the scenarios developed by us for this exercise were very general descriptions of outcomes (with the simple presumption that energy policy – and all other relevant ‘actants’ – would be ‘in line’ with these outcomes, without however going into the details of what this policy would look like)601. Similarly, a much heard critique was that the model cannot take into account ‘qualitative’ factors such as ethical considerations on the part of citizens/consumers or policy choices in fields related to energy policy (e.g. spatial arrangements, waste, etc.). Virtually all participants in the exercise made the comment that scenarios should also reflect energy policy choices at the European level, which was supposed to become the most relevant decision-making level after the liberalisation of the energy markets. As explained before, this ‘vagueness’ in the representation of actual policy choices in the scenarios was entirely intentional. By confronting participants in the exercise with a ‘pure industrial/market fiction’ (in the end, this is what a ‘perfect foresight’-model such as MARKAL is) we hoped to elicit from them commentaries and interpretations as to which ‘actants’ would bring these methodological fictions closer to reality or, conversely, render them entirely unfeasible and/or undesirable. For instance, if a user of the MCM model thought a certain scenario would be unfeasible he could express this by assigning a score of zero and a high weight to the criterion which he/she felt made that particular scenario deficient (this of course requires an explanation why the participant believes a particular scenario is unfeasible). Thus, the way in which users interpret these scenarios precisely provides valuable additional information on the way they perceive a sustainable energy future. However, it is clear from some of the comments received that this aspect of the exercise was not entirely understood and therefore should certainly require more attention if this work is to be continued in the future. Setting aside these interpretative problems, there was little discussion over the justification of the four basic strategies (nuclear power; carbon capture & storage; (more) import of electricity; or (more) renewables, cogeneration and energy conservation) as being of central importance to the future of the energy system. However, some participants (mostly representatives of environmental NGO’s) criticised the way certain technical options were represented in the MARKAL database, and/or added that other options were not sufficiently addressed by the scenarios at present. Participant A2 was of the opinion that the present scenarios focused too much on electricity production whereas according to him the more important action domains were transportation and (domestic and industrial)

601

For instance, we did not specify how the ‘rational’ behaviour of energy consumers in the ‘Rational perspective’ scenarios (formally modelled by using a 5% discount rate for all energy investment decisions) could come about. One can think of e.g. making energy audit fiscally attractive, greater communication efforts on the benefits of energy saving, and other DSM measures.

436

Chapter 7

heat production602. Options in these domains were described as being more promising candidates for bringing about a sustainable energy future. For instance, he referred to cogeneration combined with heat storage as a valuable option for extending the useful application of cogeneration techniques also to peak-load electricity production (whereas in the present model calculations electricity is treated as a ‘by-product’ of heat generation and therefore has to be fed into the electricity grid at the moment of heat production – i.e. the electricity has to be integrated in the base load)603. Speaking in more general terms, this participant was of the opinion that the options for electricity production should be functionally linked more clearly to specific consumption patterns – e.g. he gave the example of nuclear power which, according to him, could maybe in the future become an interesting option for hydrogen production; or electricity production based on renewable energy sources which would require electricity storage capacity for a greater penetration. Another representative of an environmental NGO (participant A1) expressed doubts regarding the costs of some of the electricity production options taken up in the MARKAL database. In his opinion, the investment cost for the EPR nuclear power plant was too low because financial costs were not included; conversely, he thought wind power was put at a disadvantage. He also commented that scenarios should also be able to withstand a ‘reality check’. In his opinion, the ‘import’ scenario (which takes into account import of up to 30% of Belgian electricity needs from foreign countries) would become a reality already on much shorter notice (2010-2015), due to a foreseeable overproduction of French nuclear electricity at marginal cost. Hence, participant A1 was of the opinion that all scenarios developed for the Belgian context should take into account a significant fraction of electricity import. Another ‘reality check’ according to this participant was the fact that possibilities to store CO2 would be limited in any case, hence rendering the scenarios which to a large extent rely on this option very unlikely. Both representatives of the environmental NGO’s were of the opinion that the technical potential for energy conservation was underestimated in the scenarios. Also, both mentioned the possibility of rising uranium resource prices in the future, whereas our model assumptions include a reasonably stable price over the coming 50 years. All things considered, these criticisms can be taken as indications for further improvement of the ‘robustness’ of the scenarios (cf. Section 4.1.2).

3.2

Selection of criteria

Selection of criteria proved to be the easiest part of the MCM exercise for most of the participants. The fact that (as a guideline) a maximum of 15 attributes could be selected for the exercise was not felt as unduly constraining, since the users of the model generally understood that the subsequent weights given to the selected attributes would anyway imply that only the most important ones would influence the final result of the exercise. 602

Consequently, this person felt that the discussion was somehow ‘forcefully’ directed towards nuclear power, whereas in his opinion other problems were more important and pressing. 603 Adding that peak-load electricity production could even become very interesting (in economic terms) for owners of cogenaration capacity in a free electricity market, since prices offered for peak loads could be significantly higher than the average prices taken into account in the MARKAL model.


437

The hierarchical structuring of the ‘combined value tree’ developed by us for this exercise seemed to be acceptable to all participants. This is not entirely surprising, since it reflects the by now relatively well-established three ‘pillars’ of sustainable development (ecological, economic and institutional factors). However, it remained the case for some of the attributes that different participants gave somewhat different meanings to the same attribute. Such different framings were mostly revealed in the act of ‘scoring’ the attributes (cf. Section 3.3). For instance, while the ‘degree of decentralisation’ was generally associated with opportunities to influence political decision making, one participant (D2) rather interpreted this attribute from a technical point of view – i.e. according to him, from a technical point of view centrally-produced electricity could be controlled much better than its decentralised counterpart, also because the knowledge base for centralised electricity production is considered to be much more extensive. For this participant, the ‘degree of decentralisation’ was actually misplaced as a sub criterion of ‘social, political, cultural and ethical needs’. Sometimes, one criterion was also seen as a replacement for a range of other criteria – e.g. participant G1 considered the ‘degree of decentralisation’ to be an adequate translation of his more general concerns on the political influence of large and powerful multinational energy firms. Table 42 (to be found on page 442) gives a summary of the selected criteria (organised at the intermediate level sub-groupings). The full list of selected attributes for each participant is given in Annex 5 (selected attributes are indicated in bold). Several observations regarding the framing of the sustainable energy question can be made. For instance, the majority of participants chose to include elements in their assessments which could not be reduced to strictly technical parameters. This was the case even in subgroupings which are usually conceived of in technical terms (environmental and economic impacts). For instance, participant G1 (a representative of an advisory body) did not select any criterion from the ‘economic welfare’ grouping, because he considered the quantitative criteria mentioned there to be an inadequate representation of his opinion – i.e. in his view, economic criteria should relate to ‘qualitative’ issues such as quality of work, fulfilling ‘real’ needs (instead of artificially-stimulated energy squandering), etc. In fact only two participants (D2 and E1) only relied on quantitatively-scored criteria for representing their opinion. In general, much attention was also given to the social and political ramifications of the adoption of energy technologies. In general, representatives of environmental NGO’s (A1 and A2) and (perhaps more surprisingly) members of advisory groups (G1 and G2) showed a great sensitivity to these issues, whereas ‘environmental and human health & safety’ were considered to be generally important from all perspectives. Also, different participants tended to emphasise different timeframes when pondering the intricacies of sustainable energy development. The perspectives taken by participant D2 (the representative of an employers’ organisation) and A1 (the representative of an environmental NGO) might serve as an example here. While participant D2 stressed the importance of cheap electricity on the basis of short-term competitive exigencies, participant A1 rather worked his way back from a future view of society (with strict energy savings) towards the actions that would lead to it. Seen from this point of view, cheap electricity would be harmful rather than beneficial.

438

Chapter 7

The selection of criteria was evidently also influenced by the professional interests and perspectives of the participants. For example, participant C1 (the representative of a labour union) was the only one who selected job opportunities amongst the most important criteria (with the suggestion that this criterion should be quantified objectively). Actually this participant chose to add another criterion to the list during the exercise, namely the (nonradiological) occupational risks in the nuclear sector, which he considered to become more relevant in a context of intensified competition. Another example is participant D1 (the representative of small and medium-sized enterprises), who chose amongst the most important criteria the ‘distribution of economic benefits/burdens’ (where he considered the ‘nuclear’ scenarios to be the most unfair in terms of a just distribution) and ‘economic risks’ (where he considered import of electricity to be the most ‘risky’ option). However, as mentioned, although participants did frame the problem by stressing issues relevant to themselves, they also acknowledged other areas (e.g. the ‘environmental and human health and safety’) which had to be addressed, thus building in a ‘higher-level’ justification in their reasoning. Another issue regarding the selection of criteria deserves some attention, namely the imposition of ‘performance thresholds’ as a prerequisite for the consideration of options. Under one perspective in the current exercise (that of participant A1), for instance, climate change and nuclear proliferation were formulated as two ‘meta-criteria’, in the sense that their importance transcended the importance of all other criteria. Whereas climate change was indeed included in the present exercise as a ‘hard’ boundary factor which had to be met by all scenarios (i.e. a 30% GHG emission reduction by 2050), participant A1 still felt unsatisfied because of the unambitious target proposed. Referring to the ‘German advisory council on global change’ (WBGU 2003), this participant estimated that an 80% reduction target (by 2050) would have been more appropriate in order to avoid dangerous levels of anthropogenically-induced climate change. With regard to the ‘non-proliferation’ question, participant A1 reasoned that any long-term continuation of the nuclear option eventually would have to make use of breeder reactors (in view of the uranium resource limitations). Hence, the emergence of a ‘plutonium economy’ with unacceptable proliferation risks on a worldwide scale. In other words, the satisfaction of these criteria was actually regarded as a precondition for the inclusion of an option (nuclear power) in the appraisal. In fact, this approach to the prioritisation of criteria is well-documented in the literature (see e.g. Hobbs and Meier 2000), where it is referred to as ‘lexicographic ordering’604. It is a very different approach to the analytical framework proposed here under which all criteria are in principle traded off against each other. However, this problem can be overcome by simply being transparent about the threshold condition imposed by the participant, establishing whether the scenarios uphold the threshold criteria, and then applying the remaining criteria in the same fashion as for the other participants. Participant A1 agreed to continue the exercise (even with the ‘nuclear’ scenarios) with the admonition however that his point of view was

604

I.e. the ordering of options as if they were words in a dictionary, with performance characteristics as letters.


439

‘more polarised’ than suggested by the final ranking obtained at the end of the MCM exercise. Similarly, although not presented as ‘hurdles’, it became clear from the responses of a number of participants that a number of criteria which could be subsumed under one heading had so much importance that for practical purposes they could be interpreted as ‘thresholds’. For instance, participant A2 explicitly mentioned that ‘equity’ in general was of central importance to him – which was reflected in his choice of sub criteria (‘leaving resources for development’, ‘geographic distribution of risks’, transgenerational ethics as reflected in his attribution of equal weights for short-term, mid-term and long-term economic costs, etc.). In another case (that of participant F3), this central criterion proved to be the ‘limited availability of space’ in Belgium, since most of the selected criteria (e.g. ‘land use’, ‘water use’, ‘impacts of solid waste’, etc.) reflected a concern for a possible future competition between different functional needs for a limited amount of space. Two general findings thus emerge from this analysis of criteria selection. The first is that most of the participants in the MCM exercise identify criteria which lie beyond the scope of the traditional and more technical forms of sustainability assessment (e.g. lifecycle analysis, calculation of external costs) or advisory reports on energy policy. Nevertheless these criteria are important in ‘framing’ the assessment, and therefore any attempt to limit this ‘framing’ (e.g. to one dominant perspective) will likely produce a distorted debate (forcing other ‘voices’ to present their arguments in the ‘language’ of the dominant perspective). This is most evident in the case of the socio-political criteria which, although they play an important role in the assessment of a number of the participants in our exercise, seem to be absent from the ‘official’ discourse and public debate on sustainable energy altogether. A second (and related) finding is that both the ‘direct’ effects (e.g. health impacts) as well as the ‘complex’ indirect consequences of development in the energy field (e.g. influences on political decision making) are considered to be relevant for the purposes of sustainability appraisal.

3.3

Scoring

In general, the scoring of scenarios under the various criteria represented the most difficult part of the MCM exercise for most participants. Often, comments and/or suggestions for improvement were formulated pertaining to either the attribute scores selected by us or the difficulty of arriving at qualitative scores in the absence of more ‘solid’ data. Some participants pointed out that the type of attribute measurement used in the model was inappropriate to their assessment of that particular attribute. For example, participant C1 (a representative of a labour union) found that for the issue of ‘catastrophic (nuclear) risks’ a more suitable indicator would have been the number of incidents in nuclear power plants as

440

Chapter 7

measured on the ‘International Nuclear Events Scale’ (INES)605. Concerning the second issue (the need for more ‘solid’ quantitative data), comments received during the MCM exercise highlighted areas where more technical information was felt to be relevant to the debate. For instance, participant A1 pointed out that the criterion ‘economic risks’ – defined as a qualitative criterion for our present purposes – could be estimated much more objectively. This participant principally felt that these economic risks were chiefly present for nuclear power, as he mentioned the examples of possible soaring dismantling costs (according to participant A1, these could amount to 15-50% of original investment costs), liability for nuclear accidents (not completely included in the nuclear kWh price – cf. Chapter 3), and the historical experience that nuclear power plants have shown a ‘negative learning curve’ (i.e. investment costs for the present generation of power plants have increased (due to ever increasing safety demands) rather than decreased – as can be expected from a ‘normal’ learning curve). Similarly, participant C1 felt that the ‘distribution of economic benefits/burdens’ could be quantified to a certain extent: in this case, he was referring to the issue of stranded costs and benefits of nuclear power generation in Belgium (cf. Chapter 4). As a consequence of this particular framing, all ‘market’ scenarios were given a score of zero on this particular attribute, since this participant felt that the issue of stranded benefits would not be resolved to his satisfaction in the highly competitive atmosphere suggested by these scenarios. On the issue of ‘direct or indirect subsidies needed’, participant C1 felt that an investigation of the subsidies given to the different electricity production options in the past would be relevant information to discuss future developments. Other environmental criteria such as ‘land use’ and ‘water use’, or social criteria such as ‘job opportunities’, were generally considered to be in need of further quantification. Of course, such criticisms do not fundamentally undermine the MCM method as such: since indicators can be changed easily (or given a more quantitative ‘backing’), the ‘defaults’ used by us for the present exercise should be seen rather as suggestions than as ideals. As such, they are an important result of the exercise, since the difficulties that some participants experience with particular indicators reflects their interpretation of a particular criterion and therefore do not constitute a fundamental error in the method itself. The scoring exercise also revealed the crucial importance of contextual considerations. As participants justified their scoring during the interview sessions, they frequently asked either for more information regarding the scenarios or spelled out assumptions. These assumptions often related to how well the participants thought energy options would work in practice. We can illustrate this with some examples. Perhaps the most striking example of the importance of contextual considerations is provided by a comparison between the

605

The INES scale is a means for promptly communicating to the public in consistent terms the safety significance of events reported at nuclear installations. It was designed by an international group of experts convened jointly in 1989 by the IAEA and the ‘Nuclear Energy Agency’ (NEA) of the OECD with the intention of creating a common understanding of nuclear incidents between nuclear operators, decision makers and the media. Roughly speaking, it could be compared to the (non-technical) Richter scale for communicating the strength of an earthquake.


441

perspectives of participant A1 and A2 (both representatives of environmental NGO’s). Whereas participant A1 was of the opinion that scenarios including the nuclear option would automatically lead to a higher ‘concentration of power’ in oligopolistic energy multinationals (hence, ‘nuclear’ scenarios score low on this attribute)606, participant A2, although sharing this concern in the case of the nuclear option, expressed his opinion with much more uncertainty. He reflected that the historical ‘lack of public participation’ in the decision to build nuclear power plants might very well be repeated in the case of the construction of large (offshore) wind farms, as in participant A2’s opinion, here again decisions were taken by a small circle of influential people leaving the public with only the possibility to say ‘yes or no’ to pre-defined plans. Hence, participant A2 chose a relatively large uncertainty interval for the scores on ‘control and concentration of power’ (although the ‘nuclear’ scenarios still scored lower on average). The importance of context could also be seen in other ‘technical’ criteria, for instance when scoring the ‘occupational risks’ of the nuclear (this criterion was added by participant C1) and other options (gas and coal). For participant C1, the effectiveness of regulations in a context of increasing competitive pressures proved to be the deciding factor in scoring these attributes. Hence, indicators on the effectiveness of regulatory interventions in these fields were considered to be of great relevance to the sustainability debate (rather than the historical data on the occupational risks of different electricity production technologies, taken from the ExternE studies and used by us for the present purposes). Similarly, participant A1 stressed the importance of safety culture (i.e. the organisational aspects of safety) in scoring the ‘radiological health impacts’ on workers in the nuclear sector. Thus, comments received during the scoring exercise showed many possibilities for improvement. As a result, the actual scores that were attributed in this exercise (cf. Annex 5) should not be considered definitive. Many of the criteria would either require further disaggregation and (possibly) quantification. Nevertheless, since the people involved in the exercise generally did have expertise in many different areas of energy policy and collectively represented a wide range of pertinent (sometimes technical) perspectives, the general pattern in their scoring should provide at least a pointer to the character of the technical issues at stake.

3.4

Weightings

The assigning of numerical weights to reflect the relative importance of different appraisal criteria is perhaps potentially the most problematic aspect of any multi-criteria analysis (cf. Section 3.2 and Annex 4). The overall results of the weighting procedure are displayed in Table 42, and were confirmed by participants at the end of the MCM interviews.

606

In the post-MCM telephone interview, this participant also stressed that he approached energy policy issues from the angle of political power, and he conceived the role of his organisation as one of countering the prevailing power structures.

442

I.

Chapter 7

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

Tot.

Air pollution

11

0

12

19

12

19

11

9

0

0

7/10

Occupational health

0

0

5

0

0

9

0

7

0

0

3/10

Radiological impact

20

0

12

15

12

0

11

0

19

10

7/10

Aesthetic

0

0

0

0

7

0

0

0

0

7

2/10

Other environmental

0

0

0

0

12

17

11

2

0

22

5/10

Resource use

0

0

5

0

0

0

11

16

0

0

3/10

Other energy related

11

6

0

0

0

0

11

9

19

0

5/10

% of total weight

42%

6%

35%

34%

43%

45%

55%

43%

38%

39%

Overall economic

16

34

28

33

19

20

11

29

0

19

9/10

Producer benefit

0

0

5

0

0

0

0

0

0

0

1/10

Consumer benefit

0

0

0

0

12

15

0

15

0

0

3/10

International

0

0

4

0

0

0

0

0

0

0

1/10

0

0

0

0

0

0

0

0

13

0

1/10

16%

34%

37%

33%

31%

35%

11%

44%

13%

19%

11

19

5

33

0

0

11

0

19

12

6/10

23

18

13

0

15

10

11

0

4

19

7/10

8

14

6

0

11

0

12

13

17

11

7/10

0

0

5

0

0

0

0

0

0

0

1/10

42%

51%

28%

33%

26%

10%

34%

13%

40%

42%

impacts

pressures

II.

benefit

cooperation Need for government support % of total weight III.

Individual choice/participation Institutional needs/benefits Development opportunities Jobs % of total weight

IV.

Diversification % of total weight

0

9

0

0

0

10

0

0

9

0

0%

9%

0%

0%

0%

10%

0%

0%

9%

0%

3/10

Table 42. Weightings used by all participants (intermediate level) (Note: last column indicates the number of participants choosing a criterion from that particular subgrouping)607

607 Each column represents the way a particular participant has divided his100 ‘importance points’ or weights over the intermediate-level criteria.


443

Of the ten participants, all identified at least one criterion in each of the three broad toplevel groupings (environment, economy and social/political/cultural/ethical criteria). Diversification was not considered to be a priority for six interviewees. From the average weightings assigned to the top-level groupings, it is clear that the ‘environmental & health’ concerns carried a relatively equal weight for all participants (with the exception of participant A2), while opinions on the relative importance of ‘economic’ and ‘social/political/cultural/ethical’ factors tended to vary considerably608. Furthermore, each of the three top-level groupings was dominant under at least one perspective or another609: • ‘Environmental and human health & safety’ criteria were assigned the highest priority by participants A1, D1, D2, E1, F2 and F3; • ‘Economic welfare’ criteria were assigned the highest priority by participants C1 and F3; • ‘Social, political, cultural and ethical needs’ criteria were assigned the highest priority by participants A1, A2, G1 and G2. The fact that all top-level groupings of criteria (with the exception of ‘diversification’) were dominant under at least one perspective or another reveals the magnitude of the differences in the perspectives taken by different participants. This is particularly relevant for the political/social/cultural/ethical criteria (apparently of high importance to the representatives of environmental NGO’s and advisory organisms), since these tend to be absent in the more ‘technical’ forms of sustainability assessment favoured also by some important actors in the debate. For instance, participants D2 and E1 (representative of an industrial federation and a large electricity producer respectively) only chose one criterion from the ‘social/political/cultural/ethical’ factors (since ‘decentralisation’ rather represented a technical matter of concern for participant D2). From Table 42 it also becomes clear that within these broad groupings some criteria are referred to more than others. For instance, in the ‘environmental and human health & safety’ group, ‘air pollution’ and ‘radiological impacts’ (including the long-term impacts of high-level waste) are generally considered to be very important. And within the ‘economic welfare’ grouping, the overall economic benefits of the different scenarios clearly formed an overriding priority for almost all of the participants (with the exception of participant G1). Again (as was the case with scoring), one must be careful when interpreting these weightings. For instance, the weights attributed to ‘radiological impacts’ (and in particular the long-term impacts of high-level waste) can be considered from different points of view. For some participants (e.g. participant A1 and G1) the high weight attributed to this

608

This is perhaps a bit surprising since participant A2 is a representative of an environmental NGO. However, this person felt that political factors (such as a concentration of power) carried greater priority now, also in view of reaching the most ‘difficult’ environmental targets (e.g. limiting greenhouse gas emissions by at least 50%). 609 This provides some further confirmation (beyond the approval of participants) that grouping criteria in this way provides a relatively robust structure for considering different matters of concern.

444

Chapter 7

criterion reflected a concern for transgenerational ethics, while for others (e.g. participant C1, D1 and D2) this reflects a concern that without an operational solution for the management of high-level waste it would be harder (if not impossible) to justify the construction of additional nuclear power plants610. And, as mentioned in section 3.2, there were also cases (e.g. participant A1) where the weighting procedure was considered to be somewhat irrelevant, in the sense that some criteria should be considered as ‘hurdles’ (i.e. these criteria could not be the subject of trade-offs).

3.5

Rankings

Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

P.LCS.I

LCS

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

LCS

3.

P.CS

P.LCS

P.LCS.I

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

P.LCS.I

4.

LCS

P.CS

P.CS

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

LCS

P.LCS.I

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

LCS

3.

P.CS

P.LCS.I

P.CS

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

P.LCS.I

4.

LCS

P.CS

LCS

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

Market

Rational

Table 43. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on average scores) (Note: grey areas indicate change of ranking when considering a ‚Market’ and ‚Rational’ perspective)

Figure 19 and Table 43 display the overall rankings for each of the strategies under the perspective of each of the ten participants611. Results for both the ‘Market’ and ‘Rational Perspective’ scenarios are given, with a graphic representation of both ‘optimistic’ and ‘pessimistic’ scores. The axes in this figure are scaled in order to clarify the differences in ranking orders (rather than the absolute values taken by the ranks) under pessimistic and optimistic assumptions.

610

From this perspective the absolute amount of high-level waste generated in the different scenarios is not so much of importance, but rather the amount of high-level waste for which no definite solution is at hand. 611 After we had conducted the MCM interviews, we became aware that we had misrepresented the ‘overall cost energy system’ (short-, mid- and long-term) parameter: the correct values proved to be higher than the ones we had used in the MCM exercise. However, in relative terms the new parameter value only changed the relative positions of the M.K.P.LCS and M.K.P.LCS.I scenarios. Only participant A2 and C1’s final rankings were affected by the change in parameter value.


A1 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

A1 – Ranking Rational Scenarios

MKPLCSI

A2 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

MKPCS

MKPLCS

MKPLCSI

MKPCS

MKPLCS

RKPCS

RKPLCS

RKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI

C1 – Ranking Rational Scenarios

MKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI

D1 – Ranking Rational Scenarios

D1 – Ranking Market Scenarios

MKLCS

RKLCS

A2 – Ranking Rational Scenarios

C1 – Ranking Market Scenarios

MKLCS

445

MKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI

446

Chapter 7

D2 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

D2 – Ranking Rational Scenarios

MKPLCSI

E1 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

MKPCS

MKPLCS

MKPLCSI

MKPCS

MKPLCS

RKPLCS

RKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI

F2 – Ranking Rational Scenarios

MKPLCSI

F3 – Ranking Market Scenarios

MKLCS

RKPCS

E1 – Ranking Rational Scenarios

F2 – Ranking Market Scenarios

MKLCS

RKLCS

RKLCS

RKPCS

RKPLCS

RKPLCSI

F3 – Ranking Rational Scenarios

MKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI


G1 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

G1 – Ranking Rational Scenarios

MKPLCSI

G2 – Ranking Market Scenarios

MKLCS

MKPCS

MKPLCS

447

RKLCS

RKPCS

RKPLCS

RKPLCSI

G2 – Ranking Rational Scenarios

MKPLCSI

RKLCS

RKPCS

RKPLCS

RKPLCSI

Figure 19. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (from left to right: LCS, P.CS, P.LCS, P.LCS.I; error bars indicate uncertainty)612

Several features emerge: • The perspectives taken by the different participants result in very different ranking orders across the scenarios; • For some participants (e.g. A1, F2 and G1) a clear distinction can be drawn between the pattern displayed by the three ‘non-nuclear’ strategies and the ‘nuclear’ strategy. However, for the other six distinctions between the ‘nuclear’ and ‘non-nuclear’ scenarios are less clear; • Although there are cases where the differences between scenarios are very pronounced, for most participants choosing between two or more particular scenarios turns out to be a rather difficult endeavour. From Table 43 it becomes clear that each of the individual scenarios (with the exception of the P.CS-scenarios, i.e. scenarios which include a high potential for carbon capture and

612

LCS = low carbon storage (+no nuclear phase out); P.CS = nuclear phase out (+ carbon storage allowed); P.LCS = nuclear phase out and low carbon storage; P.LCS.I = nuclear phase out, low carbon storage and import of electricity.

448

Chapter 7

storage technologies) on average performs both best and relatively bad or worst under the perspective of one or another participant: • The ‘nuclear’ strategy (LCS) performs best according to participants A2, D2 and F3, and worst according to A1, D1, F2 and G1; • The ‘renewables and cogeneration’ strategy (P.LCS) performs best according to participants A1, C1, F2, G1 and G2, while participant E1 and F3 considered this to be the second-worst solution; • The ‘import’ strategy (P.LCS.I) performs best according to participant D1 and E1, and worst according to participant D2. Table 43 also clearly indicates that overall rankings prove to be relatively stable with regard to the scenario perspective (‘Market’ or ‘Rational’) taken. Only in two cases (A2 and C1) the overall ranking was changed as a result of a different scenario perspective. However, from Figure 19 it is also clear that the perceived need for investment in nuclear power (under the perspective taken by participants A2, D2 and F3) becomes less pronounced in the ‘Rational Perspective’ scenarios than in the ‘Market’ scenarios. The final rankings (as represented here) were arrived at after a process of often intense deliberation. Participants often changed some evaluation parameters in view of their overall appreciation of certain options. These ‘overall’ appreciations sometimes proved to be very informative in their own right, in addition to the rankings derived by the ‘formal’ mathematical procedure. For instance, participant C1, when pondering the results of his MCM interview, commented that she was not against the nuclear option, provided that effective regulations were in place for guaranteeing its safety (in terms of occupational risks, nuclear incidents and finding a solution to the problem of high-level wastes). She also underscored her dislike for the ‘import’ option based on a combined argument of ensuring job opportunities for the Belgian electricity sector and diminishing the risks of economic dependence on foreign countries. On the other hand, participant D1 generally preferred the ‘import’ option, with the argument that Belgium is already almost fully dependent on foreign countries for fulfilling its primary energy needs; therefore, if ‘sustainable’ electricity could be produced abroad under more favourable conditions, he saw no need for Belgian self-sufficiency in terms of electricity production. Similarly, participant A1 commented that although he preferred the option of phasing out nuclear power and relying on a combination of energy conservation, renewables and cogeneration without increased import of electricity, he also had some doubts on the feasibility of this scenario in the Belgian context (he referred to the presence of energy-intensive industrial activities); therefore, some import of sustainable electricity might prove to be a necessity. And participants E1 and F3 suggested that actually a combination of their first and second preferred option (i.e. a combination of increasing electricity import and maintaining nuclear power production in Belgium) would be optimal from their point of view (albeit for different reasons).


449

Thus, the final rankings obtained through the MCM exercise clearly yield a multitude of divergent signals. Almost each of the different strategies are, under one perspective or another, assigned the status of the ‘optimal’ solution. Now, one might reasonably ask what could possibly be learnt from this result. This issue will be discussed further in 4.1.3. For the moment, let us simply observe that, rather than being a perverse feature of the analysis, these ‘divergent signals’ can actually be seen as an indication of the fidelity with which the MCM technique reproduces the degree of discord and expressions of uncertainty by individual participants. Of course, there is always the possibility of strategic behaviour from the part of some of the participants (i.e. the desire to deliberately influence the result of the MCM exercise in one way or another by selecting only those criteria which can reasonably be expected to support a pre-existing preference structure). The fact that participants often changed some evaluation parameters in view of their overall appreciation of certain options can be seen as an indication of such behaviour. And although strategic behaviour can be a feature in almost any form of deliberation or analysis, multi-criteria approaches are generally more vulnerable to such behaviour because the separate articulation of criteria and weightings are justified according to ‘subjective’ value perspectives. The possibility of strategic behaviour is even more pertinent for the present application of MCM, since participants were asked not only to provide weightings, but also in some cases to score scenarios themselves. For instance, participant A1’s view on nonproliferation as an absolute requirement for any future scenario of course excludes consideration of the nuclear option from the outset of the exercise. But nevertheless there are some indications of the willingness shown by participants (at least in this ‘protected’ setting – cf. Section 4.2.1) to go through the exercise in a ‘fair’ way. Participant A1, despite his objections, still proceeded with the MCM exercise. Similarly, even in the cases where there was a pronounced ‘loser’ scenario (e.g. in the rankings of participants A1, F2 and G1), their selection of criteria included some under which the disfavoured scenario still scored relatively highly. But fundamentally the point remains of course, as Stirling and Mayer (1999) remind us, that what may appear as expedient strategic assumptions under one perspective may be viewed as ‘dispassionate framing’ of the issue under another. Furthermore, this is as true of the positions taken by the participants in the exercise as it is of the position taken by us as analysts (cf. Section 4.2.2). An important advantage offered by MCM in this regard is that strategic behaviour can be identified more easily in view of the transparency and verifiability of the entire process.

3.6

Sensitivity analysis

As mentioned in section 2.2, participants had already the opportunity during the individual MCM interview sessions to experiment with changes to the weighting values which they had assigned to their chosen criteria. However, as Stirling and Mayer (1999, 2000) observe (and as confirmed by our exercise), the expression of the relative importance of different ‘matters of concern’ in simple numerical terms remains a rather unfamiliar or even counterintuitive process for some participants. To this we might add that weighting schemes were elicited after a long and sometimes laborious interview process. Thus, although all participants expressed satisfaction with their overall final weighting scheme at

450

Chapter 7

the end of the individual exercises, the possibility could not be discounted that the assignment of weightings may in some cases have been unduly truncated by fatigue or pressures of time. For these reasons, the final analysis of results should also include an exploration of ‘sensitivities’. We chose to include two different types of sensitivity analysis. The first one involved an exploration of the ‘sensitivity’ of final rankings to the uncertainties as expressed in ‘optimistic’ and ‘pessimistic’ scores for different attributes (Section 3.6.1). The second type of sensitivity analysis involved an exploration of what the final rankings would have looked like for each participant if their weightings on each of the three top-level groupings of criteria (we did not include ‘diversification’ because this criterion generally carried less weight) had been different (Section 3.6.2). 3.6.1

Sensitivity analysis: uncertainties in scoring

Table 44 and Table 45 display the overall rankings obtained for all the participants for the entire range of scenarios under both ‘optimistic’ and ‘pessimistic’ scoring assumptions. This type of sensitivity analysis reflects the impact on ranking orderings of using either ‘optimistic’ or ‘pessimistic’ scores (compared to the ranking obtained by multiplying the ‘normal’ weights with average scores). A closer investigation of these tables reveals that it is only relatively rarely that the overall uncertainties have any significant effect on the final ranking orders. According to participant A2 for instance, the ‘Market’ scenario including the nuclear option moves from being the best-performing option under ‘pessimistic’ assumptions to being second in rank under ‘optimistic’ assumptions. Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

P.LCS.I

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS.I

2.

P.LCS.I

LCS

LCS

P.LCS

P.LCS.I

LCS

P.CS

P.LCS.I

P.LCS.I

P.LCS

3.

P.CS

P.LCS

P.CS

P.CS

P.CS

P.LCS

P.LCS.I

P.CS

P.CS

LCS

4.

LCS

P.CS

P.LCS.I

LCS

P.LCS

P.CS

LCS

P.LCS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

P.LCS.I

P.LCS.I

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

P.CS

3.

P.CS

LCS

P.CS

P.CS

P.CS

P.LCS

P.LCS.I

P.CS

P.CS

P.LCS.I

4.

LCS

P.CS

LCS

LCS

P.LCS.I

P.CS

LCS

P.LCS

LCS

LCS

Market

Rational

Table 44. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on ‘optimistic’ scores) (Note: grey areas indicate change of ranking compared to ranking based on average scores)


Ranking

451

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

LCS

2.

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

P.LCS

3.

P.CS

P.LCS.I

P.LCS.I

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

P.LCS.I

4.

LCS

P.CS

P.CS

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.CS

LCS

P.LCS

LCS

2.

P.LCS.I

LCS

P.LCS.I

P.LCS

P.LCS

LCS

P.LCS

P.LCS.I

P.LCS.I

P.LCS

3.

P.CS

P.LCS.I

LCS

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

P.LCS.I

4.

LCS

P.CS

P.CS

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

Market

Rational

Table 45. Ranking of ‘Market’ and ‘Rational’ scenarios for all participants (based on ‘pessimistic’ scores) (Note: grey areas indicate change of ranking compared to ranking based on average scores)

Generally, the rankings obtained by participants A2, C1 and G2 seem to be most sensitive to uncertainties in scoring. As becomes evident from Table 46, there was a significant difference in the degree to which uncertainties were expressed by different participants. These differences do not seem to be related to the different affiliations however. Perhaps not surprisingly (in view of its novelty), significantly greater uncertainties are generally associated with the ‘carbon capture and storage’ option (P.CS), whilst the ‘import’ option (P.LCS.I) is also sometimes associated with high levels of uncertainty (especially in the ‘Market’ perspective). However, the appraisals of the individual participants seem to be the most influential factor, with nearly all options subject to the greatest uncertainties under certain perspectives. Uncertainty

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

LCS

0.18

0.20

0.22

0.01

0.35

0.19

0.46

0.39

0.25

0.23

P

0.15

0.48

0.36

0.13

0.38

0.60

0.09

0.55

0.23

0.50

P.LCS

0.06

0.25

0.18

0.05

0.27

0.18

0.06

0.32

0.19

0.29

P.LCS.I

0.11

0.44

0.23

0.05

0.38

0.19

0.16

0.49

0.17

0.37

LCS

0.13

0.16

0.17

0.03

0.25

0.18

0.20

0.26

0.24

0.16

P

0.14

0.28

0.22

0.14

0.28

0.46

0.09

0.36

0.18

0.42

P.LCS

0.08

0.19

0.15

0.04

0.28

0.16

0.10

0.34

0.17

0.31

P.LCS.I

0.08

0.29

0.16

0.06

0.27

0.14

0.10

0.30

0.14

0.28

Market

Rational

Table 46. Uncertainties expressed for each scenario according to each participant (Note: uncertainty is expressed as a ratio to the average value of ‚optimistic’ and ‚pessimistic’ ranks)

452

Chapter 7

These findings seem to suggest that it is generally not the ‘technical’ dimensions of uncertainty which play a crucial role in ranking orderings, but rather the more ‘intangible’ qualitative aspects of framing assumptions adopted by different participants. Of course, the uncertainty expressed by participants could change in view of new information obtained on the performance of different options. This issue can of course only be explored further based on a deeper examination of the scoring assumptions. 3.6.2

Sensitivity analysis: importance weightings

For this second type of sensitivity analysis, we systematically varied the weightings for each participant for each of the top-level groupings with a factor of three either up or down. The results of this sensitivity analysis were represented for each grouping of criteria as two tables exploring the effects on the overall rankings for each participant as compared to the rankings based on the original weightings and average scores (cf. Table 43). These tables are included in a separate annex to this dissertation (Annex 6). Hence, the overall difference between the lowest and the highest weighting for each top-level criterion grouping explored for each participant was therefore a factor of nine, representing a considerable difference of possible perspectives concerning the relative importance of these criteria groupings. The weighting sensitivities were examined for the top-level groupings rather than for the individual criteria themselves, because the latter exercise would have been prohibitively complex both to perform and to interpret. A closer investigation of the tables included in Annex 6 reveals that in most cases the impact of varying criteria weightings is significant. This effect is of course most outspoken in cases where the final rankings obtained in the MCM exercise revealed rather close proximities in the rankings of scenarios (e.g. participant A2, C1, and G2). However, in other cases some strong regularities can be observed (despite the significant degree of variation introduced). Sensitivity analysis on the weightings used by participant D1 for instance reveals a strong preference for the scenarios without nuclear power and low potential for ‘carbon capture and storage’ (P.LCS or P.LCS.I), since both these options alternate between the first and second spot in the rankings under different weighting assumptions. Furthermore, there are also a few examples were even these significant increases or decreases in weighting values yield relatively insignificant impacts on the final ranking. Participant A1’s final ranking for instance is only changed in case ‘economic welfare’ criteria are changed; and even then, changes are not fundamental in the sense that the nuclear option (LCS scenarios) still ranks very lowly. Other ‘sceptical’ perspectives towards nuclear power (most notably F2 and G1) are also remarkably robust with regard to final rankings (with participant F2’s preference oscillating between scenarios with or without the ‘carbon capture and storage’ option – i.e. P.CS or P.LCS). On the other hand, participant D2 and F3’s preference for a ‘nuclear’ scenario (LCS) seems to be a robust feature of the general ranking pattern.


453

4 Discussion Given these results of the individual MCM exercises, the question is of course what can be learnt from them. From this learning perspective, not only the numerical results (in terms of the final rankings obtained) are interesting. In particular, it will be clear that we do not recommend the use of multi-criteria techniques for decision-making purposes, since such use makes these techniques particularly vulnerable to strategic behaviour (cf. Section 3.5). Nevertheless, criticisms from users of the model provide insights into users’ perceptions of models in general, and into problems inherent in the application of decision-analysis techniques to ‘unstructured’ policy problems such as the demands for a more sustainable energy system. In the following sections, we will focus on both the ‘products’ (Section 4.1) used in and produced by the MCM exercise and the ‘process’ (Section 4.2) through which they were derived.

4.1

4.1.1

The ‘product’


Generally speaking, participants in the MCM exercise were satisfied with our structuring of the sustainable energy policy problem into the hierarchical structure displayed by the ‘combined value tree’. Only one participant (C1) chose to add a new criterion to the list, which reflected the specific mandate of this representative’s organisation (a labour union). The ‘combined value tree’ thus allowed representatives from very different organisations to map out the boundaries of their respective viewpoints within the same analytical framework. As mentioned, one participant (A1) mentioned the need for a more ‘objective’ hierarchical structure – i.e. some criteria obviously carried much greater weight than others in his opinion. However, it is of course difficult to propose such more ‘objective’ hierarchisation as a researcher, since this would obviously be contested from other perspectives. Another aspect of the ‘combined value tree’ which received some criticism was the degree of overlap displayed by some criteria, particularly those in the ‘social/political/cultural/ethical’ grouping. From the comments elicited during the individual MCM interviews, it is clear that participants often gave broadly similar meanings (reflecting a broad concern for more ‘democracy’ or ‘participation’ in energy policy) to such attributes as ‘degree of decentralisation’, ‘control and concentration of power’ or ‘influence on political decision making’, which did not necessarily coincide with the descriptions given by us in chapter 6. Such criticisms, rather than pointing out fundamental flaws, pave the way for further conceptual clarification (cf. Section 4.1.3) and targeted communication. In essence, what is needed would be a clarification of the effects of ‘decentralisation’ in a purely technical sense (i.e. small distributed electricity production units linked to low or medium voltage networks or operating in a isolated manner)613 on other technical issues (e.g. network stability, quality of power, system control) and non-

613

This is the way ‘decentralisation’ is conceptualised in our present scenario exercise.

454

Chapter 7

technical issues (e.g. ‘control and concentration of power’ in energy markets, ‘need for direct political intervention’ in designing a new regulatory framework, etc.)614. 4.1.2

A new direction for scenario development

In contrast to the ‘combined value tree’, it is clear from comments received during the MCM exercise that the scenarios developed as a support for this exercise were met with more criticism. Some participants felt unsure about the correct interpretation of the results. Furthermore, some participants felt that the scope of these scenarios (e.g. in terms of (technical) options included, environmental targets addressed, etc.) did not adequately represent there concerns. And our focus on a discussion of different options for electricity production was sometimes seen as a further indication of such a limited scope. As mentioned before, such criticisms were to some extent the result of an intentional move on our part. By focusing on the validity of the scenarios (in terms of a focus on the results of formal mathematical modelling rather than on elaborating qualitative scenario developments), we left open the possibility of different interpretations or ‘narratives’ on how the evolution of the energy system (as calculated by the MARKAL model) might actually take place (or conversely, run into insurmountable difficulties) in the ‘real’ world615. Thus, while certainly admitting that the present scenarios were in some ways considered to be unsatisfactory for supporting a wider deliberation on future strategic options for Belgian energy policy, this approach nevertheless enabled us to retrieve a lot of valuable information on the possible future direction for scenario development. For instance, most participants stressed the need for integrating strategic orientations for Belgian energy policy into the wider context of expected developments in the European economy and policy. This comment can be taken as a cue for enhancing the contrast in the different scenario narratives616. For instance, different ‘pictures’ for the future orientation of the Belgian economy in the European context could serve as the basis for a deliberation on future energy options. These ‘pictures’ could for instance range from a future Belgian economy which realises its growth mainly in energy-intensive industrial production to an economic model oriented more towards growth in service sector activities. Such narratives will then have to be translated into the corresponding demand levels for energy services (e.g. high-temperature heat for industry, fuel needed for transport, etc.). These energy service demand levels can consequently serve as an input for the MARKAL model, while a simple calculation using present ‘carbon intensities’ for the different energy service demands would reveal the overall effort needed in order to reach a given GHG reduction target617. 614

Such analysis can build on the groundbreaking work done by Winner (1980b) in a chapter entitled “Decentralisation clarified”. 615 These were only summarily addressed through the ‘Market’ and ‘Rational Perspective’ worldviews. 616 In agreement with Chantal Mouffe’s constant admonition (e.g. in the collection of essays taken up in Mouffe 1999) that the category of ‘the political’ (in contrast with the moral or the economic sphere) should precisely be characterised by an interplay of very diverse metaphorical accounts of society’s ordering. 617 For an approach along these lines, see e.g. the U.K. White Paper on creating a low-carbon economy (DTI and DEFRA 2003) and the recent long-term energy scenarios created for the Belgian context by Econotec and VITO (2005).


455

However, in view of the comments received, we propose a somewhat different approach than used in the present scenarios. Instead of using the MARKAL model calculations for determining the future energy system’s configuration as we have done in the present exercise (using summarily described ‘worlds’), we believe it would add to the overall transparency of the scenarios if these possible future configurations were also imposed as exogenous variables. Such configurations for the future energy system could for instance include a greater reliance on electricity as an energy vector (as is the case now in the ‘Market’ scenarios); an energy system based on the supposition that the present infrastructure (e.g. electricity transmission and distribution grids; gas terminals and pipelines; transport infrastructure) cannot be expanded (based on the supposition of intransigent spatial planning requirements and/or a lack of popular support for large-scale construction); and a future energy system mostly based on hydrogen as an energy carrier. Each possible future configuration should also include a centralised and a decentralised variant. Lastly, the MARKAL model can be used to determine the optimal use of both supply- and demand-side options for meeting the required energy demands, using the same four broad technical options we investigated in the present exercise (including sensitivity analyses)618. Thus developed, scenarios could be used as a background for discussing which policy measures would be needed in order to bring about the future as depicted in that particular scenario storyline. And of course, our ‘highly stylised’ scenarios should be completed by more in-depth modelling exercise, e.g. concerning system integration of intermittent electricity generation, security of supply, the potential for offshore wind energy, etc. (to name but a few). This way, both qualitative (narrative) and formal mathematical components of scenario building would be addressed. The latter should not be forgotten, in view of the validity of the entire scenario approach. Indeed, the majority of participants in the MCM exercise attributed a lot of weight to the overall economic cost of the different strategies. This is evident from the perspective of participants representing large industrial energy consumers (D2) or the former dominant electricity producer (E1), but also the representatives of the environmental NGO’s (A1 and A2) stress this attribute, albeit that they tend to have more attention for the long-term economic costs. Of course, cost estimates become more uncertain the farther one looks into the future. To avoid the perhaps misleading precision suggested by our present use of the existing MARKAL database, the performance figures (economic data, maximum installed capacity, technical potential for energy savings, etc.) given in this database could perhaps be based on a wider range of estimates found in the litterature. Or, as an alternative, such data could be obtained by questioning a representative sample of energy experts (e.g. using the Delphi-technique). However, although this question certainly merits further attention, empirical variability in the data set is in practice likely to be overshadowed by the even more pronounced variability in criteria weightings (as demonstrated in our pilot exercise).

618

Or ideally, different energy-economic models based on different modelling logics (e.g. equilibrium modelling vs. simulation models) could be used.

456

4.1.3

Chapter 7

Eliciting new and inclusive knowledge perspectives

If participants in the MCM exercise are free to define frames of reference other than the dominant ‘technical’ ones it logically follows that there will have to be plural domains of valid knowledge. Therefore, if we accept the ideal that the ‘best possible’ knowledge should be elicited, this implies that all participants should somehow be empowered to draw upon (expert) resources for addressing the questions they deem to be relevant to the policy question at stake. In this regard, the individual MCM exercises have clearly demonstrated that individual participants bring to bear very diverse knowledge and experiences (including ethical as well as cognitive aspects) on the issue of a sustainable energy future. The general feeling of unease experienced by a lot of participants in the MCM exercise of having to score particular attributes on the basis of an ‘intuitive’ understanding of certain matters of concern also serves as an indication of the general need for the ‘best’ kind of knowledge of each kind to be developed in deliberations. Indeed, the uncertainties expressed during the ‘scoring’ exercise can be reframed as questions requiring further deliberation. If we limit ourselves to uncertainties regarding the possible future role of nuclear power in a sustainable energy scenario, the following questions seem to be most relevant619: • Concerning external costs: What have been the direct (e.g. R&D funding, ‘soft’ investment loans, etc.) and/or indirect (e.g. limited liability for nuclear accidents) subsidies for the nuclear sector in a historic perspective, compared to other electricity production techniques (coal, oil, gas and renewables)? Is it conceivable to better internalise these externalities in the future? • Concerning nuclear safety: Is it acceptable to build nuclear power plants (with a small chance of accidents of catastrophic proportions) in densely populated and industrialised regions such as Belgium? How can we ensure that nuclear safety is regulated effectively? Furthermore, some perspectives stress the need for addressing the nuclear safety question in a global perspective – i.e. what would the consequences of an expansion of civil nuclear power worldwide be in terms of the probability of a serious nuclear accident occurring somewhere in the world (based on the results of a probabilistic safety analysis for the whole of the reactor park)? • Concerning radioactive waste management: Which solutions (incorporating a transgenerational ethic) can be envisaged for the safe disposal of high-level waste, taking into account the fact that for some waste (i.e. the bitumised medium-level waste in particular) the adequate processing technology in view of ultimate disposal is still not technically resolved? Should some measure of retrievability be foreseen? How can we guarantee that adequate funding will be available for whatever management scheme is chosen (in other words, how can the creation of a new ‘nuclear passive’ be avoided)?

619

Of course, some participants in the exercise claimed they already had the definitive answers (either affirmative or negative) to these questions. As becomes clear from Table 46, personal attitudes towards the expression of uncertainty varied considerably.


457

• Concerning non-proliferation issues: Which scenarios can be foreseen for the future non-proliferation regime faced with a worldwide expansion of civil nuclear power (based on the present once-through fuel cycle, reprocessing of spent nuclear fuel, or based on breeder reactors or other advanced fuel-cycle options)620? Is it ethically responsible to continue research into innovative nuclear reactor concepts while there is neither a guarantee yet that nuclear weapons’ stocks will be reduced, nor a worldwide prohibition on further research on military application of nuclear technology? • Concerning governance issues: Can decision making on nuclear issues (e.g. energy policy, waste management, nuclear safety, emergency planning) be made in a more transparent and accountable way than demonstrated by the (Belgian) historical experience (or can nuclear power only thrive in a more ‘closed’ corporatist structure)? Do the necessary scale of investments in nuclear infrastructure and the need for managing terrorist threats and non-proliferation requirements limit the chances of public participation in decision making? What can be learnt from a comparison with other countries that have embraced the nuclear option in the past? To what extent are the regulatory bodies (i.e. the nuclear safety authority and the national waste management organism) in Belgium truly independent or are they rather captured by their client industries? • Concerning the distribution of economic benefits and burdens: How can a better distribution of the economic benefits of nuclear power (which are perceived by some participants to have fallen mostly to the share of the major electricity generator and large industrial energy consumers in the past) be guaranteed after the existing power plants’ depreciation? Is it justified to finance the ‘nuclear passive’ by imposing levies on all consumption of electricity? • Concerning global equity: Can we simply assume that developing countries will acquire the necessary capital and technical expertise required to operate nuclear power stations in the future? Can we justify and approach which reserves the use of nuclear power only for industrialised democratic countries (i.e. will this approach lead to a two-class system of nuclear power, with certain technologies and/or fuel cycle options denied to one class of countries, while permitted in ‘safe’ countries)? Undoubtedly, these are very difficult questions to answer621. Nevertheless, if one wishes to engage in more deliberative forms of governance (as we propose), it is unlikely (in view of the results of the present MCM exercise) that they can be avoided, as the responses clearly influence the perspectives taken by many stakeholders in the debate on the future orientation of energy policy.

620

See e.g. Feiveson (2003) and MIT (2003) for reflections along these lines. We should remind the reader here of Ralston Saul’s adage that “…a true question – a question seeking truth without expecting to find more than a fragment of it – will remain clear and unforgiving over hundreds of years…” (Ralston Saul 1995, p.1). 621

458

4.1.4

Chapter 7

The ranking profiles

The conclusions that can be drawn from the ranking profiles as derived during the present MCM exercise with regard to strategic decisions concerning the future development of the Belgian energy system must of course, in view of the inherent limitations of the method, be regarded as tentative. Further investigation on a number of issues and deliberation in a more ‘political’ setting might very well change the overall complexion of the results. Still, we believe that some general insights can be gleaned from the present exercise, based on some emergent regularities in the ranking profiles of individual participants. For instance, from the ranking profiles it is clear that the ‘nuclear’ scenarios only represent a robust first choice (i.e. against the sensitivity analyses performed on the scores and weights) under those perspectives which tend to pass over the questions raised in section 4.1.3 (i.e. in the perspective of participant D2 en F3, in the latter case being further strengthened by land use concerns). Scenarios counting on the ‘carbon capture and storage’ option are generally met with scepticism – none of the ranking profiles prefers this option over all others. In contrast, and even setting aside those perspectives which are generally favourable towards scenarios relying neither on nuclear nor carbon capture and storage technology, it must not be forgotten that all scenarios use renewable electricity production up to its maximum capacity (as defined in the MARKAL database), which is substantially higher than the present renewable contribution to the supply mix. On the basis of a set of results such as these, it would be difficult to justify any regulatory intervention in electricity supply markets which acted to favour nuclear power or carbon capture and storage technologies over renewables. Of course, what counts as a measure ‘favouring’ nuclear power over renewables will be open to interpretation, but it can safely be assumed that at present, any questioning of the nuclear phase out (as stipulated by law in 2003) will be met by fierce resistance from those perspectives which tend to stress the historical ‘injustice’ of past energy policy in favour of nuclear power (over any other option). If we are allowed to venture into a personal opinion here, we would suggest that any political move with the aim of dismissing the phase-out law would seriously thwart any attempt at setting up the deliberative governance scheme we envisage. Finally, the evident discrepancies between the ‘optimal’ strategies envisaged by different perspectives at the present stage of the debate can be seen in itself as a compelling indication that some degree of diversity would be desirable. Despite the fact that most individual perspectives do not specifically address the need for diversification, we would suggest that the evident degree of discord in the appraisal of the different strategies presents a strong case for a policy intervention which deliberately fosters a degree of diversity in the energy supply mix (cf. Chapter 6). This observation raises questions over the extent to which R&D and regulatory policy should be geared towards an active encouragement of a variety of technologies and practices. In particular, questions regarding the balancing of nuclear (a distinction has to be made between existing challenges such as high-level waste management, reactor safety, radiation protection, etc. and research on innovative reactor concepts, including fusion) and non-nuclear related energy research cannot be avoided.


459

Of course, the actual pursuit of technology and/or policy options in the ‘real’ world can never be compared with the protected settings of a multi-criteria appraisal exercise. Therefore, the ‘optimal’ strategies identified under different perspectives can never function as a simple ‘yardstick’ against which to measure real-life developments. Instead, if MCM were to become a ‘running practice’ of the deliberative ‘energy agency’ we envisage, it would generate a set of possible ‘moving targets’ (cf. Chapter 6) towards which future strategies might be configured to aim. The targets for policy making identified in this way would be ‘moving’ both in the sense that they will be constantly changing when new empirical information becomes available and because of the shifting nature of the mix of perspectives. With the broad targets for policy intervention thus identified there is of course ample room for the construction of more detailed policy measures aimed at their implementation.

4.2

4.2.1

The ‘process’

Reflections related to the ‘experimental’ nature of the exercise

It is necessary to keep in mind (and this cannot be stressed enough) that the MCM process had an explicitly ‘experimental’ context. The fact that the MCM exercise was set up as an experiment rather than as an official consultation initiative (and the fact that this was very clear to all parties involved) must be seen as having a bearing on the results. Of course, the organisational procedures and structures which determine the direction that ‘real’ negotiations concerning the sustainable energy future should take (e.g. concerning the participants allowed in the debate, the roles assigned to them, the support given for substantiating their positions, etc.) will have a large influence on the drawing up of the list of ‘actants’ to be considered and the long-term energy scenarios. Furthermore, relationships will develop during the ongoing negotiations, possibly changing the identities of the participants; and they will of course be influenced by previous experiences. It is also likely that within an experimental situation some participants (especially representatives of interest groups) felt freer to ‘stretch’ their opinion to a degree that may not, at present, be felt to be tolerable in a ‘real’ consultation. Thus, any conclusions drawn from this MCM exercise must be accompanied by some necessary qualification. For instance, to some extent it can be said that participants in the exercise were self-selected, in the sense that potential participants were often confronted with other responsibilities or the constraints of busy schedules. It can be argued that as a group, they were neither a ‘statistically valid’ or otherwise representative sample, nor were they necessarily the ‘usual suspects’ who would come forward in the case of an actual energy policy controversy. Beyond this, further minor reservations may be expressed concerning the information available to the participants in the scoring exercise, the amount of time available for the weighting of criteria, and the relationship between the ranking profiles and the scenarios used for the present exercise.

460

Chapter 7

Nevertheless, as we have tried to point out in the preceding sections, there are many senses in which the principal findings of this MCM study can be regarded as robust. For instance, although only ten in number, the participants spanned a wide range of the perspectives currently reflected in the debate over sustainable energy622. They also embodied an array of expertise and institutional experience, although expertise in engineering/economics was underrepresented. The general consistency and coherence of user perspectives was guaranteed through the use of tools such as the structured ‘combined value tree’ and a constant documenting of the assumptions used in scoring different scenarios. In this way, the degree to which uncertainties and discrepancies between the positions taken by different participants have been made explicit by the MCM methodology should not be seen as a lack of robustness on the part of the exercise as a whole, but rather as a principal advantage, serving to render deliberations more complete, transparent and systematic. Therefore, if interpreted carefully, perhaps the most useful way of making sense of the results of the present exercise is as a potential guide to the design of further appraisal research in the field of sustainable energy, addressing some of the gaps, ambiguities and question marks as a support for a deliberative governance model. 4.2.2

The importance of trust

Arguably the most important success factor for the type of consultation envisaged by us is the degree of trust invested in the process proceedings by the different participants. The development of trust and understanding is essential to the whole deliberative enterprise and can be seen as at the same time a prerequisite for successful dialogue, an emergent quality and an outcome that is desirable in its own right. However, trust in a process is a very complex variable that depends on the interplay of numerous other factors. Besides the transparency of the ‘products’ of a deliberation (cf. supra) – in terms of the traceability of the results – it is for instance also essential that the processes used for consultation are as transparent and legitimate as possible to all participants involved. This requires clear and accessible information concerning the structure of the process, its aims, its remit, the identity and interests of organisers and sponsors, and the expected roles of the different participants. On this issue, the telephone interviews conducted independently afterwards by STEM clearly revealed that in our case, we were handicapped by the long time span between the different phases of our participatory fieldwork (almost two years elapsed between the first interview sessions and the completion of the MCM exercise) (Keune et al. 2004, p. 17). Most participants were unable to reconstruct a clear image of the logical sequence linking the individual interviews, the scenario workshop and the MCM exercise. Consequently, the telephone interviews revealed a large spread in the perception of different aspects of the entire participatory exercise, e.g. with regard to its aims (ranging from ‘substantial’ aims such as a discussion of the role of nuclear power in sustainable energy scenarios to 622

We repeat here that members of the FRDO were not chosen on the basis of statistical representativeness for ‘public opinion’.


461

‘methodological’ aims such as finding an adequate way to involve stakeholders in this question), the expected role of the participants (ranging from simply mapping different perspectives to seeking consensus through the use of ‘objective’ arguments), the reasons for participating (ranging from a methodological interest in the project to an interest to learn more about sustainable energy). This finding was of course exacerbated by the fact that only a minority of participants actually participated in all three phases of the research. Hence, it must be appreciated that, despite best efforts to provide clear and comprehensive briefing, participants – especially in longer processes without ‘real’ impact on decision making – will not necessarily grasp the totality of the process as envisaged by the researchers. This is certainly an important point to retain for possible future applications, albeit that it can safely be assumed that, if scenario exercises and multi-criteria appraisals would be embedded in a ‘real’ consultation procedure, the above-mentioned problems would certainly be lessened (because of the greater effort invested by stakeholders in following up the process). However, this apparent ‘fuzziness’ in the minds of the participants does not appear to have influenced their trust in the process, as most participants expressed an interest in the methods used and a satisfaction with its results (bearing in mind some critical remarks of course). The most notable exception proved to be the two representatives of environmental NGO’s, who proved to be (very) sceptical of the approach. The telephone interviews afterwards revealed that transparency concerning the identity and motivations of the organisers of a participative process is a particular matter of concern. For instance, one of the representatives of environmental NGO’s (participant A2) mentioned that a certain ‘hidden agenda’ could be expected since the sponsoring organisation for the exercise (SCK•CEN) had an interest in maintaining the level of nuclear electricity production in Belgium. He saw the fact that the discussion of scenario outcomes (in the report distributed before the MCM exercise) focused largely on electricity production and its consequences as proof of that his fear of a ‘hidden agenda’ was justified623. The other environmental NGO representative (A1) was even more outright in his accusations of a ‘hidden agenda’, seeing the whole approach a ‘long roundabout way’ set up by SCK•CEN to arrive at useful sustainability criteria for supporting a research project on a possible future fuel cycle option624. In a way, it could of course be expected that developing trust in a deliberative process involving a possible future role of nuclear power would be most challenging from the point

623

This participant (in the telephone interview) commented that “…the seeming thread in the whole exercise was to investigate the meaning of sustainable energy in the Belgian context; while in reality the intention was to prove that Belgium could not do without nuclear power…” (our translation). 624 He referred to the MYRRHA project. MYRRHA is a so-called ‘accelerator-driven system’ (ADS) aimed at providing protons and neutrons for various R&D applications. It consists of a proton accelerator coupled to a subcritical fast core. Among the applications that are considered, the MYRRHA facility will be used to enhance new R&D activities in waste transmutation. Transmutation, in brief, is the transformation of one isotope into another one by bombarding the atom with a high neutron flux. The underlying idea is to transmute long-lived fission products and minor actinides – which (due to their long half-life) are the most ‘annoying’ form of nuclear waste – into less long-lived waste, reducing the burden on the nuclear waste stocks.

462

Chapter 7

of view of involving environmental NGO’s. From the comments made by the environmental NGO participants in our present ‘pilot exercise’, it is clear that we have not been able to secure this trust. Certainly, one can take a fatalistic attitude and assume that the level of trust we are aiming for is simply an unattainable ideal. Other sociological findings seem to support such attitude. For instance, Mormont (1995) observes that environmental NGO’s sometimes have an interest in maintaining a the ‘controversial’ status of certain questions (rather than trying to solve them, e.g. through collective deliberation); in any case, Mormont argues that a substantial involvement of environmental NGO’s in deliberative procedures would also imply a substantial shift in their organisation mode (which is, in Belgium at least, rather aimed at influencing public opinion)625. However, we are of the opinion that improving trust is not unfeasible altogether. Certainly, it must be appreciated that the development of authentic trust and understanding takes time and this process cannot be short-circuited. But, according to us, ‘fertile’ conditions can be optimised, and in this sense the criticisms received regarding our ‘pilot exercise’ potentially point towards possible improvements (e.g. ensuring sufficient and independent expertise, addressing all perspectives in scenario exercises, ensuring full openness about the aims and contexts of participative processes, etc.). Also, the fact that a ‘real’ consultation procedure would probably not be conducted by a researcher affiliated with the SCK•CEN, but rather by an organisation that is seen to be independent by all participants (cf. our reflections on the role of the ‘facilitator’ in chapter 5), would certainly add to its perceived legitimacy. 4.2.3

Interaction and deliberation in the participative process

Two issues seem to be of further importance for discussing the procedural qualities (or lack thereof) of the participative process designed for the purposes of the present dissertation, that is the interaction between researchers and participants (in the interviews, scenario workshop and the MCM exercise) and the possible interactions between participants if MCM were to be adopted as part of a deliberative governance process. The first issue (interaction between researchers and participants) has been touched upon here and there already in the present chapter. As explained, we took care to restrict our intervention as researchers as far as possible to giving responses over questions of methodology and eliciting participants’ perspectives on criteria definition, the framing of different options, scoring and weighting options, etc. Most participants were satisfied with this approach as it was generally felt that the degree of self-sufficiency allowed in the process we set up allowed for their perspectives to be adequately addressed. However, as mentioned, environmental NGO representatives felt that their perspective had not been fully respected in setting up the basic options for appraisal in the MCM exercise. The other issue (interactive deliberation between participants) is potentially more important, particularly in view of setting up the deliberative governance scheme we

625

For a similar argument based on case-study research, see Fung and Wright (2003).


463

propose in chapter 5626. Of course, a crucial feature of the present exercise is the restriction of the analysis to the exploration of positions taken by individuals in a relatively isolated methodological setting627. The crucial question then becomes: ‘Can MCM help in setting up a truly deliberative environment in a ‘real’ political setting?’. This is an important question which, in view of the limitations of the present exercise, can only be answered by further research and experimenting. Let us not forget that creating a deliberative environment is a very demanding process, generally requiring the simultaneous fulfilment of a number of issues discussed here – i.e. the importance of developing inclusive knowledge (cf. Section 4.1.3), developing a degree of trust in the process (cf. Section 4.2.2), the creation of at least some common understanding, etc. Furthermore, it must be recognised that, even where the basic prerequisite conditions of trust are present, a successful deliberative environment requires ongoing commitments from participants. Generally, the level of discipline and energy required for participants to remain open to sometimes radically different perspectives and/or types of knowledge should not be underestimated. On this issue, it is clear that different participants displayed a markedly different attitude towards deliberative requirements such as a willingness to question one’s own perspective (e.g. through the expression of uncertainties) and/or adopting elements from other reference frames (e.g. by choosing criteria from different categories). Clearly, this proved to be an easier task for participants without any direct vested interest in the issue at stake (e.g. representatives of advisory councils). We would suggest that involving such people (who could function as a ‘bridge’ between the opposing ‘dominant’ perspectives) in a deliberative governance scheme could prove to be a crucial success factor. Anyway, it must not be forgotten that establishing a high-quality deliberative situation takes time and can certainly not be ‘rushed’ by a quick ‘pilot exercise’ as we undertook. 4.2.4

Using the MCM method for decision support

As mentioned before, both representatives of environmental NGO’s (A1 and A2) had a somewhat negative view of the usefulness of the MCM method for decision support. According to participant A2, the fundamental need for Belgium on its road towards a more sustainable energy future is not a more ‘balanced’ debate and decision support, but rather more communication and sensitisation (in the public opinion, firms, etc.) that a lot can be gained from saving energy (with the expectation of course that the potential for energy saving is larger than usually admitted). According to this person, the problem is one of mobilisation and stimulating creativity in a search for new solutions, rather than a more careful ‘reasonable’ approach. On the other hand, most other participants were of the opinion that the MCM method – and in particular its capacity for integrating both qualitative and quantitative information – could be a useful support for decision making,

626

As explained in chapter 2 (Section 3.4.1), deliberative democratic theories generally agree on the importance of unconstrained dialogue, inclusiveness and social learning. 627 Of course, in the scenario workshop the different participants were present in the same room. But this workshop was mainly set up for the purpose of making participants more familiar with the scenario approach adopted for the MCM exercise, rather than with the purpose of substantively influencing the entire scenario approach.

464

Chapter 7

although their approval remained at a rather ‘diffuse’ level. From our experience with the MCM model, we can nevertheless suggest some deeper insights. Indeed, experiences from our project suggest that broadly speaking, participants who had a generally positive attitude towards our approach attributed two possible meaningsto the ‘use’ of formal decision modelling in public agencies628. Some of the participants regarded the methodology primarily suitable for instrumental use – in other words, they considered decision analysis as a way to construct a frame of reference for evaluating decision alternatives. Participants favouring this approach usually chose ‘technical’ (i.e. quantitative) criteria from the value tree in order to arrive at a more ‘objective’ decision. Others saw a more conceptual use for formal decision modelling – i.e. for structuring and decomposing policy problems, aggregating expert knowledge from different areas, taking multiple objectives into account explicitly, etc. Of course, this second meaning corresponds more to the use we ourselves had in mind. Extending this understanding to all participants thus would require a greater communicative effort. Regarding the conceptual use of the MCM model we envisage, it can be said that one of the key findings of this exercise is hardly suprising: how participants ‘framed’ the decision problem (the overall worldview they brought to the analysis as reflected in the mix of criteria they considered to be important) proved to be decisive for the ranking of the energy scenarios. This result is quite consistent with the view of many experienced decision analysts. An official manual by the British ‘Department of Transport, Local Government and the Regions’ (DTLR 2001) even argues that …the time spent determining the criteria in any multi-criteria analysis is the most important time of all, and generally much more so than excessive fine-tuning of the numerical detail of the models themselves…

Similarly, the qualities displayed by MCM would likely be disputed by few, at least in principle629. In this respect, it might be asked what precisely the application of MCM to our topic might achieve. To this (and similar) thorny questions, we would counter that perhaps the greatest benefit of the application of MCM in the debate on sustainable energy might simply be a demonstration effect to those responsible for decision making that different actors’ assessments frame the decision problem in a very different way. We agree with Stirling (1997) that those advocating an instrumental use of decision-analytic models will mainly be sceptical of MCM because of what is perceived to be its ‘overly ambitious’ character (i.e. achieving a certain degree of clarity in interminable socio-technological controversies).

628

One should not forget of course the possibility of strategic use of formal decision models. Such use will of course generally not be admitted openly, but even when strategic use is in play, it presents no serious difficulty if embedded in an ongoing deliberative practice as we ultimately envisage (cf. Section 3.5). 629 I.e. the possibility for pluralistic assessment, a large degree of attention for the process of criteria selection (and the inherent multi-dimensionality of the problem at hand), an open acknowledgement of the unavoidable subjectivity (but not irrationality!) of the results obtained, a more pragmatic humility in the face of uncertainties, transparency and accessibility to independent scrutiny and wider participation (besides experts in certain domains), and open-endedness (Stirling 1997)


465

However, according to us, the most ‘ambitious’ part of MCM is the explicit attempt of introducing deliberate social choices into a ‘scientific’ method. In a sense, it has been a recurring idea throughout this dissertation that almost irrespective of the particular analytic techniques employed, all approaches to technology, policy or environmental appraisal incorporate implicit political choices. Therefore, making these choices more visible is for our purposes an indispensable (and nonnegotiable) quality, which might of course be resisted by some. Provided that it is clearly communicated that the outcomes of an MCM exercise are not to be interpreted as providing the basis for a decision (in one way or another), MCM according to us does provide a useful extension to other existing approaches that allow problems to be ‘mapped’ in a formal sense and then debated. Compared to other approaches, MCM has the attraction that it uses a process that draws upon a significant body of scientific literature on multi-attribute utility theory (and hence, will likely be credible to scientific experts) but is also broadly familiar among wider audiences. Furthermore, because of its structured and quantitative form, it allows aspects of problems to be explored that other mapping procedures do not. Therefore, especially if the method were to be supplemented in the future with more precise indicators for some of the criteria (possibly determined by a panel of experts subject to approval by stakeholder representatives in the new ‘energy agency’), it might well serve as a ‘boundary object’ or framework able to ‘sustain’ the debate on sustainable energy in a structured way over time. Thus, if the appraisal process is understood (by all!) to be iterative, reflexive and trustworthy, involving no requirement that the perspective taken at any one moment becomes ‘irrevocable’, then we might hypothesise that there can be a corresponding degree of trust in the method as such. Similarly, under such conditions stakeholders might come to understand that there is no danger of particular weightings becoming ‘reified’ or ‘manipulated’ by political adversaries. Such perceptions are important since truthfulness of the participants’ judgments – i.e. the weights and scores used in the MCM exercise truly represent what a participant thinks about the scenarios, rather than being an attempt to manipulate a final result – is an essential requirement for the meaningfulness of a conflict analysis as presented in this chapter.

5 Summary and conclusions In this chapter we implemented and tested multi-criteria mapping, a method stemming from the broad research tradition of multi-criteria appraisal and multi-attribute utility theory, as a possible tool for deliberative governance. The initial attraction of this tool lies in its capacity to interactively explore uncertainties in sustainable energy policy in relation to different scientific framings, stakeholder perspectives and broad strategic options. In individual interviews, a ‘mapping’ of different long-term energy scenarios for Belgium (developed specifically for the purpose of these MCM interviews) has been carried out. The MCM exercise, as applied here, consisted of four main stages: attribute selection (in which actors identified around ten main areas of concern for sustainable energy policy); attribute rating (in which an assessment is made of the relative importance of the different

466

Chapter 7

concerns); option evaluation (in which the actors are asked to score the different energy scenarios for each of the attributes); and experimentation (in which actors are encouraged to explore and modify their point of view based on a graphical and tabular representation of the results). It is clear that this limited ‘pilot exercise’ does not enable us to resolve all of the questions that might be raised over the mode of implementation in an institutionalised context for an operational MCM procedure. Nevertheless, we believe some relevant conclusions can be drawn on the capacity of MCM to function as a ‘boundary object’ in the present debate on the future of the Belgian energy system – i.e. its capacity to remedy the current lack of transparent interaction at the interface between science, politics and society. Firstly, we have given a lot of attention to the development of a ‘combined value tree’. Here, we have tried to strike a balance between a focus on overly prescriptive ‘technical’ criteria whose spurious precision can foster interminable controversy (cf. Chapter 3) and unduly ambiguous approaches which fail to allow the drawing of any meaningful conclusions. Blueprints, best foreign practices, international codes and standards, harmonisation, etc. for deriving sustainability criteria can do the trick only for some of the narrowly technical issues (e.g. health impacts). But on a whole, the list of criteria and indicators needs to be adapted to – and requires a process of discovery about – local needs and capabilities. Here, Boltanski and Thévenot’s ‘commonwealth model’ proved to be invaluable in setting up a coherent hierarchy of different matters of concern based on both ‘technical’ (mobilising actants in the industrial and/or market commonwealth) and more ‘political’ concerns (mobilising actants in the civic commonwealth). Participants in the MCM exercise were generally able to reflect their perspective through an appropriate choice of criteria, suggesting that the present combined value tree can be seen as a first positive step in the direction of a reconciliation of ‘science’ and ‘subjective factors’. Unfortunately, another important aspect of the approach – i.e. the development of scenarios as a support for deliberative discussion – has proven to be more unsatisfactory630. Here, requirements have to be met which are often difficult to reconcile. Some participants stress the need for validity (as attained through formal mathematical modelling), others have stressed the need for stretching the scope of these scenarios and providing more contrasted narratives. Furthermore, mathematical modelling is often seen to conflict with demands of transparency. These criticisms thus point out a need for further improvement in (interactive) scenario development, an issue that would certainly require more attention (and resources) in the future. At present, some broad suggestions have been given in section 4.1.2.

630

Thus, referring to Figure 18, the present situation would fall into the ‘second case’ – i.e. one of agreement on the list of ‘actants’ to be taken into account, but disagreement on the future actants’ association to common worlds.


467

The individual MCM exercises have confirmed the centrality of the issue of ‘framing’ in this kind of sustainability appraisal. Participants took more time to discuss the framing (e.g. through a selection of criteria, contextual assumptions when scoring options, etc.) than in any other phase. However, generally a need was also expressed for more ‘expert’ support in scoring the criteria deemed to be relevant under their perspective, e.g. concerning the choice of data sources or appropriate methodologies for addressing the issue. Thus, in the future, it would certainly be advisable to complement to use a variety of specialists (subject to the approval of the different stakeholders) to score the scenarios on the various criteria. That way, MCM could be used for structuring and clarification of individual opinions and for providing a common framework in which all the diverse aspects linked to the general sustainability concept could be addressed. It could also enable focusing the debate on the crucial issues and eliminating unimportant factors, and making different perspectives understandable to others by making judgments explicitly open to criticism. Ideally, MCM would be part of the broader deliberative governance mechanism we have developed in chapter 5, where due decision-aiding ‘platforms’ are foreseen to let actors express their perspectives, experiences and knowledge requirements. Perhaps we are now in a better position to explain what we really expect from the application of MCM in the broader governance context exposed in chapter 5. The role of the central ‘energy agency’ should be conceived of as one of seeking balance between a number of well-articulated ‘matters of concern’ (i.e. sufficiently supported by evidence in the form of indicators, expert witnesses, etc.). For this, a number of ‘needs’ have to be addressed, such as the provision of a forum for participation (with all the procedural guarantees this entails) and the development of the state-of-the-art relevant (i.e. inclusive) knowledge (i.e. the energy agency will have to be sufficiently staffed by independent experts). Important as addressing these needs might be, they should not lead us to believe that the ‘energy agency’s’ primary task is to find the definite answer to the questions elicited by the search for more sustainable forms of energy production and consumption. Rather, we submit that its primary task lies in trying to establish the areas of ‘reasonable doubt’. In performing this task the ‘energy agency’ would be quite unique, in the sense that it provides something which escapes the political decision maker (who, whether he/she likes it or not, must govern) or the individual expert (who, ideally at least, should remain within the boundaries of his/her scientific discipline). Therefore we believe that one of the greatest needs today in facing the sustainability challenge is to find ways, even very simple mechanisms, that will help to insert this ‘reasonable doubt’ into the public debate. The results of our MCM ‘experiment’ seem to suggest (taking into account all caveats however) that an ‘improved’ MCM version (along the lines suggested in section 4) could take up this role.

CHAPTER 8 SUMMARY, CONCLUSIONS, RECOMMENDATIONS AND FURTHER RESEARCH In this chapter, we first summarise the main lessons and conclusions resulting from our These lessons are subsequently translated into policy research (Section 1). recommendations (Section 2) and suggestions for further research (Section 3).

1 Summary and conclusions After the relatively quiet eighties, energy has gradually forced its way back on the political agenda on the international, regional as well as the national level. Whether swamped with tales of the catastrophic consequences of climate change, nuclear accidents or the more insidious poisoning of our environment, or startled by sudden price spikes on the international energy resource markets, the realisation that we have to modify our habits of energy ‘production’ and use in a quite radical way slowly gains hold over the collective consciousness. On the level of international politics, this ‘collective disquiet’ – projected on the larger canvas of global interactions between economy, society and the environment – has instigated a search for a new legitimation of the development idea itself. On a conceptual level, a new development paradigm called ‘sustainable development’ was launched, rising rapidly into prominence after the publication of the ‘Brundtland report’ (WCED 1987). Despite the often significantly different stresses in the literature on the subject, appeals to ‘the’ (or should we say ‘a’) principle of sustainable development nevertheless refer to a more or less stable ensemble of characteristics. Taken together, and formulated at a fairly general level of ‘everyday’ or ‘manifest’ understanding, these characteristics reflect a commonsense view on the shifting focus of development priorities: 1. development cannot be simply equated with economic growth; 2. issues of intergenerational equity have to be addressed; 3. consequently, more attention should be paid to the long-term effects of a proposed development (i.e. prospective analysis); 4. development problems should be analysed in an integrated way, taking into account the interplay between its economic, social, institutional and ecological components; 5. possible impacts on human health or the environment should be at the centre of attention; 6. a principle of intragenerational ethics dictates that one should avoid passing on the negative impacts of local developments to other people on the globe (taking into account ‘common but differentiated responsibilities’); 7. when assessing these impacts, a principle of precaution should be adhered to; and 8. proposed developments should provide ample opportunity for public participation at all stages.

470

Chapter 8

Now of course, these characteristics are all rather ‘soft’ and ‘vague’ and hence open up the way for (possibly conflicting) further theoretical interpretations and elaborations once the principle is put into operation. And yet, as we have argued in chapter 1, one should avoid seeking solace in precise definitions for two reasons. The first is that these characteristics cannot be understood in an abstract sense, far removed from practical application in a specific context. They can only be understood through use (and misuse) – and that is also why we took the time to discuss some practical cases in this dissertation. Secondly, these characteristics cannot be defined in their ‘true’ sense by the tools of lexicography, but only in their relationship to each other. Each attempt at defining sustainable development, whether it is motivated by an ideological drive towards finding some kind of an ‘ethical fundament’ (thereby denying the deep divisions of value pluralism in contemporary society) or by pragmatic considerations (e.g. only allowing established technical elaborations of the concept), is at fault in the sense that it seeks to make the plural a priori singular – that is, before any ‘construction’ has taken place. In contrast, we have defended the inherent ‘vagueness’ and ‘semantic slipperiness’ of sustainable development as an asset rather than a drawback, in the sense that it at least creates an opening towards collective thinking and acting (even if it turns out later that some perspectives simply cannot be reconciled). In chapter 1, we have argued that the truly political importance of overarching ‘ideologies’ such as sustainable development lies in the recognition of demands of ‘entities’ which have been by and large left out of the official political discourse before (e.g., in the case of sustainable development, ‘ecosystems’, ‘future generations’, ‘the poor’, etc.). And of course this makes such concepts essentially contestable and open to political struggle once the terms under discussion become subjected to more precise interpretations. Accepting the principle of sustainable development as a starting point for reflection (which can only be substantiated further by a constant dialogue in practical cases) also implies accepting a correspondingly ‘open’ problem framing. In this dissertation, we have developed our reflections based on the practical case of the ongoing (nuclear) energy debate in Belgium. The overarching research question guiding our research therefore became: ‘Can nuclear energy contribute to a development process leading to sustainability – if so, how; and how can this question be answered?’. From this broad question the following sub-questions can be derived: • ‘Which meanings are attributed to sustainable development in general and in particular in the context of energy policy?’; • ‘Which scientific approaches and theories have been advocated and to what extent do they address the full scope of questions raised by demands for more sustainability, and in particular those concerning the institutional dimension of public policy making? What are the strong and weak points of these approaches; which lessons can be learnt?’; • ‘What then is an ‘appropriate’ scientific/political methodology or procedure to address sustainability questions?’. Throughout our search for answers to these questions we have followed three research strands, which however constantly intermingled.

Conclusions, recommendations and further research

471

In a first strand, we have contributed to the ongoing debate between different (meta-) theoretical philosophical perspectives on the relationship between (technological) development (a project underpinned by deeply-rooted assumptions about the nature of modernity) and the environmental challenge. We argued in favour of a constructivist position, enabling us to conceive of sustainable development in terms of a never-ending collective learning experience, based on ‘agonistic’ interactions between different perspectives on sustainability. The second research strand has consisted of a number of case studies, exploring how nuclear energy is positioned in practice in the sustainable energy debate. Empirical data were collected on three levels. Firstly, on the level of ‘science for policy making’, we have analysed a large-scale European research programme attempting to reveal the ‘true costs’ of different energy options. Secondly, on the level of national energy policy, we have made a reconstruction of the policy development process leading to the Belgian decision to phase out nuclear power (framed as a move towards sustainable development). Thirdly, on the level of the societal debate on (sustainable) energy policy in Belgium, we have analysed the positions taken by some of the most important stakeholders in the debate. In this case, it was instructive to analyse if and how the advent of sustainable development as a guiding idea for future developments could produce any significant changes in a historically polarised debate. Finally, in the third research strand we have tried to combine the insights gleaned from both theoretical and empirical investigations into a practical proposal for a new governance structure for sustainable energy set in the institutional context of Belgian energy policy. In the following paragraphs, we look back on the central findings of these three research strands in terms of providing answers to the first two research questions. The third research question, which deals with the more practical aspects of the overarching question in terms of a proposal for a new governance structure, will be dealt with in the ‘recommendations’ section of this chapter (Section 2). 1. The first research strand we dealt with sustainable development on a rather abstract and ethereal philosophical level. We felt that such ‘abstract’ conceptual analysis was needed since the conceptual deep structures in a piece of theorising determine, among other things, its specific validity claims and, correspondingly, the distribution of covert and overt burdens of proof. As Latour reminds us, the language used in a sociological account is the equivalent of the natural scientist’s laboratory – hence, we saw it as a necessity for us as researchers to be clear about the ‘laboratory tools’ we were going to use when approaching our subject. In chapter 1, we have argued in favour of a constructivist philosophical stance, understood not as a systematic theory but rather as a set of attitudes aimed against particular basic philosophical principles that might form obstacles to a ‘productive’ solution of problems and collective learning. These theses can be summarised as ‘antifoundationalism’ (in the sense that, fundamentally, all our convictions are of a provisional nature and hence susceptible to repeal or review) and ‘anti-essentialism’ (in the sense that

472

Chapter 8

constructivism elides the common distinction between practical and theoretical knowledge – or, for that matter, other binary distinctions such as nature/culture, fact/value, subject/object, context/content, etc. – by arguing that the practices involved in the deployment of knowledge are always pivotal considerations). The great advantage offered by constructivist insights is that they have succeeded in opening up the ‘black box’ of technological and scientific developments and have revealed the intimate intertwinement of technology and society in minute detail. Our scientific and technological knowledge is not so much the mirror image of an inexorable process of modernity entering into a ‘reflexive’ phase (i.e. the essentialist position defended by Beck) or of a particular configuration of powerful social interests (i.e. the social constructivist position), but rather mediates on behalf of hybrid entities called ‘actants’ (a notion introduced by Latour to cancel out the traditional subject/object distinction), and therefore also inevitably transforms them (i.e. the constructivist – without the ‘social’ – position). However, it is clear that, while these basic constructivist insights are helpful for avoiding certain inflexible philosophical stances on a meta-theoretical level, one also needs a more ‘positive’ base to start from in order to look for constructive solutions to the sustainability problem on the more practical level of governance. If absolute knowledge is not available due to a lack of metaphysical or transcendent guarantees, this does not mean that we are left at the mercy of universal doubt. Knowledge can be more or less reliable (or useful, productive, etc.), and this reliability can only be assessed in terms of the collective learning experience itself. In view of our research interests, this meant that we needed a theory which allowed offering an interpretation of the interplay between scientific and institutional and/or political practices commonly encountered in the field of governance for sustainable energy. In chapter 1, we initially proposed to use Boltanski and Thévenot’s commonwealth model as a ‘reliable’ knowledge framework fit for our research purposes, though we could not yet fully prove this point at that stage. It is only through the account given of the subsequent phases of our investigation, oriented at both theory and practice, that the reader can form his or her own opinion on the model’s utility. 2. After laying out our meta-theoretical foundations (or rather, anti-foundations), we have continued our theoretical reflections with an investigation of technology governance from a constructivist point of view. The constructivist vantage point helps us in conceiving of governance not only as a method for choosing the adequate ‘means’ for furthering some given (and generally accepted) ‘ends’, but also crucially as a cooperative scheme (i.e. a complex composed of rules of interaction, beliefs, procedures etc.) which delineates ‘relevant’ knowledge for the policy problem at hand, as well as the scope of ‘admissible’ arguments and the corresponding in- or exclusion of perspectives in the formulation of governance strategies. Fundamentally, all governance practices raise boundaries between certainty and uncertainty, fact and value, and relevance and marginality. In chapter 2, we identified four ideal-typical governance schemes. These were construed not as reflections of actual political practice – which will always escape any ‘simple’ attempt at classification – but rather as ‘scenarisations’ aimed to be helpful in noticing and understanding the one-sidedness of certain ‘idealised’ approaches.


473

The expert-based governance scheme operates according to a (deep-seated) logic that political authority and decision making can be separated from the scientific authority provided by experts, based on disciplinary competence. It is an efficient scheme for arriving at solutions when ‘facts’ are generally undebated. However, difficulties arise when value choices are (often tacitly) delegated to expert committees, or when abstract scientific insights (and the models of human behaviour they often incorporate) prove to be ineffective when applied to multi-dimensionality of practical contexts. Governance by aggregation seeks a resolution of policy problems through negotiations between different ‘interests’, but is mostly fraught with difficulties when applied beyond a short-term perspective. A controversial definition of the nature of a policy problem (in terms of ‘incommensurable’ value positions) and its foreseeable evolution in the future makes the identification of ‘interests’ a hazardous undertaking. Governance by pacification represents a widespread political strategy to tackle (politically organised) value positions of a different kind (which cannot be reduced to the common ‘interest’ denominator), and has shown its value in the past, also in Belgian energy policy. Nevertheless, be it from a moral (e.g. pacifying relations between social groups through more or less ‘distorted’ communicative settings), an empirical (e.g. a possible lack of sufficient learning capacity in the context of environmental or technological controversies) or a pragmatic (e.g. a possible lack of real influence on decision making) point of view, one can only be sceptical about the potential of such attempts at pacification for pushing towards a change in the direction of sustainable development – at least when they are not connected to a wider participatory practice. A last scheme – governance by deliberation – is strong in its insistence on giving a ‘voice’ to the voiceless and promoting the values of ‘unconstrained dialogue’ and ‘social learning’, as advocated by Habermas in his theory of communicative action. However, Habermas’s perspective, despite its intrinsic appeal for dealing with ‘unstructured’ policy problems, does not effectively counter the challenges of sustainable development. This is because Habermas attempts to ground the search for solutions in terms of generalising and abstracting processes, which tend to deny the true nature of a political conflict, characterised by ineradicable conditions of tension, conflict and contrast. In conclusion, each of the four governance schemes is to a greater or lesser extent deficient in addressing some questions raised by ‘radically uncertain’ conditions. As a result, the governance scheme proposed in chapter 5 – although greatly inspired by these four schemes – still differs in some crucial aspects. 3. Governance by aggregation has found its most theoretically founded emanation in the sustainability debate in the economic theory of externalities. The pricing of power generation externalities, it is often argued, is necessary for making consistent and meaningful ‘rational’ choices between technologies with a view on making development more sustainable. The basic idea behind the calculation of external costs is that all possible impacts of power generation technologies should somehow be defined and managed within the logic of the ‘market commonwealth’. After all, the ‘market commonwealth’ enables measurement of objects on a universally understood numerical scale of ‘grandeur’ (i.e. money). Be it for reasons of inherent communicative simplicity, political expedience or

474

Chapter 8

pragmatic considerations, the view that political interventions for ‘steering’ energy systems towards a more sustainable future should limit itself to internalising externalities certainly holds sway over many actors in the debate, and thus certainly justifies its inclusion as a case study in our dissertation. Using Boltanski and Thévenot’s insights, it is of course easy to level a fundamental critique against this approach. In this vein, (critical) literature on the subject often has recourse to the well-rehearsed argument that pricing externalities is simply ‘unethical’ – i.e. people reasoning from different ‘commonwealths’ will tend to have a fundamentally different ‘ethical’ outlook on the potential use (and desirability) of performing external cost calculations – hence limiting its use in decision-making contexts. In chapter 3 we have argued that this argument certainly merits some attention, but should not be seen as an absolute obstacle to any serious consideration of the value of external cost calculations. This is because it fails to make a distinction between applying external cost calculations in the ‘context of discovery’, the ‘context of justification’ and the ‘context of application’. In the context of discovery, research on external costs is useful if only because the economics of externalities provides us with a rigorous and coherent framework enabling us to actively seek out entities left out of the ‘economic collective’. Furthermore, it instigates a search for the right instruments (e.g. questionnaires, mathematical models, or even political institutions) for giving a ‘voice’ to these entities. The above-mentioned ‘ethical’ critique effectively limits the use of external cost calculations in the context of justification, at least when this justification is viewed as unique. However, if the externality ‘storyline’ (including all necessary presuppositions and simplifications for integrating ‘actants’ within one logic) is presented alongside other possible ‘storylines’ (based on other commonwealth logics), accompanied by a candid statement on any remaining uncertainties and/or ambiguities, a dialogue becomes possible (given the right procedural safeguards – cf. Section 2). Finally, starting from such ‘purified’ accounts, compromises between different positions can be sought in the context of application, e.g. by limiting the use of external cost calculations to certain impacts, certain policy domains, etc. These considerations have led us in chapter 3 to introduce a distinction between a ‘broad’ cost-benefit analysis, comparing ‘costs’ and ‘benefits’ of different ‘scenarisations’, and a more ‘narrow’ costbenefit analysis, reflecting the more common use of this term. 4. In chapter 4, we have analysed how problem structuring proceeded in a ‘real-life’ political setting – i.e. policy development in the case of the Belgian decision nuclear phaseout decision. This decision took place against the background of the progressive inclusion of energy policy in the logic of the market commonwealth, as dictated by European directives on the liberalisation of energy markets. We have shown that the justification given for the phase-out decision was based on an attempt to recast the policy problem in a well-structured technical mould. This was evident from its self-proclaimed reliance on expert opinion (as expressed in the AMPERE report), limitations on the possibilities for ethical debate, treatment of the policy question within the mandate of existing bureaucratic organisations, etc. A detailed analysis has shown however that this ‘technical’ treatment could only be achieved by leaving some ‘white spots’ and/or ambiguities in the


475

justifications given. This finding suggested that other possible views on the role of nuclear power in a sustainable development perspective existed which had to be actively ‘suppressed’ or ‘blurred’ (in order to proceed ‘as if’ there was a consensus). This finding was further strengthened by the empirical part of this chapter, which shows that sustainable energy (and the role of nuclear power therein) is an essentially contested concept, and furthermore that there were virtually no connecting or translating links between the divergent concept and problem framings. Further analysis (based on interviews with key stakeholders in the debate) indeed revealed the ‘unstructuredness’ of the problem. We have identified three ideal-typical positions in the debate (which of course we cannot do justice to here in a few lines). The manager for a large part frames his arguments within the confines of the industrial and market commonwealth. He sees economic growth and technological advance as the most important component of sustainable development, to the extent that actions that might seriously endanger possibilities of growth or competitiveness in general must be discouraged. Governments should set up a stable framework; business will then take up its responsibilities through ‘sustainable entrepreneurship’, ensuring relationships based on trust and consent with concerned parties (labour unions, stockholders, employees, local residents, etc.). There is no reason why electricity generators owning nuclear power plants could not be part of this. The controllist on the other hand believes more attention should be given to the institutional embedding of technology in society. The controllist fears that in the future, nuclear power will be ‘inevitable’ if one wants to respect post-Kyoto commitments and still foster economic growth. Rather, acceptance (or rejection) should be based on a democratic debate with the representatives of concerned parties, under conditions of full transparency. For now, the controllist believes, these conditions have not been fulfilled; too much is left in the dark: costs of decommissioning, costs of high-level waste management, the real costs of the business-as-usual scenario, etc. – all ‘great unknowns’. The reformist perspective sees the evolution of the Belgian electricity sector as an ongoing social process in which scientific knowledge, technological innovation (or the lack thereof in renewable energy technologies) and corporate profit reinforce each other in deeply entrenched patterns, patterns that, according to this perspective, bear the unmistakable stamp of political and economic power. In terms of Boltanski and Thévenot’s commonwealth model, people and objects are artificially kept in a state of permanent ‘misery’: perfectly valid technical options (e.g. renewable energy options) are underdeveloped and ‘rational’ behaviour (e.g. energy saving) discouraged (i.e. ‘misery’ in the industrial commonwealth), the true costs of energy use are being concealed (i.e. ‘misery’ in the market commonwealth), and people are kept in a state of political apathy (i.e. ‘misery’ in the civic commonwealth). Seen from the reformist perspective, nuclear power is not merely the symbol of this social order; it is a true embodiment of that order. Summarising the general picture, it becomes apparent that conflict rather than mutual exchange was the dominant dynamic in the debate surrounding the phase-out decision. Exclusive relations between the different perspectives were caused by competing rationalities on the one hand and the governance framework on the other. Therefore,

476

Chapter 8

conditions for collective learning to occur were severely hampered, since each of these perspectives tends to make use of other different methodological approaches, different problems framings, different data to support their vision, etc. The findings also suggest that, in theory as well as in practice, the sustainable energy debate is dominated at the level of ‘official’ policy-making initiatives by approaches which, metaphorically speaking, tend to conceptualise problems within the boundaries of one or two commonwealths (which of course facilitates the definition of practical solutions). The question is whether these insights can be turned against themselves in order to open up a space for improvement. Therefore, the remaining chapters (Chapters 4-7) were concerned with setting out a proposal for a new governance structure (conceived for the Belgian context) and exploring its practical implications in a tentative way. The findings of these more practicallyoriented chapters (answering our third research question) are reformulated here in the form of recommendations for policy makers.

2 Recommendations The recommendations offered in the following paragraphs should be seen as a contribution to the search for a better model of governance for sustainability, set in the more specific context of (Belgian) energy policy. This model (or cooperative scheme) does not replace the other models we have discussed in chapter 2. Rather, it should be conceived of as an addition to these models, in the sense that it makes possible a substantiated choice between the different cooperative schemes depending on a collective definition of the problem at hand. Furthermore, our proposal is an attempt to unite the particular strong points of each of these models, without inheriting the weaknesses and/or running into contradictions. These strong points are set out as fundamental requirements at the beginning of chapter 5: relying on sound science (reconceptualised in terms of discovering contextually-situated ‘wise decisions’ rather than universally valid theoretical claims), bridging the sciencepolicy boundary (based on an understanding of ‘science-for-policy’ as a negotiation between science and policy communities in the acceptance of problem formulations, methods, adequacy of information, reported uncertainties, etc.), an overriding duty to the ‘good common world’ through collective learning (based on temporary, not permanent, closure), an inclusive dynamic (i.e. actively seeking out ‘under-represented’ perspectives) resting on multi-tiered learning experiences (i.e. not limited to the sphere of ‘official’ policy making), quality control (concerning issues of fairness, competence and transparency), and a candid discussion and representation of uncertainties. However, it is of course not sufficient to merely proclaim an adherence to such ‘pious’ requirements – they have to be realised ‘in the field’ through actual socio-material practices. In the following paragraphs we set out the most important recommendations regarding both theory and practice for realising these requirements ‘in the field’.


477

Recommendation 1: New platforms need to be created for a more genuinely inclusive debate about sustainable energy development. In particular, the creation of a new ‘energy agency’ committed to sustainable development is long overdue. This ‘energy agency’ must serve as a consultative body with an output towards formal democratic institutions, but still must be located at some ‘distance’ from the deliberations going on in these institutions.

A new institution for sustainable energy policy could take many different forms. However, the most important point is that the new body should be able to play its role as a ‘boundary organisation’ between the different ‘spheres’ (the political system, the scientific world, organised stakeholders, etc.) and ‘levels’ (ranging from national energy policy to everyday life world experiences). In connecting to the ‘political system’, the ‘energy agency’ we envisage must – whilst not having a privileged access to the ‘common good’ or the ‘public interest’ – fulfil an essential role of focusing towards institutionalised decision making in legislative bodies. This focusing has two aspects. Firstly, it should help in bringing together all relevant perspectives, arguments and learning experiences. Secondly, it should be able to give a more detailed consideration to arguments than can for instance be expected from the individual decision maker who is often enmeshed in different pursuits at the same time. Ad a)

Ad b)

Ad c)

Ad d)

The ‘energy agency’ has to perform the following tasks: deciding on the ‘entry right’ of particular ‘actants’ (cf. Recommendation 2), equipping these ‘actants’ with the best possible ‘representatives’ in order to better articulate their demands (cf. Recommendation 3), supporting deliberations on strategic choices for the future energy system (cf. Recommendation 4), and achieving partial and provisional closure on desirable measures followed by continuous monitoring and evaluation (cf. Recommendation 5). Performance of these tasks will crucially depend on having adequate procedural mechanisms in place for compensating the distorting effects of self-interest and power (cf. Recommendation 6). Here, we already point out the absolute need for independent and peer-reviewed expertise (e.g. the ‘energy agency’ must be equipped with a sufficiently staffed scientific secretariat) and the provision of sufficient resources (e.g. in terms of time and money) for carrying out the tasks mentioned. The ‘energy agency’ must also take up its function as a nexus between international commitments (e.g. European directives) and national performance. In particular, it has to take into account the ‘boundary conditions’ implied by the liberalisation of energy markets in the European context. Whilst the creation of an ‘energy agency’ by the federal government certainly constitutes a priority, it is questionable whether by itself it will have the power and authority to innovate on the scale required. A wider range of initiatives, in a wider

478

Chapter 8

range of interactions between energy technology and different user constituencies (e.g. energy companies, civil society groups, consumers, etc.) is needed as a matter of urgency (cf. Recommendation 7).

Recommendation 2: ‘Entry right’ to the deliberations taking place within the energy agency must not be based solely on established routines, disciplinary problem framings, interest positions or opinions. Boundaries with the ‘outside world’ must be kept flexible and open. In particular, we recommend inscribing a formal commitment to the precautionary principle in the ‘energy agency’s’ mandate as a necessary (but not sufficient) boundary condition governing its functioning.

This second recommendation can be read as a more practical consequence of the philosophical anti-foundationalist stance developed in chapter 1 of the present dissertation. It is of course easier to posit this recommendation than it is to realise it in practice. In this regard, it is no coincidence that arguably the largest part of actual policy-making initiatives takes place in settings that allow some measure of predictability for the outcomes of the debate – be it that the possible options are already well-defined, or that decision making takes place within a group with a homogenous commitment, or that decision mandates are strictly limited, etc (cf. our discussion of the National Energy Committee’s functioning in Chapter 2). Furthermore, the history of the nuclear debate in Belgium (and elsewhere) has led to more or less polarised positions and distrustful relationships between the principal actors involved. Therefore, strong procedural guarantees for ‘entry right’ for new ‘actants’ (in Latourian terms) should be put in place in order to avoid a reproduction of the exclusive dynamics witnessed in the past. Ad a)

Ad b)

Ad c)

Arguably the best chance for opening up deliberations to new perspectives lies in harnessing the potential for institutional innovation contained in the precautionary principle. An understanding of the precautionary principle as a rationale for action (during the whole process of selecting appropriate goals, working out feasible alternatives and developing adequate policy measures) is gradually developing in a number of policy domains (most notably those related to the environment and human health), both at the international and (in particular) the EU level, both on a political and juridical level. This embedment in established legal doctrine also implies that, as a legal principle, precaution cannot forego the requirements of reasoned and consistent rule making (i.e. taking into account principles such as proportionality, non-discrimination, and considering the scientific ‘state-of-the-art’) thus rendering it less vulnerable to particularistic political aspirations. Adopting a precautionary approach (i.e. an anti-foundationalist ‘doubting’ gesture) is however not sufficient. This approach must be explicitly linked to a sustainable development dynamic (i.e. an affirmative gesture). More specifically, the


Ad d)

479

sustainability dimensions of (international) solidarity (e.g. concerning issues of unjust distributions of risks and/or benefits in the international community) and institutional management (e.g. distribution of responsibilities for preventive measures, liability regimes, etc.) of proposed technological development paths should not be overlooked. This implies the establishment of a link with sustainable development targets (expressed through the development of criteria and indicators – cf. Recommendation 3 and 6), visions (in plural) for a sustainable energy future (cf. Recommendation 4), and the ‘matters of concerns’ expressed in an ‘enlarged public sphere’ (made up of socio-material practices – cf. Recommendation 7). Collective learning of course implies the possibility of revision – therefore, neither the list of ‘actants’ allowed in the debate nor the rules of interaction should be fixed permanently. Moreover, new ‘actants’ should be sought after consciously. Therefore, adequate institutional provisions must be put in place for keeping deliberations open to possible ‘surprises’, e.g. by organising an early warning function in public services, ensuring the independence of regulatory bodies from special political or economic interests, supporting independent academic research, or through dedicated long-term monitoring of specific impacts (cf. Chapter 5).

Recommendation 3: Deliberations taking place in the ‘energy agency’ can be facilitated by the creation of specific ‘boundary-spanning’ objects (e.g. value trees, sustainable energy criteria and indicators), or techniques (e.g. multi-criteria mapping).

In this dissertation we have strongly advocated the use of a sustainable energy indicator set, (organised into a value tree made up of sustainability criteria) as a crucial enabling element (or ‘boundary-spanning object’ with accountability to the multiple ‘social worlds’) for facilitating the ‘collective learning experience’ we envisage with our model process, in the sense that it allows the development of a protocol for revealing whether improvement towards more sustainability is actually achieved. In practice however, the development of such indicator sets is all too often predicated on the transposition of existing approaches and/or visions. Such practices are not responsive enough to context-dependent dynamics of the sustainable energy debate. This finding calls for a different approach. Ad a)

The list of sustainable energy criteria and indicators must create an opening towards other knowledge perspectives than the traditionally dominant technical/economic perspective. In this respect, it is for instance striking that the ‘civic’ argument, though playing an important role in many stakeholder perspectives, is largely absent from the ‘official’ political debate on sustainable energy. This finding points out the necessity of developing adequate institutional criteria for sustainable energy development (in particular, a more consistent and scientifically-sound articulation of characteristics of centralised/decentralised energy systems is needed). This will require the incorporation of new forms of

480

Ad b)

Ad c)

Chapter 8

social and humanities expertise, to complement those of existing technical/economic specialists. Therefore, the list of criteria to be taken into account must be developed by combining both top-down (i.e. drawing upon established – mostly technical/economic indicator sets) and bottom-up (i.e. responsive to the dynamics of the societal debate) approaches in a reflexive equilibrium: relevant criteria and/or indicators can be drawn from interviews with some of the key stakeholders in the debate, from the policy measures already in force or planned (e.g. policy declarations), from the scientific literature (e.g. ‘traditional’ risk analysis), and from sustainable development indicator sets abroad (e.g. developed by the IEA, EC, UNCSD, etc.). In addition, the ‘energy agency’ should instigate a search for criteria and indicators which are sensitive enough to detect responses to innovations or proposed policy measures at an early stage. The list of criteria and indicators to be taken into account must be determined by thinking back and forth between problem scanning and strategic vision assessment (cf. Recommendation 4), with ultimate policy goals being set only at a later stage of the decision process. A technique such as the multi-criteria mapping approach applied in this dissertation (cf. Chapter 7) has shown its usefulness for this purpose.

Recommendation 4: Strategic priority-setting for constructing the ‘sustainable energy commonwealth’ must rely on a ‘broad’ cost-benefit analysis. Such analysis must be informed by scientific foresight, provided that current foresight practices are modified in order to incorporate a wider range of normatively-inspired visions and insights in the dynamics of technological change.

A key element in our reasoning has been that strategic orientations for sustainability in the energy sector should be based on a ‘broad’ cost-benefit analysis (as opposed to more ‘narrow’ forms of cost-benefit analysis based on a monetarisation of all possible impacts) informed by socio-technical foresight exercises, i.e. ‘scenarios’. In particular, we have argued that the major advantage offered by such ‘broad’ cost-benefit analysis lies in the fact that it can bypass the difficulties engendered by a reference to particular scenarios as representing some form of objectivity (e.g. based on trend extrapolations, econometric modelling, etc.) and instead openly discuss the normative framings embedded in these scenarios. Scenarios thus constitute ‘boundary objects’ in the debate in more than one sense: they are attempts to span (but not erase!) divisions between science and policy (offering ‘scientifically validated’ orientations for policy making); science and ethics (showing that certain ethical requirements can or cannot be realised in practice; and conversely, by providing ethical orientation to research choices); and the past, present and the future (linking the future to the past and present).


Ad a)

Ad b)

Ad c)

Ad d)

481

It is therefore of vital importance that government should adapt existing foresight approaches to reflect broader normative needs in evaluating the wider impacts of energy technologies on the shaping of society, as part of the process of prior evaluation and R & D. In particular, our interactions with stakeholders have also revealed that sociotechnical foresight should also learn from history, i.e. through a thorough analysis of previous action – considering how decisions have played out in the past, and looking for key variables that might represent themselves in the future (cf. Section 3 – Suggestion 1 for further research). As put forward in actor-network theory, a new technological system will only succeed when it is able to attract a whole universe: a network of socio-material relationships has to be put together, persuaded and enlisted; and having a coherent and/or convincing vision or scenario at hand is often of vital importance in such processes. Discounting scenarios on the basis that they would lack reality or involve value judgments thus would constrain the perspectives encountered in decision making and also condition which knowledge-producing practices will be favoured. Therefore, encouraging the development of contrasting scenarios – e.g. by providing resources for stakeholders in the debate to develop their own visions – becomes a matter not only of guaranteeing ‘political pluralism’ but also ‘epistemological pluralism’. However, such scenarios should still provide an opportunity for mutual criticism and learning. In this regard, we have demonstrated the practical use of Boltanski and Thévenot’s commonwealth model as a common framework for playing out differences (on an epistemological, ethical, empirical, etc.) level. Strategic priorities for sustainable energy policy could for instance be developed by working out the consequences of each commonwealth logic (or at least the most relevant ones) in a speculative manner (whilst staying aware of course that the ‘real world’ can never correspond to such pure ideal types), thus revealing areas where consensus might be achievable, as well as areas of deep-seated conflict.

Recommendation 5: Negotiations towards the ‘sustainable energy commonwealth’ need to be built on the active cultivation of conditions from which trusted interactions can emerge. This recommendation does not entail a vision of political negotiations resting on trust (trust as a goal), but rather proposes a complementary analysis of policy problems from the point of view of the quality of interactions leading to their formulation, prioritisation and diagnosis.

The ‘energy agency’ should be encouraged to define itself as a collectivity whose members are committed to crafting solutions to shared problems. The development of trust and understanding is essential to the whole deliberative enterprise and can be seen at the same

482

Chapter 8

time as a prerequisite for successful deliberations, an emergent quality and an outcome that is desirable in its own right. However, trust in a process is a very complex variable that depends on the interplay of numerous other factors. Our reconstruction of the policy development cycle in the case of the Belgian nuclear phase out clearly shows that if a social actor cannot find any form of ‘resonance’, the probability for the investment of trust in a decision-making body will be very low. In such cases, information will be used as a weapon. The creation of trust therefore crucially depends on a number of preconditions. Ad a)

Ad b)

Ad c)

Ad d)

This demanding recommendation can only be realised with the aid of strong procedural guidelines assuring the overall ‘quality’ of the interactions, among them for instance the guarantee that scientific data serving as an input for the debate must be the result of a thorough review process. Procedural guidelines must also ensure transparency and accountability to larger ‘outside’ audiences, and provide guarantees for a ‘fair and competent’ process. Besides the transparency of the ‘products’ of a deliberation – in terms of the traceability of the results – it is also essential that the processes used for consultation are as transparent and legitimate as possible to all participants involved. This requires clear and accessible information concerning the structure of the process, its aims, its remit, the identity and interests of organisers and the sources of funding, and the expected roles of the different participants. These should be formally enshrined in an ‘agreement of understanding’ between the process participants. Practically speaking (in view of securing enough political ‘weight’), organised stakeholder groups will have to figure prominently in the new ‘energy agency’. In chapter 5 we have argued that both the criteria of openness towards other perspectives and productivity of small group interactions can be reconciled by ensuring that different ideal-typical roles – to wit scientific experts, ethicists, stakeholder representatives, administrators, facilitators and politicians – each contribute in a meaningful way to each of the tasks set out for the ‘energy agency’ (i.e. managing flexibility and openness to the ‘outside world’, problem structuring, broad cost-benefit analysis, etc.). These roles are not fully determined by professional or institutional attachments. Therefore, at each stage of the deliberations, the rules of interaction must specify and uphold the principle of adequate role representation (a task set out for the ‘facilitator’), and the attribution of roles should be clear to each actor involved. For trust building it is crucial that negotiations start in an early phase, so that actors have the possibility of real (instead of marginal) influence on (political) decision making. In this regard, framing deliberations under the umbrella of ‘sustainable development’ has the potential advantage of ‘reframing’ policy problems in new and unaccustomed terms. Increased attention at the earliest development stages to the possible congruence of new energy technologies and products with wider sustainability requirements will also be increasingly important for the execution of the sustainable energy development policies and precautionary approaches to which the Belgian government is already committed.


Ad e)

483

Our findings also clearly reveal that in the controversial nuclear energy debate uncertainties (e.g. in data sets, results of modelling exercises) are used as strategic resources in ‘warlike’ interactions (e.g. undermining arguments of an adversary, hiding uncertainties in one’s own assessments, etc.). Therefore, deliberations in the ‘energy agency’ should be supported by ‘uncertainty management’ techniques informing the participants (and outside audiences) on the ‘robustness’ of particular pieces of evidence used.

Recommendation 6: Political decisions are required for defining standards for measuring progress towards sustainability in the energy sector. Such standards should be ‘transformable’ in light of new evidence. Policy measures aimed at reaching these standards must display a degree of reversibility, flexibility and diversity.

Implicating different actors and unusual perspectives in the debate by no means implies that we are proposing to overturn the entire functioning of the political system. In particular, it does not relieve political representatives of their responsibility to take decisions, since they are the only legitimate actors for provisionally ‘closing’ the debate. Furthermore, consensus on possible strategies for sustainable energy development is no absolute requirement for the functioning of our governance model; rather the governance model should first and foremost provide a platform for playing out the differences between perspectives (before seeking out issues where a consensus could be developed). Thus, the outcomes of the collective learning process (canalised through the functioning of a ‘central’ energy agency) do not supplant the politicians’ power and duty to take decisions, but they do drive back the power to avoid taking a decision or to avoid taking into account the voices of all relevant ‘actants’. Ad a)

Adopting an anti-foundationalist view on sustainability (as we propose to do) implies recognition that ‘sustainability standards’ (i.e. long-term goals) – though necessary for orienting choices – should best be formulated so that they can be transformed in light of new insights. This can be achieved by leaving some room for interpreting the standards in different ways. The ongoing learning process must be organised to incorporate opportunities to dispute the content of strategic decisions and draw new ‘actants’ into the discussions. Ad b) Furthermore, in order to support a collective learning experience, policy measures should also uphold certain qualities, most notably reversibility (i.e. avoiding lockin effects), flexibility (i.e. the ability to enlist a policy measure for different purposes) and diversity (i.e. adopting a number of disparate options in parallel). Ad c) Policy measures must also be subject to a periodic review. This review should concern the intended results of the policy measures themselves, as well as an evaluation of changes in the knowledge base and the rules of interaction.

484

Chapter 8

Recommendation 7: Collective learning can be facilitated by considering the following points: the flow of information from the ‘central energy agency’ to the ‘periphery’; the establishment of robust but flexible connections with other ‘collective experiments’; and securing allegiance to all stages of the policy development cycle.

Rather than conceiving of a strategy for sustainable development in deterministic terms (i.e. political planning will have a series of predictable effects), constructivism takes an anti-deterministic stance based on both normative and empirical arguments. In a normative sense, it is necessary to keep the boundaries of the ‘collective’ open and flexible. In an empirical sense, it is clear that de facto, many actors are involved in technological developments (starting from basic research to the ultimate embedment of technology in society), and so there are no institutions capable of exerting direct control over all possible interactions. In political science, the image of the ‘policy network’, with governments implied as ‘modulators’ rather than top-down authoritative ‘directors’, has almost become commonplace. Furthermore, many processes of technology adaptation take place without a direct relation with government initiatives or concerns. And conversely, attempts at ‘modulation’ are not always limited to government interventions. The point is that, although this multi-actor process always takes place, it is not necessarily reflexive in nature – i.e. possibilities for adaptive learning are not always grasped. In this dissertation, we have argued that government’s main role should be to take up responsibility for inducing such ‘reflexivity’ by establishing connections and information flows with and between a diverse array of ‘entities’ that are deemed to be relevant for the purposes of sustainable energy governance. In addition, through these connections these entities should get at least an insight, and possibly also some influence, in the process taking place within the walls of the ‘energy agency’ (e.g. why are certain strategic choices proposed?). Ad a)

Ad b)

This role of establishing connections should not be seen as a replacement for the more ‘traditional’ forms of government intervention – i.e. direct regulation and intervention through market mechanisms – but rather as an addition, in particular in policy domains where these ‘traditional’ mechanisms have proved to be unsuccessful (e.g. local dialogue experiments on low-level waste management replacing the ‘traditional’ decide-announce-defend model). Prospective mapping and foresight (emphasised in our ‘pilot exercise’) are important, but essentially ‘passive’ activities (i.e. without any necessary implication for action). They can be useful for identifying promising options (embodying certain ‘visions’ of sustainability), but articulation of user ‘preferences’ will only occur in an actual confrontation with the new technology in real-world conditions. Therefore, government should initiate programmes of experimental testing of the potential of new deliberative mechanisms for engaging


Ad c)

Ad d)

485

with ‘the public’, which are now being proposed mainly in academic circles or on the level of local government (cf. Section 3 – Suggestion 2 for further research). Although the implications of this final recommendation admittedly are not fully developed in this dissertation, we did discuss some examples of such ‘collective experiments’. For instance, strategic niche management builds on the idea that ‘forceful’ demonstrations of technology (e.g. pilot experiments) are important in their own right in inducing debate and learning. We have also pointed out the essential role of businesses in carrying forward any long-term goals identified at the government level. Such institutional innovation in interactions between government and the public sphere would not have to start from scratch. By now, a large and growing body of ‘inclusive’ or ‘participatory’ approaches to technology policy (in a broad sense) have been tried out. However, seen from our constructivist perspective, much more careful attention should be given to both the social and material dimensions of these practices than displayed at present by a certain propensity in literature on the subject (influenced by Habermas’s discourse ethics) to conceive of participatory procedures as a ‘bolt on’ input to ‘traditional’ decision making.

3 Suggestions for further research With this brief summary of the findings and recommendations for policy makers we gradually arrive at the boundaries of our research. We have tried to arrive at a tenable philosophical position for studying sustainability debates, and have studied combined scientific and political practices for ‘closing’ such debates in the context of (nuclear) energy policy in Belgium. From this research, we have tried to derive a practical proposal for an ‘improved’ governance model adapted to the complex (i.e. combined scientific/ethical/political/etc.) questions raised by demands for a more sustainable energy system. Of course, in view of the quite formidable range of subjects covered by the sustainability discourse, we have to acknowledge that our work has been only partial and many gaps remain. In the following paragraphs, we would therefore like to suggest what we believe to be some areas where further research might be warranted. They relate to the past, the present and the future of the sustainability debate. 1. At first sight, it could seem strange that we invoke the past (even the past before ‘sustainable development’ was officially coined as a political leitmotiv) when discussing what seems to be an exclusively future-oriented concept. However, the issue is here that our institutions in society have a de facto historic in-built future perspective (as a necessity for operating meaningfully); therefore, the challenge becomes here to connect these ‘inbuilt’ future orientations with ‘sustainability’ as a collectively-defined future goal. Our analysis has clearly revealed that this sense of ‘historical injustice’ certainly plays a role in the present debate. On the level of national politics, and applied to the issue of nuclear power, an oft-mentioned problematic feature of the current situation is that Belgian government (and its taxpayers) are forced into a policy of compensation for earlier

486

Chapter 8

decisions. In a number of nuclear files, the (strong) conviction is that non-negligible profits (for the former dominant electricity generator) have been privatised, while (economic) risks and losses have been shifted to taxpayers or captive customers. Funding the ultimate and safe disposal of radioactive waste (including decommissioning of the existing power plants), remediation of the ‘nuclear passive’, state-funded nuclear research and development, etc. all figure as examples often cited by certain stakeholders in the debate. Therefore, we believe that a first necessary move towards a more productive dialogue between different stakeholders would firstly entail a complete (as far as possible) analysis of the externalities of the nuclear fuel cycle in a historic perspective, and comparing these to the externalities of other electricity-production technologies. Secondly, an important institutional research question would be to study which governance mechanisms could be envisaged for avoiding (as far as possible) the materialisation of new externalities in the future. 2. A second research track we would like to offer for consideration is an investigation of the possible ‘interfaces’ between the ‘enlarged public sphere’ (made up of socio-material practices) and the sphere of political decision-making. In this dissertation we have argued that, although the unruliness of interactions in this enlarged public sphere is to some extent unavoidable and even desirable, there are nevertheless opportunities for more carefully planned ‘collective experiments’ aimed at providing more specific guidance for the implementation of ‘sustainable’ policy measures. We believe that, in analogy to the more traditional notion of an experiment in exact-scientific terms, planning ‘collective experiments’ should at least obey the following guidelines: • The ‘theory’ behind the experiment should be plausible – i.e. evidence and common sense suggest that the specified ‘experimental’ activities will lead to the desired outcomes (therefore, these outcomes should also be clearly articulated); • The required collective experimentation is doable – i.e. the initiative has adequate financial, technical, political, institutional and human resources to implement the research strategy; and • Results of collective experimentation are testable – i.e. the pathways of change are specific and complete enough, with measurable indicators and specified pre-conditions, to track the progress in a credible and useful way. As mentioned before, a great deal of ‘experimentation’ at the interface between the ‘public sphere’ and ‘politics’ has already been tried out (e.g. various participatory technology assessment tools, environmental conflict mediation, ethical codes for companies, corporate social responsibility schemes, etc.), although a broadly shared theoretical frame for understanding these approaches is still lacking. Thus, further work would first have to systematically collect and analyse such collective experiences, and explore the conditions under which governance arrangements based on a logic of inclusion might likely gain a foothold. Also, an important research venture would be to explore the conditions under which such participatory experiments might assume exemplary value and make a real impact on the collective construction of the future energy system.


487

3. A final research suggestion relates to experimenting with socio-technical foresight exercises. As argued, socio-technical foresight exercises can be seen as prime examples of a ‘hybrid fora’ where ethics, science, economic interests, and politics (and much more) meet head-on. As a correlate, such exercises have to serve many different purposes at once. This makes the creation of foresight scenarios as a platform for dialogue an inherently complex undertaking. Without pretending to be complete, different issues have to be taken into account and articulated simultaneously, such as: • The role and limits of scientific expertise, especially in long-term decision-making about energy; • The internal structure and functioning of the models used needs to be adapted to the purpose of the foresight exercise; • The philosophy behind scenario construction (e.g. backcasting vs. forecasting) needs to be clearly understood and accepted by every actor involved; • The choice and the role of stakeholder groups in providing insights for foresight exercises must be decided upon in function of assuring a pluralistic perspective; • Quality assurance of the process and the results needs to be defined; • The resources (time, people, money) available to do the modelling/scenario work will have a significant impact on the overall quality of the results; • The communication of the results of foresight exercises (to decision makers, stakeholders, etc.) must the subject of careful planning; • The role and importance of the wider political context must be taken into account; and • Integration of different areas of scientific expertise must be organised. Furthermore, it is important to address these issues from the beginning, i.e. in setting-up the process (e.g. stakeholders will not simply accept results coming out of ‘a black box’model). This in turn invites reflection on what has been common practice so far in the field of socio-technical foresight, and to derive from this ‘benchmarking’ exercise rules of good practice.

4 Epilogue Hidden behind an apparently simple plea in favour of sustainable development lies a daunting complexity of (possibly conflicting) goals, expectations, experiences, conceptions of the self and society, and so on. This is certainly true in the field of complex technological questions (such as those relating to energy choices for the future). In this dissertation we have argued in favour of a constructivist perspective on sustainability. We have reached into the domain of governance, but it was only a reaching, in the sense that we were looking for those conditions that might allow ‘collective experimentation’ to emerge, without however prescribing the direction this experimentation should take. There are no easy (ethical, economic, political, technological, etc.) answers to the inherently ‘complex’ question of sustainability. However, we have argued that, rather than being depressed by this state of affairs, one should feel invigorated by the possibilities lying before us. It is too early to say where these will take us, but there are strong indications

488

Chapter 8

that something is stirring, and that it will be difficult to revert to the old expertise or politics ‘as usual’…

REFERENCES & BIBLIOGRAPHY Adam, B., Beck, U. and Van Loon, J. (Eds.) (2000), The Risk Society and Beyond. Critical Issues for Social Theory, Sage, London. Afgan, N., Carvalho, M. and Hovanov, N. (2000), “Energy system assessment with sustainability indicators”, Energy Policy, Vol. 28, pp. 603-612. Aibar, E. (1996), “The evaluative relevance of social studies of technology”, Techné: Journal of the Society for Philosophy and Technology, Vol. 1, No. 3-4. Akrich, M. (1995), “User representations: Practices, methods and sociology”, in Rip, A., Misa, Th. and Schot, J. (Eds.), pp. 167-184. AMPERE Commission (2000), Rapport van de commissie voor de Analyse van de Productiemiddelen van Elektriciteit en de Re-oriëntatie van de Energievectoren (AMPERE), Besluiten en Aanbevelingen, Executive Summary, Syntheserapport en Hoofdrapport, available at . Anderson, K. (2001), “Reconciling the electricity industry with sustainable development: Backcasting – a strategic alternative”, Futures, Vol. 33, pp. 607-623. Ayres, R., van den Bergh, J. and Gowdy, J. (2001), “Strong versus weak sustainability: economics, natural sciences and ‘consilience’”, Environmental Ethics, Vol. 23, pp. 155-168. Badiou, A. (2005), De ethiek. Essay over het besef van het Kwaad (Nederlandse vertaling), Uitgeverij Ijzer, Utrecht. Barbé, L. (2005), Kernenergie in de wetstraat – Dissectie van de deals, available at . Barber, B. (1984), Strong Democracy. Participatory Politics for a New Age, University of California Press, Los Angeles. Barnes, B. (1977), Interests and the Growth of Knowledge, Routledge, London. Barrett, Ch. And Grizzle, R. (1999), “A holistic approach to sustainability based on pluralism stewardship”, Environmental Ethics, Vol. 21, pp. 23-42. Bataille, Ch. and Birraux, C. (2003), Rapport sur la durée de vie des centrales nucléaires et les nouveaux types de réacteurs, Office Parlementaire d'Evaluation des Choix Scientifiques et Technologiques, Paris. Bateson, G. (1979), Mind and Nature – A Necessary Unity, Wildwood House, London. Beck, U. (1992), Risk Society. Towards a New Modernity, Sage, London. Beck, U. (1995), Ecological Politics in an Age of Risk, Polity Press, Cambridge. Beck, U. (1997) (selected and edited by M. Hajer and M. Schwartz), De wereld als risicomaatschappij. Essays over de ecologische crisis en de politiek van de vooruitgang, de Balie, Amsterdam. Beck, U. (1999), World Risk Society, Polity Press, Cambridge. Beck, U., Giddens, A. and Lash, S. (Ed.) (1994), Reflexive Modernisation: Politics, Tradition and Aesthetics in Modern Social Order, Stanford University Press, Stanford. Bergmans, A. (2005), Van ‘de burger als beleidssubject’ naar ‘de burger als partner’: de Belgische queeste naar een lange-termijnoplossing voor het beheer van het laagradioactief en kortlevend afval, unpublished PhD thesis, Faculteit Politieke en Sociale Wetenschappen, Universiteit Antwerpen (UA), Antwerpen. Berloznik, R., Vancolen, D., Van Rensbergen, J., Valenduc, G., Vendramin, P. and Marion, J.-Y. (1996), Wetenschappelijk onderzoek en duurzame ontwikkeling, Eindverslag aan het DWTC, Vlaamse Instelling voor Technologisch Onderzoek (VITO), Mol. Beroepsfederatie der Elektriciteitsproducenten (BFE) (2003), Jaarverslag 2003, BFE, Linkebeek. Bérubé, G. and Villeneuve, F. (2002), “Ethical dilemmas and the decision-making process. Is a consensus realistic?”, Energy Policy, Vol. 30, pp. 1285-1290. Bijker, W. (1987), “The social construction of bakelite: Toward a theory of invention”, in Bijker, W., Hughes, Th. and Pinch, T. (Eds.), pp. 159-187. Bijker, W. (1992), “The social construction of fluorescent lighting, or how an artifact was invented in its diffusion stage”, in Bijker, W. and Law, J. (Eds.), Shaping Technology/Building Society: Studies in Sociotechnical Change, MIT Press, Cambridge, pp. 75-102. Bijker, W. (1995), Of Bycicles, Bakelite and Bulbs: Toward a Theory of Sociotechnical Change, MIT Press, Cambridge. Bijker, W. (2001), “Social construction of technology”, in Smelser, N. and Baltes, P. (Eds.), International Encyclopedia of the Social & Behavioral Sciences, Vol. 23, Elsevier Science Ltd., Oxford/Amsterdam/ etc., pp. 15522-15527. Bijker, W., Hughes, Th. and Pinch, T. (Eds.) (1987), The Social Construction of Technological Systems: New Directions in the Sociology and History of Technology, MIT Press, Cambridge. Bijker, W. and Pinch, T. (2002), “SCOT answers: other questions ! A response to Nick Clayton”, Technology and Culture, Vol. 43, pp. 361-369.

490


Blass, E. (2003), “Researching the future: Method or madness?”, Futures, Vol. 35, pp. 1041-1054. Bloor, D. (1976), Knowledge and Social Imagery, Routledge/Kegan Paul, London. Bobbio, N. (1987), The Future of Democracy, University of Minnesota Press, Minneapolis. Boehmer-Christiansen (1995), “Reflections on scientific advice and EC transboundary pollution policy”, Science and Public Policy, Vol. 22, No. 3, pp. 195-203. Boholm, A. (1996), “The many faces of ‘The Cultural Theory of Risk’: An anthropological critique”, paper presented at the 1996 Annual Meeting of the Society for Risk Analysis (SRA) – Europe. Boltanski, L. (1982), Les cadres. La formation d’un groupe social, Minuit, Paris. Boltanski, L., Darré, Y. and Schiltz, M. (1984), “La dénonciation”, Actes de la recherche en sciences sociales, No. 51, pp. 3-40. Boltanski, L. and Thévenot, L. (Eds.) (1989), Justesse et justice dans le travail, CEE – PUF, Paris. Boltanski, L. and Thévenot, L. (1991), De la justification. Les économies de grandeur, Gallimard, Paris. Bombaerts, G. (2004), Waste depositionism. A philosophical inquiry on technoscientists and nuclear waste, unpublished PhD thesis, Faculteit Letteren en Wijsbegeerte (Vakgroep Wijsbegeerte en Moraalwetenschap), Universiteit Gent (UGent), Gent. Boonekamp, P. (2002), Energy and emission indicators. International inventory and assessment, Report ECNC-02-072, Energie Centrum Nederland (ECN), Amsterdam. Börlin, M. (1994), “Technology assessment and the interests of future generations”, in Agius, E. and Busuttil, S. (Eds.), What Future for Future Generations?, The Foundation for International Studies, University of Malta, Malta, pp. 267-290. Bossel, H. (1999), Indicators for Sustainable Development: Theory, Method, Applications, International Institute for Sustainable Development, Winnipeg. Boudourides, M. (2003), “Participation under uncertainty”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 39-69. Bovy, M., Laes, E. and Eggermont, G. (2003), “Science, neutrality and objectivity in nuclear research: Shared references or ideal types?”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 433-443. Boulanger, P.-M., Thomas, P.-Y., van Assche, J. and De Ridder, B. (2003), Het participatief opmaken van indicatoren voor duurzame ontwikkeling in België, Institut pour le Développement Durable (IDD)/ Centrum voor Duurzame Ontwikkeling (CDO), Ottignies/Gent. Bourdeau, P., Laponche, B., Morrison, R., Mortensen, J. and Savelli, P. (2001), Assessment of the AMPERE Commission report by an international peer review group, available at Brey, Ph. (1997), “Social constructivism for philosophers of technology: A shopper’s guide”, Techné: Journal of the Society for Philosophy & Technology, Vol. 2, No. 3-4, pp. 56-80. Bruggink, J. and van der Zwaan, B. (2002), “The role of nuclear energy in establishing sustainable energy paths”, International Journal of Global Energy Issues, Vol. 18, Nos. 2/3/4. Bruyer, V., Wallenborn, G., Bachus, K., De Bruyn, T. and Gysen, J. (2002), Draagvlakenquête duurzame ontwikkeling – Tweede editie, Eindrapport in opdracht van de Federale Raad Duurzame Ontwikkeling (FRDO), Institut de Gestion de l’Environnement et d’Aménagement du Territoire (IGEAT, ULB)/Hoger Instituut Voor de Arbeid (HIVA, KULeuven), Brussel/Leuven. Bruyninckx, H. (2002), Towards a social pact in sustainability matters: Concluding research remarks on participation in the Belgian sustainable development policy context, Executive summary of a research project sponsored by the Federal Science Policy Department (HL/DD/01), Brussels. Buchanan, A. (1996), “Choosing who will be disabled: Genetic intervention and the morality of inclusion”, Social Philosophy and Policy, Vol. 13, No. 2, pp. 18-46. Buchdahl, J. and Raper, D. (1998), “Environmental ethics and sustainable development”, Sustainable Development, No. 6, pp. 92-98. Burgess, J., Collins, K., Harrison, C., Munton, R. and Murlis, J. (1999), An Analytical and Descriptive Model of Sustainable Development, The Environment & Society Research Unit (University College London), London. Burms, A. and De Dijn, H. (1990), De rationaliteit en haar grenzen – Kritiek en deconstructie, Universitaire Pers Leuven/Van Gorcum, Leuven/Assen/Maastricht. Burms, A. and De Dijn, H. (1993), “Tijdsdiagnose: Overspannen daadkracht”, in van Boxsel, J., Gruppelaar, J., Achterhuis, H. and Vorstenbosch, J. (Eds.), Techniek, voorlopigheid en verlegenheid. Ethiek en technologisch aspectenonderzoek, Nederlandse Organisatie voor Technologisch Aspectenonderzoek (NOTA), Den Haag, pp. 63-73. Button, M. and Mattson, K. (1999), “Deliberative democracy in practice: Challenges and prospects for civic deliberation”, Polity, Vol. XXXI, No. 4, pp. 609-637. Callicott, J. (1994), “Towards a global environmental ethic”, in Brown, N. (Ed.), Ethics and Agenda 21: Moral Implications of the Global Consensus, United Nations Publications, New York.

References & Bibliography 491 Callon, M. (1987), “Society in the making: The study of technology as a tool for sociological analysis”, in Bijker, W., Hughes, Th. and Pinch, T. (Eds.), pp. 83-103. Callon, M. (1991), “Techno-economic networks and irreversibility”, in Law, J. (Ed.), A Sociology of Monsters: Essays on Power, Technology and Domination, Routledge, London/New York, pp. 132-161. Callon, M. (1995), “Technological conception and adoption network: Lessons for the CTA practitioner”, in Rip, A., Misa, Th. and Schot, J. (Eds.), pp. 307-327. Callon, M. (1998), “An essay on framing and overflowing : Economic externalities revisited by sociology”, in Callon, M. (Ed.), The Laws of the Markets, Blackwell Publishers, Oxford, pp. 244-269. Cambrosio, A. and Limoges, C. (1991), “Controversies as governing processes in technology assessment”, Technology Analysis & Strategic Management, Vol. 3, No. 4, pp. 377-396. Carlé, B. and Hardeman, F. (2002), Veiligheid en risicoperceptie. Resultaten van de opiniepeiling van november 2002 in België, Report BLG-938, Studiecentrum voor Kernenergie (SCK•CEN), Mol. Carter, A. (1999), A Radical Green Political Theory, Routledge, London. Centre for Economic Studies (CES) (2001), Markal, a model to support greenhouse gas reduction policies, Final Report nr. CG/DD/221 & CG/DD/222, Federal Office for Scientific, Technical and Cultural Affairs, Brussels. Centre for Economic Studies (CES) and Vlaams Instituut voor Technologisch Onderzoek (VITO) (2001), The Belgian MARKAL database, Final Report nr. CG/DD/221 & CG/DD/222, Federal Office for Scientific, Technical and Cultural Affairs, Brussels. Charpin, J.-M., Dessus, B. and Pellat, R. (2000), Etude économique prospective de la filière électrique nucléaire, rapport au Premier Ministre, Paris. Chawla, S. (1991), “Linguistic and philosophical roots of our environmental crisis”, Environmental Ethics, Vol. 13, pp. 253-262. Chermack, T. and van der Merwe, L. (2003), “The role of constructivist learning in scenario planning”, Futures, Vol. 35, pp. 445-460. Chiesura, A. and de Groot, R. (2003), “Critical natural capital: a socio-cultural perspective”, Ecological Economics, Vol. 44, pp. 219-231. Coglianese, C. (1998), “Implications of liberal neutrality for environmental policy”, Environmental Ethics, Vol. 20, pp. 41-60. Cohen, S., Demeritt, D., Robinson, J. and Rothman, D. (1998), “Climate change and sustainable development: Towards dialogue”, Global Environmental Change, Vol. 8, No. 4, pp. 341-371. Collingridge, D. (1980), The Social Control of Technology, St. Martin’s Press, New York. Colorni, A., Paruccini, M. and Roy, B. (Eds.) (2001), Multiple Criteria Decision Aiding, Office for Official Publications of the European Communities, Luxembourg. Commissie van Beraad inzake Kernenergie (1976), Eindrapport, Ministerie van Economische Zaken, Brussel. Commissie voor de Regulering van de Elektriciteit en het Gas (CREG) (2001), Studie over de provisies en de fondsen in de nucleaire sector, Rapport F010315-CDC-024, CREG, Brussel. Cooms, R. (1995), “Firm strategies and technological choices”, in Rip, A., Misa, Th. and Schot, J. (Eds.), pp. 331-345. Cooney, R. (2004), The precautionary principle in biodiversity conservation and natural resource management: An issue paper for policy makers, researchers and practitioners, IUCN, Gland/Cambridge. Courcelle, Ch. and Gusbin, D. (2001), Energievooruitzichten 2000-2020. Federaal Planbureau, Brussel. Cowell, F. (1995), Measuring inequality (second edition), Prentice Hall/Harvester Wheatsheaf, London. Crabbe, Ph. (1997), “Sustainable development: Concept, measures, market and policy failures at the open economy, industry and firm levels”, CANSEE 97 (CANadian Society for Ecological Economics) Proceedings, Mac Master University, Hamilton. Craye, M., Funtowicz, S. and van der Sluijs, J. (2004), “A reflexive approach to dealing with uncertainties in environmental health risk science and policy”, International Journal for Risk Assessment and Management, Vol.4, No. 2. pp 216-236. Dammers, E. (2000), Leren van de toekomst. Over de rol van scenario’s bij strategische beleidsvorming, Eburon, Delft. Damveld, H. and van den Berg, R. (1998), Kernafval en kernethiek: maatschappelijke en ethische aspecten van (de terughaalbare) opslag van kernafval, Metra, Groningen. Damveld, H. (2003), “Nuclear waste and ethics”, in Andersson, K. (Ed.), VALDOR – VALues in Decisions On Risk – Proceedings, June 9-13, Stockholm, pp. 128-139. Deblonde, M. (2001), Economics as a Political Muse. Philosophical Reflections on the Relevance of Economics for Ecological Policy, Kluwer Academic Publishers, Dordrecht/Boston/London. De Dijn, H. (2003), Taboes, monsters en loterijen. Over ethiek in de laatmoderne tijd, Uitgeverij Pelckmans, Kapellen.

492


Degros, Ch. (2004), La responsabilité civile et l’assurance de l’exploitant d’une installation nucléaire: Un défi renouvelé, Mémoire de Diplôme d’Etude Spécialisée en droit et économie des assurances (unpublished), Université Catholique de Louvain (UCL), Faculté de Droit, Louvain-La-Neuve. De Jonge, W., Mazijn, B. and Van Assche, J. (2000), “Milieugebruiksruimte: operationalisering van een vaag begrip”, in Mazijn, B. (Ed.), pp. 1-68. De Laet, M. (1995), “Technologie en samenleving in een antropologisch perspectief”, in Achterhuis, H., Geurts, J., Rip, A. and Smits, R. (Eds.), Technologie en samenleving, Garant, Apeldoorn/ Leuven. de Meere, F. (1996), U kunt gerust gaan slapen… Denkbeelden over technologie, risico’s en samenleving, Eburon, Delft. De Nocker, L., Torfs, R. and Wouters, G. (2000), “Externe kosten van elektriciteitsproductie”, E-tijdschrift, 116e Jaargang (speciaal nummer), pp. 85-93. Department for Transport, Local Government and the Regions (DTLR) (2001), Multi-Criteria Analysis: A Manual, DTLR, available at . de Sadeleer, N. (1999), Les principes du pollueur-payeur, de prévention et de précaution. Essai sur la génèse et la portée juridique de quelques principes de droit de l’environnement, Editions Bruylant, Bruxelles. de Vries, B. (1996), “Contouren van een duurzame toekomst”, in Gimeno, P., Weiler, R. and Holemans, D. (Eds.), pp. 31-72. D’haeseleer, W. (2003), “Future electricity generation in Belgium, the nuclear phase out and its greenhouse gas consequences”, Physicalia Magazine, No. 25, pp. 63-75. Dietz, F., Hafkamp, W. and van der Straaten, J. (1994), Basisboek milieu-economie, Boom, Amsterdam/ Meppel. Dobson, A. (1996), “Environmental sustainabilities: An analysis and a typology”, Environmental Politics, Vol. 6, No. 3, pp. 401-428. Dodier, N. (1993), “Review article: Action as a combination of ‘common worlds’”, Sociological Review, Vol. 41, No. 3, pp. 556-571. Dosse, F. (1995), L’empire du sens. L’humanisation des sciences sociales, La Découverte, Paris. Douglas, M. and Wildavsky, A. (1982), Risk and Culture: An Essay on the Selection of Technical and Environmental Dangers, University of California Press, Berkeley. Dratwa, J. (2002), “Taking risks with the precautionary principle. Food (and environment) for thought at the European Commission”, Journal of Environmental Policy and Planning, Vol. 4, pp. 197-213. Dryzek, J. (2000), Deliberative Democracy and Beyond: Liberals, Critics, Contestations, Oxford University Press, Oxford. Durbin, P. (1996), “Can there be a best ethic of sustainability?”, Techné: Journal of the Society for Philosophy and Technology, Vol. 2, pp. 49-60. Duyvendak, W., De Haan, I., Manschot, H., Mol, A. and Prins, B. (1992), “Camembert, rechtvaardigen en geweldloze liefde. Het politiek-filosofisch-economisch-sociologisch-antropologisch raster van Luc Boltanski en Laurent Thévenot”, Krisis, Juni, pp. 5-22. Eckersley, R. (2004), The Green State, MIT Press, Cambridge. Econotec and Vito (2005), Key assumptions for subsequent calculation of mid and long-term greenhouse gas emission scenarios in Belgium, Final report prepared in the framework of a study carried out for the Federal Public Service of Public Health, Food Chain Safety and the Environment, Econotec/Vito, Brussels/Mol. Eeckhoudt, L., Schieber, C. and Schneider, T. (2000), “Risk aversion and the external costs of a nuclear accident”, Journal of Environmental Management, Vol. 58, pp. 109-117. Ekins, P., Simon, S., Deutsch, L., Folke, C. and de Groot, R. (2003), “A framework for the practical application of the concepts of critical natural capital and strong sustainability”, Ecological Economics, Vol. 44, pp. 165-185. Elam, M. (1994), “Anti anti-constructivism, or laying the fears of Langdon Winner to rest”, Science, Technology and Human Values, Vol. 19, pp. 101-106. Elam, M. and Bertilsson, M. (2003), “Consuming, engaging and confronting science: The emerging dimensions of scientific citizenship”, European Journal of Social Theory, Vol. 6, No. 2, pp. 233-251. Elliott, D. (2003), Energy, Society and Environment, Routledge, London/New York. Elliot, R. (Ed.) (1995), Environmental Ethics, Oxford University Press, Oxford. Energie Centrum Nederland (ECN) (2000), Energietechnologie in het spanningsveld tussen klimaatbeleid en liberalisering, Rapport ECN-C-00-020, ECN, Petten. Epstein, S. (1994), “Integration of the cognitive and the psychodynamic unconscious”, American Psychologist, Vol. 49, pp. 709-724. Erreygers, G. (1984), Coal versus Nuclear. A Struggle for Power Evaluation and Comparison of Major SideEffects, Studiecentrum voor Economisch en Sociaal Onderzoek (SESO), Universitaire Faculteiten SintIgnatius (UFSIA), Antwerpen.

References & Bibliography 493 Espejo, R. (2003), “The RISCOM model – Boundaries and contextualisation”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 444-451. Espejo, R. and Stewart, N. (1998), “Systemic reflections on environmental sustainability”, Systems Research and Behavioural Science, Vol. 15, pp. 483-496. Eurobarometer (1993), Europeans and the Environment in 1992, Opinion survey carried out in the Eurobarometer framework, European Commission DG XI, Brussels. European Commission (EC) (1993), The Fifth Environment Action Plan – Towards Sustainability, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (1995a), ExternE: Externalities of Energy, Vol. 1-6, Project report EUR 1652016525, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (1995b), An Energy Policy for the European Union, COM (95) 682, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (1999a), ExternE: Externalities of Energy, Vol. 7-10, Project report EUR 1908319086, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (1999b), The TRUSTNET framework : A new perspective on risk governance, Project report EUR 19136 EN, European Commission, Brussels. European Commission (EC) (2000), Communication from the Commission on the Precautionary Principle, COM(2000) 1, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2001a), The Sixth Environment Action Programme – Environment 2010: Our Future, Our Choice, COM(2001) 31 final, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2001b), European Governance: A White Paper, COM(2001) 428, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2001c), Green Paper – Towards a European Strategy for the Security of Energy Supply, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2002), Energy: Issues, Options and Technologies, Eurobarometer Special Survey, Doc EUR 20624, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2003), External Costs. Research Results on Socio-Environmental Damages Due to Electricity and Transport, Project report EUR 20198, Office for the Official Publications of the European Communities, Luxembourg. European Commission (EC) (2005), ExternE: Externalities of Energy, Methodology 2005 Update, project Report EUR 21951, Office for the Official Publications of the European Communities, Luxembourg. European Environmental Agency (EEA) (2002), Energy and Environment in the European Union, EEA, Copenhagen. Eyckmans, J. and Pepermans, G. (2003), “Is er een toekomst voor kernenergie in België”, Samenleving en Politiek, Jaargang 10, No. 9, pp. 13-25. Eyre, N. (1997), “External costs – What do they mean for energy policy?”, Energy Policy, Vol. 25, No. 1, pp. 85-95. Federaal Planbureau (FPB) – Task Force Duurzame Ontwikkeling (1999), Op weg naar duurzame ontwikkeling? – Federaal rapport inzake duurzame ontwikkeling, Federaal Planbureau, Brussel. Federaal Planbureau (FPB) – Task Force Duurzame Ontwikkeling (2002), Een stap naar duurzame ontwikkeling? – Federaal rapport inzake duurzame ontwikkeling 2002, Federaal Planbureau, Brussel. Fedichem (2000), Geen duurzame ontwikkeling zonder chemie, Fedichem, Brussel. Feenberg, A. (1999), Questioning Technology, Routledge, New York. Feiveson, H. (2003), “Nuclear power, nuclear proliferation, and global warming”, Forum on Physics & Society, January issue, available at . Ferry, L. (1992), Le nouvel ordre écologique: l’arbre, l’animal et l’homme, Grasset et Fasquelle, Paris. Festenstein, M. (2004), “Deliberative democracy and two models of pragmatism”, European Journal of Social Theory, Vol. 7, No. 3, pp. 291-306. Fischedick, M., Nitsch, J. et al. (2002), Langfristszenarien für eine nachhaltige Energienutzung in Deutschland, Umweltsbundesambt, Berlin. Fischer, F. (1980), Politics, Values and Public Policy. The Problem of Methodology, Westview Press, Boulder. Fiske, A. and Tetlock, Ph. (1997), “Taboo trade-offs: Reactions to transactions that transgress the spheres of justice”, Political Psychology, Vol. 18, No. 2, pp. 255-297. Flueler, Th. and Scholtz, R. (2004), “Socio-Technical Knowledge for Robust Decision Making in Radioactive Waste Governance”, Risk Decision and Policy, Vol. 9, pp. 129-159. Forester, J. (1993), Critical Theory, Public Policy and Planning Practice, SUNY Press, Albany. Forester, J. (1999), The Deliberative Practitioner: Encouraging Participatory Planning Processes, MIT Press, Cambridge. Foster, J. (Ed.) (1997), Valuing Nature? Ethics, Economics and the Environment, Routledge, London/New York.

494


Foster, J. (1997), “Environment and creative value”, in Foster, J. (Ed.), pp. 232-246. Fraunhofer Institute for Systems and Innovation Research (2003), “Beheer van de energievraag” in het raam van de door België te leveren inspanningen om de uitstoot van broeikasgassen te verminderen, Ministerie van Economische Zaken, Brussel. Fujimura, J. (1996), Crafting Science: A Sociohistory of the Quest for the Genetics of Cancer, Harvard University Press, Cambridge. Fung, A. and Wright, E. (2003), “Thinking about empowered participatory governance”, in Fung, A. and Wright, E. (Eds.), Deepening Democracy: Institutional Innovations in Empowered Participative Governance, Verso, London, pp. 3-42. Funtowicz, S. and Ravetz, J. (1990), Uncertainty and Quality in Science for Policy, Kluwer, Dordrecht. Funtowicz, S. and Ravetz, J. (1993), “The emergence of post-normal science”, in von Schomberg, R. (Ed.), Science, Politics and Morality : Scientific Uncertainty and Decision Making, Kluwer Academic Publishers, Dordrecht. Gandy, M. (1999), “Rethinking the ecological leviathan: environmental regulation in an age of risk”, Global Environmental Change, No. 9, pp. 59-69. Garrod, G. and Willis, K. (1999), Economic Valuation and the Environment – Methods and Case Studies, Edward Elgar, Cheltenham. Georgescu-Roegen, N. (1971), The Entropy Law and the Economic Process, Harvard University Press, Cambridge. German Advisory Council on Global Change (WBGU) (2003), World in transition: Towards sustainable energy systems, Summary for policy makers, Berlin. Gibbons, M., Nowotny, H., Limoges, C., Schwartzman, S., Scott, P. and Trow, M. (1994), The New Production of Knowledge: The Dynamics of Science and Research in Contemporary Societies, Sage, London. Gibson, R. with Hassan, S. Holtz, S. Tansey, J. and Whitelaw, G. (2005), Sustainability Assessment. Criteria, Processes and Applications, Earthscan Publishers, London. Giddens, A. (1979), Central Problems in Social Theory. Action, Structure and Contradiction in Social Analysis, Macmillan, London. Giddens, A. (1990), The Consequences of Modernity, Polity Press, Cambridge. Giddens, A. (1991), Modernity and Self-Identity, Polity Press, Cambridge. Giere, R. (1993), “Science and technology studies: Prospects for an enlightened postmodern synthesis”, Science, Technology and Human Values, Vol. 18, pp. 102-112. Gieryn, Th. (1995), “Boundaries of science”, in Jasanoff, Sh., Markle, G., Petersen, J. and Pinch, T. (Eds.), Handbook of Science and Technology Studies, Sage, Thousand Oaks, pp. 393-443. Gimeno, P. (1998), Mensen onder andere. Een inleiding tot de milieuethiek, Acco, Leuven /Amersfoort. Gimeno, P., Weiler, R. and Holemans, D. (Eds.) (1996), Ontwikkeling en duurzaamheid – De uitdaging ontleed, Technologisch Instituut/VUBPress, Antwerpen/Brussel. Glasbergen, P. and Blowers, A. (1995), Environmental Policy in an International Context. Perspectives on Environmental Problems, Arnold, London. Glorieux, E. (Ed.) (2002), De Belgische uitstap uit de kernenergie – Deel I t/m III, Luc Lemiengre, Brussels. Godard, O. (1994), “Le développement durable, paysage intellectuel”, Nature Sciences Société, No. 2, pp. 309322. Godard, O. (Ed.) (1997), Le principe de précaution dans la conduite des affaires humaines, Maison des Sciences de l’Homme, Paris. Godard, O. (2003), “Développement durable et principes de légitimité”, Cahier No. 2003-003, Ecole Polytechnique, Paris. Godlovitch, S. (1998), “Things change : so whither sustainability ?”, Environmental Ethics, Vol. 20, pp. 291304. Goorden, L. (2003), “Draagvlak of voedingsbodem. Het hoe en waarom van publieksparticipatie”, paper presented at the FRDO Symposium (25 March), Brussel, available at . Gosseries, A. (2001), “What do we owe the next generation(s) ?”, Loyola of Los Angeles Law Review, Vol. 35, pp. 293-354. Greenpeace (2001), Ontmanteling van kerncentrales en beheer van kernafval bij de opening van de elektriciteitsmarkt. Bespreking van het bestaande juridisch-financieel vacuüm en de tekortkomingen van het voorstel-Deleuze van vandaag, Nota aan de ministerraad, Brussel. Grimston, M. and Beck, P. (2002), Double or Quits? The Global Future of Civil Nuclear Energy, Earthscan Publications, London. Grin, J., van de Graaf, H. and Hoppe, R. (1997), Interactieve technology assessment – Een eerste gids voor wie het wagen wil, Werkdocument 57, Rathenau Instituut, Den Haag. Grin, J. and Grunwald, A. (2000) (Eds.), Vision Assessment: Shaping Technology in 21st Century Society, Springer Verlag, Berlin/Heidelberg/New York.

References & Bibliography 495 Grint, K. and Woolgar, S. (1995), “On some failures of nerve in constructivist and feminist analyses of technology”, Science, Technology and Human Values, Vol. 20, pp. 286-310. Grunwald, A. (2000), “Technology policy between long-term planning requirements and short-ranged acceptance problems. New challenges for technology assessment”, in Grin, J. and Grunwald, A. (Eds.), pp. 99-148. Grunwald, A. (2004), “Strategic knowledge for sustainable development: The need for reflexivity and learning at the interface between science and society”, International Journal for Foresight and Innovation Policy, Vol. 1, Nos. 1/2, pp. 150-167. Guijt, I., Moiseev, A. and Prescott-Allen, R. (2001), IUCN Resource Kit for Sustainability Assessment, International Union for the Conservation of Nature (IUCN) Monitoring and Evaluation Initiative, Geneva. Gusbin, D. and Hoornaert, B. (2004), Energievooruitzichten voor België tegen 2030, Planning Paper 95, Federaal Planbureau, Brussel. Guston, D. (2001), “Boundary organisations in environmental policy and science: an introduction”, Science, Technology and Human Values, Vol. 26, pp. 399-408. Habermas, J. (1996), Between Facts and Norms. Contributions to a Discourse Theory of Law and Democracy, Polity Press, Cambridge. Hacking, I. (1999), The Social Construction of What?, Harvard University Press, Cambridge. Hajer, M. (1995), The Politics of Environmental Discourse. Ecological Modernisation and the Policy Process, Clarendon Press, Oxford. Haldi, P.-A., Frei, Ch., Beurskens, L. and Zhuikova, N. (2002), “Multicriteria/multi-stakeholders comparative assessment of electricity generation scenarios in the sustainability context: A Swiss case study”, International Journal of Sustainable Development, Vol. 5, Nos. 1-2, pp. 102-124. Hämäläinen, R. (1990), “A decision aid in the public debate on nuclear power”, European Journal of Operational Research, Vol. 48, pp. 66-76. Hämäläinen, R. and Karjalainen, R. (1992), “Decision support for risk analysis in energy policy”, European Journal of Operational Research, Vol. 56, pp. 172-183. Harremoës, P, Gee, D., McGarvin, M., Stirling, A., Keys, J., Wynne, B. and Guedes Vaz, S. (Eds.) (2002), Late Lessons from Early Warnings: The Precautionary Principle 1896-2000, Environmental Issue Report No. 22, European Environmental Agency (EEA), Copenhagen. Harribey, J.-M. (1998), Le développement soutenable, Economica, Paris. Harries, C. (2003), “Correspondence to what? Coherence to what? What is good scenario-based decision making?”, Technological Forecasting and Social Change, Vol. 70, pp. 797-817. Hashemian, H. (2002), “Optimized maintenance and management of ageing of critical equipment in nuclear power plants”, in Ruan, D. and Fantoni, P.F. (Eds.), Power Plant Surveillance and Diagnostics – Applied Research with Artificial Intelligence, Springer-Verlag, Berlin/Heidelberg/New York, pp. 41-49. Hecq, W. and Vouche, M. (1984), “De elektriciteitsproductie en haar weerslag op het milieu”, Consensus – Informatietijdschrift over Energie, No. 3-4, pp. 109-123. Helm, D. (2002), “Energy policy: security of supply, sustainability and competition”, Energy Policy, Vol. 30, pp. 173-184. Heyes, A. (1995), “PRA in the nuclear sector. Quantifying human error and human malice”, Energy Policy, Vol. 23, No. 12, pp. 1027-1034. Hisschemöller, M. (1993), De democratie van problemen. De relatie tussen de inhoud van beleidsproblemen en methoden van politieke besluitvorming, VU Uitgeverij, Amsterdam. Hisschemöller, M. and Hoppe, R. (1996), “Coping with intractable controversies: The case of problem structuring in policy design and analysis”, Knowledge and Policy, Vol. 8, No. 4, pp. 40-60. Hobbs, B. and Meier, P. (2000), Energy Decisions and the Environment. A Guide to the Use of Multicriteria Methods, Kluwer Academic Publishers, Boston/Dordrecht/London. Hohmeyer, O. (1988), Social Costs of Energy Consumption, Springer Verlag, Berlin. Holemans, D. (1999), Ecologie en burgerschap. Pleidooi voor een nieuwe levensstijl, Uitgeverij Pelckmans, Kapellen. Holland, A. (1994), “Natural capital”, in Attfield, A. and Belsey, R. (Eds.), Philosophy and the Natural Environment, Cambridge University Press, Cambridge. Holland, A. (1997), “Substitutability. Or, why strong sustainability is weak and absurdly strong sustainability is not absurd”, in Foster, J. (Ed.), pp. 122-134. Hourcade, J. (1997), “Précaution et action séquentielle de la décision face aux risques climatiques de l’effet de serre”, in Godard, O. (Ed.), pp. 260-294. Hughes, T. (1987), “The evolution of large technical systems”, in Bijker, W., Hughes, T. and Pinch, T. (Eds.), pp. 51-82. Hunt, J., Littlewood, D. and Thompson, B. (2003), Developing participatory consultation – A review of learning from four experimental dialogue processes, IEPPP Report 4.10, Institute for Environment, Philosophy and Public Policy, Lancaster University, Lancaster.

496


Inglehart, R. (1977), The Silent Revolution. Changing Values and Political Styles among Western Publics, Princeton University Press, Princeton. Inglehart, R. (1990), Cultural Shift in Advanced Industrial Society, Princeton University Press, Princeton. Interdepartementale Commissie Duurzame Ontwikkeling (ICDO/CIDD) (2000), Federaal plan duurzame ontwikkeling 2000-2004, ICDO/CIDD, Brussel. Interdepartementale Commissie Duurzame Ontwikkeling (ICDO/CIDD) (2004), Federaal plan duurzame ontwikkeling 2004-2008, ICDO/CIDD, Brussel. International Atomic Energy Agency (IAEA) (2003), Guidance for the evaluation of innovative reactor concepts and fuel cycles, Report of phase 1A of the INPRO project, Vienna. International Atomic Energy Agency (IAEA), United Nations Department of Economic and Social Affairs (UNDESA), International Energy Agency (IEA), EUROSTAT and European Environment Agency (EEA) (2005), Energy Indicators for Sustainable Development: Guidelines and Methodologies, IAEA, Vienna. International Energy Agency (IEA) (1993), Expert Workshop on Life-Cycle Analysis of Energy Systems: Methods and Experience, OECD, Paris. International Energy Agency (IEA) (1994), Electricity Supply Industry: Structure, Ownership and Regulation in OECD Countries, OECD, Paris. International Energy Agency (IEA) (2001), Energy Policies of IEA Countries – Belgium 2001 Review, OECD, Paris. International Energy Agency (IEA) (2004), World Energy Outlook 2004, OECD/IEA, Paris. International Institute for Applied System Analysis (IIASA) (1995), Global energy perspectives to 2050 and beyond, IIASA, Vienna. International Panel on Climate Change (IPCC) (2001), Climate Change 2001: Impacts, Adaptation, and Vulnerability, Summary for policy makers, Cambridge University Press, Cambridge. Isaac, H. (1996), A propos de l’épistémologie et de la méthodologie de l’ouvrage de L. Boltanski & L. Thévenot “De la justification. Les économies de grandeur”, Cahier de Recherche No. 34, Centre de Recherche Economique Pure et Appliquée (CREPA), Université Paris Dauphine, Paris. Ishii, A. (2004), “Diplomatory science : Modeling the hybrid character of policy-advisory science for diplomacy”, in Biermann, F., Campe, S. and Jacob, K. (Eds.), Proceedings of the 2002 Berlin Conference on the Human Dimensions of Global Environmental Change “Knowledge for the Sustainability Transition. The Challenge for Social Science”, Global Governance Project, Amsterdam/Berlin/Potsdam/Oldenburg, pp. 261-270. Jacob, G. (1990), Site Unseen: The Politics of Siting a Nuclear Waste Repository, University of Pittsburgh Press, Pittsburgh. Jacobs, D. (1996), “Rome ne fut pas batie en ce jour. Le modèle des ‘cités’ selon Boltanski et Thévenot et le débat sur les droits et la religion en Flandre ”, Social Compass, Vol. 43, No. 4, pp. 537-561. Jacobs, M. (1997), “Environmental valuation, deliberative democracy and public decision-making institutions”, in Foster (Ed.), pp. 211-231. Jacobs, M. (1998), “Sustainable development as a contested concept”, in Dobson, A. (Ed.), Fairness and Futurity, Oxford University Press, Oxford, pp. 21-45. Jaeger, C. (1994), Taming the Dragon: Transforming Economic Institutions in the Face of Global Change, Gordon&Breach, New York. Jasanoff, Sh. (1987), “Contested boundaries in policy-relevant science”, Social Studies of Science, Vol. 17, No. 2, pp. 195-230. Jasanoff, Sh. (1990), The Fifth Branch: Science Advisers as Policy Makers, Harvard University Press, Cambridge. Jegen, M. and Wüstenhagen, R. (2001), “Modernise it, sustainabilise it! Swiss energy policy on the eve of electricity market liberalisation”, Energy Policy, Vol. 29, pp. 45-54. Joerges, B. (1999), “Do politics have artefacts?”, Social Studies of Science, Vol. 29, No. 2, pp. 411-431. Jones, M., Hope, Ch. and Hughes, R. (1990), “A multi-attribute value model for the study of UK energy policy”, Journal of the Operational Research Society, Vol. 41, No.10, pp. 919-929. Jörissen, J., Rink, D. and Kneer, G. (2000), “Nachhaltigkeit und Wissenschaft: Synopse zur Umzetsung des Leitbildes der Nachhaltigkeit in konzeptionellen Studien”, TA-Datenbank-Nachrichten, Vol. 9, No. 2, pp. 7-13. Joss, S. and Bellucci, S. (Eds.) (2002), Participatory Technology Assessment – European Perspectives, Centre for the Study of Democracy (CSD) in association with the Swiss Centre for Technology Assessment (TASwiss), London. Kahneman, D., Slovic, P. and Tversky, A. (1982), Judgement Under Uncertainty: Heuristics and Biases, Cambridge University Press, Cambridge. Keeney, R., Raiffa, H. and Meyer, R. (1976), Decisions with Multiple Objectives: Preferences and Value Trade-offs, John Wiley, New York.

References & Bibliography 497 Keeney, R., Renn, O. and von Winterfeldt, D. (1987), “Structuring West Germany’s energy objectives”, Energy Policy, August Issue, pp. 352-362. Kempf, H. (1994), La baleine qui cache la forêt, La Découverte, Paris. Kettner, M. (2002), “The disappearance of discourse ethics”, in von Schomberg, R. and Baynes, K. (Eds.), pp. 201-218. Keulartz, J., Schermer, M., Korthals, M. and Swierstra, T. (2004), “Ethics in technological culture: A programmatic proposal for a pragmatist approach”, Science, Technology and Human Values, Vol. 29, pp. 3.-29. Keune, H., Goorden, L., Van Fleteren, M. and Couder, J. (2004), Projectevaluatie SCK-SuND, Eindrapport in opdracht van het SCK•CEN, STEM, Universiteit Antwerpen, Antwerpen. Klein, H. and Kleinman, D. (2002), “The social construction of technology: structural considerations”, Science, Technology & Human Values, Vol. 27, pp. 28-52. Klinke, A. and Renn, O. (2001), “Precautionary principle and discursive strategies: Classifying and managing risks”, Journal of Risk Research, Vol. 4, No. 2, pp. 159-174. Knapen, D. (2002), “Kernenergie, een technologie uit de vorige eeuw”, Vlaams Marxistisch Tijdschrift, Vol. 36, No. 2, pp. 29-35. Kopfmüller, J. (2001), “Nachhaltige Entwicklung im Energiebereich”, TA-Datenbank-Nachrichten, Vol. 10, No. 3, pp. 6-11. Korthals, M. (1994), Duurzaamheid en democratie. Sociaal-filosofische beschouwingen over milieubeleid, wetenschap en technologie, Boom, Amsterdam/Meppel. Krewitt, W. (2002), “External costs of energy – Do the answers match the questions? Looking back at 10 years of ExternE”, Energy Policy, Vol. 30, pp. 839-848. Krupnick, A., Markandya, A. and Nickell, E. (1993), “The external costs of nuclear power: Ex ante damages and lay risks”, American Journal of Agricultural Economics, December Issue, pp. 1273-1279. Kuhn, T. (1970), The Structure of Scientific Revolutions, University of Chicago Press, Chicago. Kunsch, P. (2001), “Externalities and internalisation of radioactive waste producing activities – The analogy with environmental practices”, Proceedings of the ICEM’01 Conference (30 Sept. – 4 Oct.), Bruges (available on CD-ROM). Laes, E. and Meskens, G. (2001a), “Nuclear energy and sustainable development: Contradiction or challenge?”, Proceedings of the Global 2001 Conference “Back-end of the Fuel Cycle: From Research to Solutions” (913 Sept.), Paris (available on CD-ROM). Laes, E. and Meskens, G. (2001b), “Sustainability and radioactive waste management: Challenges for the future”, Proceedings of the ICEM’01 Conference (30 Sept. – 4 Oct.), Bruges (available on CD-ROM). Laes, E., Meskens, G., D’haeseleer, W. and Weiler, R. (2002a), “The Belgian nuclear phase out as a strategy for sustainable development: Unstructured problems, unstructured answers?”, in Biermann, F., Campe, S. and Jacob, K. (Eds.), Proceedings of the 2002 Berlin Conference on the Human Dimensions of Global Environmental Change “Knowledge for the Sustainability Transition. The Challenge for Social Science”, Global Governance Project, Amsterdam/Berlin/Potsdam/ Oldenburg, pp. 271-284. Laes, E. and Meskens, G. (2002b), Sustainable development and nuclear energy: Summary of interviews, Report R-3682, SCK•CEN, Mol. Laes, E. and Meskens, G. (2003), “An architecture for sustainability assessment: The case of nuclear energy in Belgium”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 501-511. Laes, E., Meskens, G., D’haeseleer, W. and Weiler, R. (2004a), “Trust as a central paradigm for advisory science: The case of the Belgian nuclear phase out”, International Journal of Sustainable Development, Vol. 7, No. 1, pp. 1-26. Laes, E., Meskens, G. and Ruan, D. (2004b), “Planning for sustainability in the Belgian electricity sector: A multi-criteria mapping exercise”, in Ruan, D. (Ed.), Intelligent Sensory Evaluation, Springer Verlag, Berlin/Heidelberg/New York, pp. 311-326. Laes, E., Chayapathi, L., Meskens, G. and Eggermont, G. (2004c), Kernenergie en maatschappelijk debat, Studie in opdracht van het viWTA – Samenleving en Technologie, Vlaams Instituut voor Wetenschappelijk en Technologisch Aspectenonderzoek (viWTA), Brussel. Laes, E., Meskens, G., D’haeseleer, W. and Weiler, R. (2005), “Adressing uncertainty and inequality in nuclear policy”, Journal of Enterprise Information Management, Vol. 18, No. 3, pp. 357-376. Lafaye, C. and Thévenot, L. (1993), “Une justification écologique ?”, Revue Française de Sociologie, Vol. XXXIV, pp. 495-524. Lako, P., Ybema, J. and Seebregts, A. (1998), The long-term potential of fusion power in Western Europe – MARKAL scenarios until 2100, Report ECN-C-98-071, Energie Centrum Nederland (ECN), Amsterdam. Larrère, C. (1997), Les philosophies de l’environnement, Presses Universitaires de France, Paris. Lash, S., Szerszynski, B. and Wynne, B. (1996), Risk, Environment and Modernity: Towards a New Modernity, Sage, London.

498


Latour, B. (1987), Science in Action, Harvard University Press, Cambridge. Latour, B. (1993), We Have Never Been Modern, Harvester Wheatsheaf, New York. Latour, B. (1994), “Une sociologie sans objet? Note théorique sur l’interobjectivité”, Sociologie du travail, Vol. 36, No. 4, pp. 587-607. Latour, B. (1995), “Moderniser ou écologiser? A la recherche de la septième cité”, Ecologie Politique, No. 13, pp. 5-27. Latour, B. (1999), “When things strike back – A possible contribution of science studies to the social sciences”, British Journal of Sociology, Special Milennium Issue, Vol. 51, No. 1, pp. 105-123. Latour, B. (2003a), “Du principe de précaution au principe de bon gouvernement : Vers de nouvelles règles de la méthode expérimentale”, unpublished paper, Ecole des Mines, Paris. Latour, B. (2003b), “The promises of constructivism”, in Idhe, D. (Ed.), Chasing Technoscience: Matrix of Materiality, Indiana Series for the Philosophy of Technology, University of Indiana Press, pp. 27-46. Latour, B. (2004a), The Politics of Nature. How to Bring the Sciences into Democracy?, Harvard University Press, Cambridge. Latour, B. (2004b), “A dialogue on actor-network theory with a (somewhat) socratic professor”, in Avgerou, C., Ciborra, C. and Land, F. (Eds.), The Social Study of Information and Communication Systems, Oxford University Press, Oxford, pp. 62-76. Latour, B. (2004c), “Un dialogue sur les deux systèmes de sociologie”, forthcoming in ‘Actes du Colloque de Cerisy’, available at Laurans, Y. (2001), “Economic evaluation of the environment in the context of justification conflicts: development of concepts and methods through examples of water management in France”, International Journal of Environment and Pollution, Vol. 15, No. 1, pp. 94-115. Lavrysen, L. and Maes, F. (2000), Institutioneel kader voor duurzame ontwikkeling en milieubeleid, in Mazijn, B. (Ed.), pp. 144-179. Law, J. and Mol, A. (1995), “Notes on materiality and sociality”, Sociological Review, Vol. 43, No. 2, pp. 274294. Lélé, S. (1991), “Sustainable development: A critical review”, World Development, Vol. 19, No. 6, pp. 607621. Leroy, P. (1979), Kernenergie – Milieuconflict of godsdienstoorlog?, Departement politieke en sociale wetenschappen, Universitaire Instelling Antwerpen, Antwerpen. Leroy, P. and Nelissen, N. (2000), “Milieubeleid en beleidswetenschappen”, in Driesen, P. and Glasbergen, P. (Eds.), Milieu, samenleving en beleid, Elsevier, Den Haag, pp. 71-93. Lierman, S. (2004), Voorzorg, preventie en aansprakelijkheid. Gezondheidsrechtelijke analyse aan de hand van het gebruik van ioniserende straling in de geneeskunde, Intersentia, Antwerpen/Oxford. Lindblom, Ch. (1965), The Intelligence of Democracy. Decision Making Through Mutual Adjustment, Free Press, New York. Löfstedt, R. and Frewer, L. (Ed.) (1998), The Earthscan Reader in Risk and Modern Society, Earthscan Publications, London. Lomberg, B. (2001), The Skeptical Environmentalist – Measuring the Real State of the World, Cambridge University Press, Cambridge. London Economics (2004), Structure and Functioning of the Belgian Electricity Market in a European Perspective, Final Report to the CREG, London Economics, London. Lovins, A. (1977), Soft Energy Paths: Towards a Durable Peace, Ballinger, Cambridge. LTI Research Group (Eds.) (1998), Long-Term Integration of Renewable Energy Sources into the European Energy System, Physica-Verlag, Heidelberg/New York. Mambrey, P., Plateau, M. and Tepper, A. (1995), Technikentwicklung durch Leitbilder. Neue Steuerungs- und Bewertungsinstrumente, Campus Verlag/St. Martin’s Press, Frankfurt am Main/ New York. MacKenzie, D. (1990), Inventing Accuracy: A Historical Sociology of Nuclear Missile Guidance, MIT Press, Cambridge. MacKenzie, D. and Wajcman, J. (Eds.) (1985), The Social Shaping of Technology, Open University Press, Milton Keynes. Markandya, A. (1995), “Quantifying the risks of nuclear electric energy”, in Hohmeyer, O., Richard, L. and Ottinger, K. (Eds.), Social Costs and Sustainability. Valuation and Implementation in the Energy and Transport Sector, Proceedings of an International Conference (Ladenburg, 27-30 May), Springer Verlag, Berlin, pp. 137-160. Massachusetts Institute of Technology (MIT) (2003), The Future of Nuclear Power, MIT, Cambridge. Masuhr, K. (1995), “The external costs of energy use: Internalization without the state?”, in Hohmeyer, O., Richard, L. and Ottinger, K. (Eds.), Social Costs and Sustainability. Valuation and Implementation in the Energy and Transport Sector, Proceedings of an International Conference (Ladenburg, 27-30 May), Springer Verlag, Berlin, pp. 528-546. Maturana, H. and Varela, F. (1992), The Tree of Knowledge, Shambala, Boston.

References & Bibliography 499 Mazijn, B. (Ed.) (2000), Duurzame ontwikkeling meervoudig bekeken, Academia Press, Gent. Meadows, D.H., Meadows, D.L., Randers, J. and Behrens, W. (1972), Limits to Growth, Universe Books, New York. Merton, R. (1973), The Sociology of Science: Theoretical and Empirical Investigation, University of Chicago Press, Chicago. Mestrum, F. (2003), “Poverty reduction and sustainable development”, in Nierynck, E., Vanoverschelde, A., Bauler, T., Zaccaï, E., Hens, L. and Pallemaerts, M. (Eds.), Making Globalisation Sustainable? The Johannesburg Summit on Sustainable Development and Beyond, VUBPress, Brussels, pp. 51-64. Michiels, J. (2003), ”De wet van 31 januari 2003 houdende de geleidelijke uitstap uit de kernenergie voor industriële elektriciteitsproductie”, Milieu-en Energierecht, No. 2, pp. 86-101. Mieg, H. (2002), “The precarious role of scenarios in global environmental politics. Political options versus scientific projections”, in Biermann, F., Campe, S. and Jacob, K. (Eds.), Proceedings of the 2002 Berlin Conference on the Human Dimensions of Global Environmental Change “Knowledge for the Sustainability Transition. The Challenge for Social Science”, Global Governance Project, Amsterdam/Berlin/Potsdam/Oldenburg, pp. 67-75. Miller, C. (2005), “New civic epistemologies of quantification: Making sense of indicators of local and global sustainability”, Science, Technology and Human Values, Vol. 30, pp. 403-432. Mormont, M. (1995), “Expertise scientifique et action publique. Le cas du changement climatique dans trois pays Européens”, Natures, Sciences, Sociétés, Vol. 3, No. 1, pp. 16-25. Mormont, M. (1999), “Un questionnement durkheimien du développement durable”, Environnement et Société, No. 21, pp. 69-79. Mormont, M. (2002), “Préface : Du concept au mode d’emploi”, in Zaccaï, E., pp. 9-16. Mortier, F. and Raes, K. (1997), Een kwestie van behoren. Stromingen in de hedendaagse ethiek, Mys & Breesch, Gent. Moss, R. and Schneider, S. (2001), Uncertainties in the IPCC Third Assessment Report: Recommendations to Lead Authors for More Consistent Assessment and Reporting, IPCC guidance paper, available at . Mouffe, Ch. (1993), The Return of the Political, Verso, London/New York. Mouffe, Ch. (1999), “Deliberative democracy or agonistic pluralism?”, Social Research, Vol. 66, No. 3, pp. 745-758. Munda, G. (1997), “Environmental economics, ecological economics, and the concept of sustainable development”, Environmental Values, Vol. 6, pp. 213-233. Naert, L. (2004), “Diep in de put?” Is de definitieve berging van hoogradioactief afval rechtvaardig tegenover de toekomstige generaties? Uitwerking en toepassing van een transgenerationele ethiek gebaseerd op het werk van John Rawls, Licentiaatsverhandeling, Faculteit Letteren en Wijsbegeerte, VUB, Brussel. Nakicenovic, N., Grübler, A. and McDonald, A. (Eds.) (1998), Global Energy Perspectives, Cambridge University Press, Cambridge. Nationaal Instituut voor de Statistiek (NIS) and Federaal Planbureau (FPB) (1996), Bevolkingsvooruitzichten 1995-2050, NIS-FPB, Brussel. Nationaal Instituut voor de Statistiek (NIS) and Federaal Planbureau (FPB) (1997), Huishoudens en gezinnen – Monografie nr. 4., NIS-FPB, Brussel. National Research Council (NRC) (1989), Improving Risk Communication, National Academy Press, Washington D.C. National Research Council (NRC) (1995), Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options, National Academy Press, Washington D.C. National Research Council (NRC) (1996), Understanding Risk. Informing Decisions in a Democratic Society, National Academy Press, Washington D.C. National Research Council (NRC) (2001), The Disposition of High-Level Waste and Spent Nuclear Fuel, National Academy Press, Washington D.C. Negri, A. (1993), “Relire Boltanski et Thévenot: Sociologie et philosophie politique”, Futur Antérieur, No. 1920, available at . Neuray, J.-F. (2003), “La sortie programmée du nucléaire”, Aménagement-Environnement, No. Spécial ‘L’Energie’, pp. 46-49. Neyens, J., Devriendt, N., Dewilde, L., Dooms, G. and Neys, W. (2004), Is er plaats voor hernieuwbare energie in Vlaanderen?, Studie in opdracht van het viWTA – Samenleving en Technologie, Vlaams Instituut voor Wetenschappelijk en Technologisch Aspectenonderzoek (viWTA), Brussel. Nierynck, E., Vanoverschelde, A., Bauler, T., Zaccaï, E., Hens, L. and Pallemaerts, M. (Eds.) (2003), Making Globalisation Sustainable. The Johannesburg Summit: Sustainable Development and Beyond, VUBPress, Brussels. Niestroy, I. (2005), Sustaining Sustainability. A Benchmark Study on National Strategies Towards Sustainable Development and the Impact of Councils in Nine EU Member States, Lemma Uitgeverij, Utrecht.

500


Nieto, C. (1996), “Towards a holistic approach to the ideal of sustainability”, Techné: Journal of the Society for Philosophy and Technology, Vol. 2, No. 2, pp. 41-48. Nuclear Energy Agency (NEA) (2000), Nuclear Energy in a Sustainable Development Perspective, OECD, Paris. Nuclear Energy Agency (NEA) (2003), Nuclear Electricity Generation: What are the External Costs?, OECD, Paris. O’Connor, J. (1994), “Is capitalism sustainable?”, in O’Connor, M. (Ed.), Is Capitalism Sustainable? Political Economy and the Politics of Ecology, Guilford, New York. O’Connor, J. (1998), Natural Causes. Essays in Ecological Marxism, Guilford Press, New York. O’Riordan, T. (1988), “The politics of sustainability”, in Turner R. (Ed.), Sustainable Environmental Management. Principles and Practice, Belhaven Press, London, pp. 29-50. Organisation for Economic Co-operation and Development (OECD) (2002), OECD Programme on Sustainable Consumption – Policy Case Study Series: Participatory Decision making for Sustainable Consumption, Report ENV/EPOC/WPNEP(2001)17/FINAL, OECD, Paris. Ostrom, E. (1992), Crafting Institutions for Self-Governing Irrigation Systems, Institute for Contemporary Studies, San Fransisco. Ottinger, R., Wooley, D., Robinson, N., Hodas, D. and Babb, S. (1990), Environmental Costs of Electricity, Oceana Publications, New York. Owen, A. (2004), “Environmental externalities, market distortions and the economics of renewable energy technologies”, The Energy Journal, Vol. 25, No. 3, pp. 127-156. Oxford Economic Research Associates Ltd. (OXERA) (2000), Policy, Risk and Science: Securing and Using Scientific Advice, Research Report 295/2000, Health and Safety Executive, Norwich. Pallemaerts, M. (1995), “De opkomst van het begrip ‘duurzame ontwikkeling’ in het internationaal juridisch en politiek discours: Een conceptuele revolutie”, Recht en Kritiek, Vol. 21, No. 4, pp. 60-77. Parry, G. (1996), “The characterization of uncertainty in probabilistic risk assessment of complex systems”, Reliability Engineering and System Safety, No. 54, pp. 119-126. Paskaleva, K. (2000), “Operationalizing integrative sustainability in national policy frameworks”, TADatenbank-Nachrichten, Vol. 9, No. 2, pp. 14-20. Pearce, D. (1998), Economics and Environment. Essays on Ecological Economics and Sustainable Development, Edward Elgar, Cheltenham/Northampton. Pearce, D. (2002), “Energy policy and externalities: An overview”, in Nuclear Energy Agency (NEA), Externalities and Energy Policy: The Life Cycle Approach, Workshop proceedings (15-16 November 2001), OECD, Paris, pp. 23-44. Pearce, D. and Turner, R. (1990), Economics of Natural Resources and the Environment, John Hopkins University Press, Baltimore. Pearce, D., Bann, C. and Georgiou, S. (1992), The Social Costs of Fuel Cycles, report for the UK Department of Trade and Industry, University College, London. Pearce, D., Groom, B. Hepburn, C. and Koundouri, Ph. (2003), “Valuing the future. Recent advances in social discounting”, World Economics, Vol. 4, No. 2, pp. 121-141. Peeters, J. (1997), “Duurzame ontwikkeling: contradicties en alternatieven”, Oikos, Vol. 1, No. 3, pp. 51-79. Pellizzoni, L. (2003), “Knowledge, uncertainty and the transformation of the public sphere”, European Journal of Social Theory, Vol. 6, No. 3, pp. 327-355. Pels, D. (1996), “The politics of symmetry”, Social Studies of Science, Vol. 26, pp. 277-304. Pepper, D. (1998), “Sustainable development and ecological modernization: A radical homocentric perspective”, Sustainable Development, No. 6, pp. 1-7. Pezzey, J. (1992), “Sustainability: An interdisciplinary guide”, Environmental Values, Vol. 1, No. 4, pp. 321362. Pezzoli, K. (1997), “Sustainable development: A transdisciplinary overview of the literature”, Journal of Environmental Planning and Management, Vol. 40, pp. 549-574. Pierre, J.P. et al. (2002), “Lettre à la rédaction. Avis de l’Institut de protection et de sûreté nucléaire sur l’utilisation de la dose collective”, Radioprotection, Vol. 37, pp. 77-84. Pinch, T. (1996), “The social construction of technology: A review”, in Fox, R. (Ed.), Technological Change: Methods and Themes in the History of Technology, Harwood Academic Publishers, Australia, pp. 17-35. Pinch, T. and Bijker, W. (1984), “The social construction of facts and artefacts: Or how the sociology of science and the sociology of technology might benefit from each other”, Social Studies of Science, Vol. 14, pp. 399-441. Pinch, T. and Bijker, W. (1987), “The social construction of facts and artefacts: Or how the sociology of science and the sociology of technology might benefit from each other”, in Bijker, W., Hughes, T. and Pinch, T. (Eds.), pp. 17-50. Plouchard, G. and Fradet, A. (2005), “Le stockage du CO2 – Evaluation du potentiel de capture et du stockage du CO2 dans le monde ”, Les Cahiers du CLIP, No. 17, pp. 3-40.

References & Bibliography 501 Proost, S. and Van Regemorter, D. (2000), How to achieve the Kyoto target in Belgium – modelling methodology and some results, CES Working Paper Series nr. 2000-09, Centre for Economic Studies (CES), KULeuven, Leuven. Proost, S. and Van Regemorter, D. (2001), “What do the AMPERE results imply for future electricity production in Belgium – An analysis with the MARKAL model”, Revue E-Tijdschrift, 117e jaargang, No. 3, pp. 41-48. Rabl, A. (1996), “Discounting of long-term costs: What would future generations prefer us to do?”, Ecological Economics, Vol. 17, pp. 137-145. Radetzki, M. (2000), “Coal or nuclear in new power stations: The political economy of an undesirable but necessary choice”, The Energy Journal, Vol. 21, pp. 135-147. Ralston Saul, J. (1995), The Unconscious Civilization, The Free Press, New York/London/Toronto/Sydney/ Singapore. Raskin, P., Swart, R. and Robinson, J. (2002), “Navigating the sustainability transition: The future of scenarios”, in Biermann, F., Campe, S. and Jacob, K. (Eds.), Proceedings of the 2002 Berlin Conference on the Human Dimensions of Global Environmental Change “Knowledge for the Sustainability Transition. The Challenge for Social Science”, Global Governance Project, Amsterdam/Berlin/Potsdam/Oldenburg, pp. 53-66. Rauschmayer, F. (2000), “Ethics of multicriteria analysis”, International Journal of Sustainable Development, Vol. 3, No. 1, pp. 16-25. Rawls, J. (1971), A Theory of Justice, Harvard University Press, Cambridge. Rawls, J. (1996), Political Liberalism, Columbia University Press, New York. Redclift, M. (1987), Sustainable Development – Exploring the Contradictions, Methuen, London/New York. Rehg, W. and Bohman, J. (2002), “Discourse and democracy”, in von Schomberg, R. and Baynes, K. (Eds.), pp. 31-60. Renn, O. (1998), “The role of risk communication and public dialogue for improving risk management”, Risk Decision and Policy, Vol. 3, No. 1, pp. 5-30. Renn, O., Webler, Th. and Wiedemann, P. (Eds.) (1994), Fairness and Competence in Citizen Participation – Evaluating Models for Environmental Discourse, Kluwer Academic Publishers, Dordrecht/Boston/London. Renn, O. and Klinke, A. (1999), “Risk evaluation and risk management for institutional and regulatory policy”, in Stirling (Ed.), Vol. II, pp. 12-35. Renn, O., Dreyer, M. and Klinke, A. (Eds.) (2003), Regulatory Strategies and Research Needs to Compose and Specify a European Policy on the Application of the Precautionary Principle, Executive Summary of a EUproject (PrecauPri), Stuttgart. Reynebeau, B. (2000), Het energievraagstuk in België: De opkomst van de kernenergie, unpublished master’s thesis, Faculteit Letteren & Wijsbegeerte, UGent, Gent. Rip, A. (1999), “Contributions from social studies of science and constructive technology assessment”, in Stirling, A. (Ed.), Vol. II, pp. 94-122. Rip, A., Misa, Th. and Schot, J. (Eds.) (1995), Managing Technology in Society. The Approach of Constructive Technology Assessment, Pinter Publishers, London. Rip, A. and Talma, S. (1998), “Antagonistic patterns and new technologies”, in Disco, C. and van der Meulen, B. (Eds.), Getting New Technologies Together, Walter de Gruyter, Berlin, pp. 285-306. Risk Assessment Review Group (RARG) (1978), Risk Assessment Review Group Report to the U.S. Nuclear Regulatory Commission, Report NUREG/CR-0400, USNRC, Washington. Rist, G. (1996), Le développement. Histoire d’une croyance occidentale, Presses de Science Po, Paris. Roberts, J. (1990), “Privatization, pluralism and PWR’s – Looking beyond market forces”, Technology Analysis & Strategic Management, Vol. 2, No. 3, pp. 237-251. Robinson, J. (1982), “Energy backcasting: A proposed method of policy analysis”, Energy Policy, Vol. 10, No. 4, pp. 337-344. Romerio, F. (2005), “Risk analysis in the field of energy problems”, Rapports de recherche du CUEPE, No. 6, Université de Genève. Rosa, E. and Clark, D. (1999), “Historical routes to technological gridlock : Nuclear technology as prototypical vehicle”, Research in Social Problems and Public Policy, No. 7, pp. 21-57. Roth, E., Morgan, G., Fischhof, B., Lave, L. and Bostrom, A. (1990), “What do we know about making risk comparisons?”, Risk Analysis, Vol. 10, No. 3, pp. 375-387. Rotmans, J. (2004), Transitiemanagement. Sleutel voor een duurzame samenleving, Van Gorcum, Assen. Rotmans, J. and de Vries, B. (1997), Perspectives on Global Change. The TARGETS Approach, Cambridge University Press, Cambridge. Rotmans, J., Kemp, R., van Asselt, M., Geels, F., Verbong, G. and Molendijk, K. (2000), Transities & transitiemanagement. De casus van een emissiearme energievoorziening, International Centre for Integrative Studies (ICIS) / Maastricht Economic Research Institute on Innovation and Technology (MERIT), Universiteit Maastricht, Maastricht.

502


Rotmans, J. and van Asselt, M. (2001), “Uncertainty management in integrated assessment modeling: towards a pluralistic approach”, Environmental Monitoring and Assessment, No. 69, pp. 101-130. Rowe, G. and Frewer, L. (2004), “Evaluating public participation exercises: A research agenda”, Science, Technology and Human Values, Vol. 29, No. 4, pp. 512-557. Ruggie, J. (1998), Constructing the World Polity: Essays on International Institutionalisation, Routledge, London. Sachs, W. (Ed.) (1992), The Development Dictionary: A Guide to Knowledge as Power, Zed Books, London/ New York. Sachs, W. (1999), Planet Dialectics: Explorations in Environment & Development, Zed Books, London/New York. Sachs, W. (Ed.) (2002), The Jo’burg Memo – Fairness in a Fragile World. Memorandum for the World Summit on Sustainable Development, Heinrich Böll Foundation, Berlin. Salo, A. (1999), “On the role of decision analytic modelling”, in Stirling, A. (Ed.), Vol. II, pp. 124-141. Sandin, P. (2004), Better Safe than Sorry: Applying Philosophical Methods to the Debate on Risk and the Precautionary Principle, Theses in Philosophy from the Royal Institute of Technology, Stockholm. Schieber, C. and Schneider, Th. (2002), “The external costs of the nuclear fuel cycle”, in Nuclear Energy Agency (NEA), Externalities and Energy Policy: The Life Cycle Approach, Workshop proceedings (15-16 November 2001), OECD, Paris, pp. 111-129. Schleisner, L. (2000), “Comparison of methodologies for externality assessment”, Energy Policy, Vol. 28, pp. 1127-1136. Schmitter, P. (1995), “The irony of modern democracy and the viability of efforts to reform its practice”, in Wright, E. (Ed.), Associations and Democracy, Verso, London, pp. 167-183. Schrader-Frechette, K. (1980), “The role of applied philosophy of science in assessment”, Science, Technology and Human Values, Vol. 6, No. 33, pp. 33-50. Schrader-Frechette, K. (1985), Science Policy, Ethics and Economic Methodology, D. Reidel Publishing Company, Dordrecht/Boston/Lancaster. Schrader-Frechette, K. (1991), “Ethical dilemmas and radioactive waste: A survey of the issues”, Environmental Ethics, Vol. 13, pp. 327-343. Sellars, W. (1963), Science, Perception and Reality, Routledge & Kegan Paul, London. Shi, T. (2004), “Ecological economics as a policy science: rhetoric or commitment towards an improved decision-making process on sustainability”, Ecological Economics, No. 48, pp. 23-36. Sismondo, S. (1993), “Some social constructions”, Social Studies of Science, Vol. 23, pp. 479-513. Sjöberg, L. (1997), “Explaining risk perception: An empirical evaluation of cultural theory”, Risk Decision and Policy, Vol. 2, No. 2, pp. 113-130. Sjöberg, L. (2003), “Risk perception is not what it seems. The psychometric paradigm revisited”, in Andersson, K. (Ed.), VALDOR 2003 – VALues in Decisions On Risks – Proceedings, Stockholm, pp. 1429. Slocum, N. (2003), Participatory Methods Toolkit – A Practitioner’s Manual, King Baudouin Foundation/ Flemish Insitute for Science and Technology Assessment (viWTA), Brussels. Sloman, S. (1996), “The empirical case for two systems of reasoning”, Psychological Bulletin, Vol. 119, No. 1, pp. 3-22. Slovic, P. (1993), “Perceived risk, trust and democracy”, Risk Analysis, Vol. 13, No. 6, pp. 675-682. Slovic, P. (1999), “Trust, emotion, sex, politics and science: Surveying the risk-assessment battlefield”, Risk Analysis, Vol. 19, No. 4, pp. 689-701. Slovic, P., Fischhoff, B. and Lichtenstein, S. (2000), “Cognitive processes and societal risk taking”, (reprint from Slovic et al. 1976) in Slovic, P. (Ed.), The Perception of Risk, Earthscan Publications, London, pp. 32-50. Smil, V. (2000), “Perils of long-range energy forecasting: Reflections on looking far ahead”, Technological Forecasting and Social Change, No. 65, pp. 251-264. Smithson, M. (1989), Ignorance and Uncertainty: Emerging Paradigms, Springer Verlag, New York/Berlin. Smits, R. and Leyten, J. (1991), Technology Assessment : waakhond of speurhond ?, Uitgeverij Kerkebosch, Zeist. Söderholm, P. and Sundqvist, Th. (2003), “Pricing environmental externalities in the power sector: ethical limits and implications for social choice”, Ecological Economics, Vol. 46, pp. 333-350. Stake, R. (1995), The Art of Case-Study Research, Sage Publications, Thousand Oaks. Stengers, I. (1993), L’invention des sciences modernes, Champs Flammarion, Paris. Stengers, I. (1999), “Le développement durable: une nouvelle approche?”, Alliage, No. 40, pp. 31-39. Stenmark, M. (2002), “The relevance of environmental ethical theories for policy making”, Environmental Ethics, Vol. 24, pp. 135-148. Stirling, A. (1994), “Diversity and ignorance in electricity supply investment – Adressing the solution rather than the problem”, Energy Policy, March Issue, pp. 195-216.

References & Bibliography 503 Stirling, A. (1997), “Multi-criteria mapping. Mitigating the problems of environmental valuation?”, in Foster, J. (Ed.), pp. 186-210. Stirling, A. (1998), “On the economics and analysis of diversity”, Working Paper No. 28, Science Policy Research Unit (SPRU), University of Sussex, Brighton. Stirling, A. (Ed.) (1999a), On Science and Precaution in the Management of Technological Risk (two volumes), Final Report of a project for the EC Institute for Prospective Technological Studies (JRC-IPTS) under the auspices of the European Science and Technology Observatory network (ESTO), report EUR 19056/EN/1+2, Brussels. Stirling, A. (1999b), “The appraisal of sustainability: Some problems and possible responses”, Local Environment, Vol. 4, No. 2, pp. 111-135. Stirling, A. and Mayer, S. (1999), Rethinking risk: A pilot multi-criteria mapping of a genetically modified crop in agricultural systems in the UK, SPRU Report No. 21, Science and Policy Research Unit (SPRU), University of Sussex. Stirling, A. and Mayer, S. (2000), “A precautionary approach to technology appraisal? A multi-criteria mapping of genetic modification in UK agriculture”, TA-Datenbank Nachrichten, Vol. 9, No. 3, pp. 39-51. Stuurgroep Maatschappelijke Discussie Energiebeleid (1983), Het eindrapport van de brede maatschappelijke discussie energiebeleid, Stenfert Kroeze, Leiden. Summerton, J. (1996), “On watts, wires and webs. Reshaping electricity?”, in Ingelstam, L. (Ed.), Complex Technological Systems, Swedish Council for Planning and Coordination of Research, Stockholm. Sundqvist, Th. (2004), “What causes the disparity of electricity externality estimates?”, Energy Policy, Vol. 32, pp. 1753-1766. Task Force Duurzame Ontwikkeling (1999), Een stap naar duurzame ontwikkeling? Federaal rapport inzake duurzame ontwikkeling 1999, Federaal Planbureau, Brussel. Task Force Duurzame Ontwikkeling (2002), Een stap naar duurzame ontwikkeling? Federaal rapport inzake duurzame ontwikkeling 2002, Federaal Planbureau, Brussel. Thévenot, L. (2002), “Conventions of coordination and the framing of uncertainty”, in Fullbrook, E. (Ed.), Intersubjectivity in Economics, Routledge, London, pp. 181-197. Tilman, V. (1998), “Le concept de développement durable”, Actualités du Droit, pp. 267-362. Thomas, S., Fischedick, M. and Wolters, D. (2001), Methodological comments on the report of the Commission AMPERE, Wuppertal Institut für Klima, Umwelt & Energie GmbH, Wuppertal. Thompson, M., Ellis R., and Wildavsky, A. (1990), Cultural Theory, Westview Press, San Francisco/Oxford. Thompson, P. (1986), “Uncertainty arguments in environmental issues”, Environmental Ethics, Vol. 8, pp. 5975. Torfs, R. (2001), Externe kosten van elektriciteitsproductie. Fase 3 van het CO2 project, Finaal rapport 2001/IMS/R/123, Vlaamse Instelling voor Technologisch Onderzoek (VITO), Mol. Torfs, R., D'haeseleer, W. and Mathieu, Ph. (2000), Emissies en energieverbruik van energiedragers, Etijdschrift, 116e Jaargang, Speciaal nummer (herfst 2000), pp. 45-49. Torgeson, D. (1999), The Promise of Green Politics – Environmentalism and the Public Sphere, Duke University Press, Durham. Turkenburg, W. (2004), Energietransitie richting duurzaamheid: Het technologisch perspectief, VROMRaad/ Energieraad/Copernicus Instituut (Universiteit Utrecht), Den Haag/Utrecht. U.K. Department for Trade and Industry (DTI), Department for Transport and Department for Environment, Food and Rural Affairs (DEFRA) (2003), Our Energy Future – Creating a Low Carbon Economy, White Paper, published by TSO, London/Etc. United Nations Commission on Sustainable Development (UNCSD) (1996), Indicators of Sustainable Development: Framework and Methodologies, United Nations, New York. United Nations Conference on Environment and Development (UNCED) (1992), Agenda 21, Nederlandse vertaling (1993): Ministerie van Volkshuisvesting, Ruimtelijke Ordening en Milieubeheer en het Ministerie van Buitenlandse Zaken (Directoraat Generaal Internationale Samenwerking), Den Haag. United Nations Development Programme (UNDP) (1999), Human Development Report, Oxford University Press, New York. United Nations Framework Convention on Climate Change (UNFCCC) (1997), Kyoto Protocol to the United Nations Framework Convention on Climate Change. United States National Academy of Sciences (USNAS) (1999), Our Common Journey: A Transition Towards Sustainability, report of the Board on Sustainable Development of the National Research Council, National Academy Press, Washington D.C. United States Nuclear Regulatory Commission (USNRC) (1975), Reactor Safety Study: An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants, USNRC Report WASH-1400, Washington D.C. United States Office of Technology Assessment (USOTA) (1994), Studies of the Environmental Costs of Electricity, OTA, Washington D.C.

504


Valentin, A. and Spangenberg, H. (1999), Indicators for sustainable communities, Wuppertal Institut für Klima, Umwelt & Energie GmbH, Wuppertal. Van Assche, J. (2000), “Afwenteling van socio-ecologische problemen, partim energie”, in Mazijn, B. (Ed.), pp. 261-283. van Asselt, M. (2000), Perspectives on Uncertainty and Risk – The PRIMA Approach to Decision Support, Kluwer Academic Publishers, Boston/Dordrecht/London. van Brakel, J. (1998), Interculturele communicatie en multiculturalisme, Universitaire Pers Leuven/Van Gorcum, Leuven/Assen. Van de Graaf, H. en Hoppe, R. (1996), Beleid en politiek – Een inleiding tot de beleidswetenschap en de beleidskunde, Coutinho, Bussum. Vandenabeele, J. (1999), Sociale verantwoording en het debat over landbouw, milieu en natuur. Een kritisch empirisch onderzoek naar het leren van volwassenen in het perspectief van duurzaamheid, unpublished PhD thesis, Faculteit Psychologie en Pedagogische Wetenschappen (Afdeling Sociale Pedagogiek en Gezinspedagogiek), KULeuven, Leuven. Vandenabeele, J., Couder, J., Van Fleteren, M. and Bongaerts, V. (2005), Determinanten van het huishoudelijk energiegebruik, Studie in opdracht van het viWTA – Samenleving en Technologie, Vlaams Instituut voor Wetenschappelijk en Technologisch Aspectenonderzoek (viWTA), Brussel. van den Daele, W. (2000), “Interpreting the precautionary principle – Political versus legal perspectives”, in Cottam, M., Harvey, D., Pape, R. and Tait, J. (Eds.), Foresight and Precaution – Vol. 1, Proceedings of ESREL 2000, SARS and SRA-Europe Annual Conference (Edinburgh, 15-17 May 2000), Balkema, Rotterdam, pp. 213-222. van der Heijden (1997), Scenarios: The Art of Strategic Conversation, John Wiley Publishers, New York. Van Hulle, F. et al. (2004), Optimal offshore wind energy developments in Belgium, Final report (CP/21), Belgian Science Policy, Brussels. Vanooteghem, H. (2005), “Elektriciteitsbronnen gewikt en gewogen”, Het Ingenieursblad, No. 8-9, pp. 16-22. Varner, G. (1994), “Environmental law and the eclipse of land as private property”, in Ferré, F. and Hartel, P. (Eds.), Ethics and Environmental Policy: Theory Meets Practice, University of Georgia Press, Athens, pp. 142-160. Verbruggen, A. (1986), “De besluitvorming inzake de bouw van kerncentrales in België”, in Liber Amicorum Prof. Pierre-Henri Virenque, Studiecentrum voor Economisch en Sociaal Onderzoek (SESO), UFSIA, Antwerpen. Verbruggen, A. (2004), “Tradable green certificates in Flanders (Belgium)”, Energy Policy, Vol. 32, pp. 165176. Verbruggen, A. and Vanderstappen, E. (2003), “Electricity sector restructuring in Belgium”, in Glachant, J.-M. and Finon, D. (Eds.), Competition in European Electricity Markets – A Cross-Country Comparison, Edward Elgar Publishers, Cheltenham/Camberley/Northampton, pp. 285-310. Verburg, R. and Wiegel, V. (1997), “On the compatibility of sustainability and economic growth”, Environmental Ethics, Vol. 19, pp. 247-265. Vinck, D. (1995), Sociologie des sciences, Armand Colin, Paris. Visser ‘t Hooft, H. (1991), “Développement technologique et responsabilité envers les générations futures”, Archives de Philosophie du Droit, Tome 36 (Droit et Science), pp. 31-47. Vlek, Ch. (1990), Beslissen over risico-acceptatie, Gezondheidsraad, Den Haag. von Schomberg, R. (2002), “The erosion of our value spheres: The ways in which society copes with scientific, moral, and ethical uncertainty”, in von Schomberg, R. and Baynes, K. (Eds.), pp. 219-238. von Schomberg, R. (2006), “The precautionary principle and its normative challenges”, in Fischer, E., Jones, J. and von Schomberg, R. (Eds.), Implementing the Precautionary Principle: Perspectives and Prospects, Edward Elgar Publishers, Cheltenham/Camberley/Northampton, pp. 19-42, in press. von Schomberg, R. and Baynes, K. (Eds.) (2002), Discourse and Democracy. Essays on Habermas’s Between Facts and Norms, State University of New York Press, New York. Voorspools, K., Brouwers, E. and D'haeseleer, W. (2000), “Energy content and indirect greenhouse gas emissions embedded in ‘emission-free’ power plants: Results for the Low Countries”, Applied Energy, Vol. 67, pp. 307-330. Vranckx, A. (2000), Wetenschap voor beleid: Een communicatieprobleem?, VUBPress, Brussel. Webler, Th. (1994), “‘Right’ discourse in citizen participation: An evaluative yardstick”, in Renn, O., Webler, Th. and Wiedemann, P. (Eds.), pp. 35-86. Weiler, R. and Holemans, D. (1993) (Ed.), Bevrijding of bedreiging door wetenschap en techniek, Pelckmans, Kapellen. Weiler, R. and Holemans, D. (1994) (Ed.), Gegrepen door techniek, Pelckmans, Kapellen. Weiler, R. and Van Ooteghem, H. (1999), “Technologie: Roerganger van onze risicomaatschappij?”, Het Ingenieursblad, No. 8-9, pp. 36-42.

References & Bibliography 505 Wetenschappelijke Raad voor het Regeringsbeleid (WRR) (1994), Duurzame risico’s: een blijvend gegeven, Standaard Uitgeverij, Den Haag. Weyer, J. (2005), “Staatliche Förderung von Groβtechnikprojecten. Ein dysfunktionaler Anachronismus im Zeitälter der Globalisierung?”, Technikfolgenabschätzung – Theorie und Praxis, Vol. 14, No. 1, pp. 18-25. Wildavsky, A. and Dake, K. (1990), “Theories of risk perception: who fears what and why?”, Daedalus, No. 112, pp. 41-60. Williams, R. and Edge, D. (1996), “The social shaping of technology”, Research Policy, No. 25, pp. 865-899. Winner, L. (1980a), “Do artefacts have politics?”, Daedalus, No. 109, pp. 121-136. Winner, L. (1980b), The Whale and the Reactor. A Search for Limits in an Age of High Technology, The University of Chicago Press, Chicago/London. Winner, L. (1993), “Upon opening the black box and finding it empty: Social constructivism and the philosophy of technology”, Science, Technology and Human Values, Vol. 18, pp. 362-378. Woolgar, S. (1991), “The turn to technology in social studies of science”, Science, Technology and Human Values, Vol. 16, pp. 20-50. Woolgar, S. (1993), “What’s at stake in the sociology of technology? A reply to Pinch and Winner”, Science, Technology and Human Values, Vol. 18, pp. 523-529. Woolgar, S. and Cooper, G. (1999), “Do artefacts have ambivalence? Moses’ bridges, Winner’s bridges and other urban legends in S&TS”, Social Studies of Science, Vol. 29, No. 3, pp. 433-449. World Commission on Environment and Development (WCED) (1987), Our Common Future, Oxford University Press, Oxford. World Energy Council (WEC) (2000), Energy and the challenge of sustainability, WEC, London/Etc. World Wide Fund for Nature (WWF) (2000), Climate Change and Nuclear Power, WWF, Gland. Wynne, B. (1992), “Uncertainty and environmental learning”, Global Environmental Change, Vol. 2, pp. 111127. Wynne, B. (1995), “Technology assessment and reflexive social learning: Observations from the risk field”, in Rip, A., Misa, Th. and Schot, J. (Eds.), pp. 19-36. Wynne, B. (2001), “Managing scientific uncertainty in public policy”, paper presented at the “Biotechnology and Global Conference: Crisis and Opportunity” (26-28 April), Cambridge, available at Zaccaï, E. (1999), “Caracteristiques du développement durable: Un essai de synthese”, Les Cahiers du CEDD, No. 4, Centre d’Etudes du Développement Durable (CEDD), Bruxelles. Zaccaï, E. (2002), Le développement durable – Dynamique et constitution d’un projet, PIE Lang, Bruxelles/ Bern/Berlin/Frankfurt/New York/Oxford/Wien. Zaccaï, E. and Missa, J. (2000) (Eds.), Le principe de précaution. Significations et conséquences, Editions de l’Université de Bruxelles, Bruxelles. Žižek, S. (1998), Pleidooi voor intolerantie (Nederlandse vertaling), Uitgeverij Boom, Amsterdam. Zuinen, N. (2004), Indicateurs pour un développement durable : aspects méthodologiques et développements en cours, Working paper 4-04, Bureau Fédéral du Plan, Bruxelles.

ANNEX 1 – CASE-STUDY METHODOLOGY Besides a broad theoretically-inspired literature study, case-studies on particular interactions in the field of sustainable energy governance occupy a central position in our thesis. A case-study is in principle no different from a survey or an experiment in the sense that it also aims at scientific explanation through empirical research. However, in a case-study, a certain phenomenon is studied in its ‘natural’ setting within the context of a dynamic process. An experimental set-up or a survey on the other hand aims to reduce (or nullify) this inherent interrelation of phenomenon and context. For our purposes however, it was important to explore the interlacement between phenomenon and context as we aimed to explore how the concept of ‘sustainable energy’ acquires its particular meaning in networks of actors, scientific knowledge, power and influence, cultural backgrounds, etc. An empirical-analytical approach, working within the confines of a limited amount of variables and hypotheses, would simply prove to be too restrictive for our purposes. Case-study research is generally subject to some possible prejudices. First of all, it is often said to lack rigour and provide limited basis for scientific generalisation, largely because of the supposed inherent subjectivity of case-study research. Secondly, it generally takes long to do case-study research while still a limited amount of actors can be followed; and thus, the research is said to generally pay off little in the advancement of social practice (Stake 1995). While these remarks may have some element of truth, we maintain that a rigorous case-study design can offer high-quality results. The quality of the results is, just as for experiments or surveys, dependent on a careful consideration of all steps of case-study research methodology. These include: designing the casestudy protocol (selection of ‘relevant’ participants, designing the research protocol), data collection, data analysis and validation. We will go into the details of each of these steps in the following sections.

1. Selection of participants For the interviews, it was our intention not only to analyse visions more systematically, but also to include a greater plurality of visions. Of course, the question is then: which actors to involve and why? The question becomes particularly relevant in a highly institutionalised context, i.e. a context where the rules of interaction have been defined relatively clearly and policy processes unwind according to a clearly defined logic. Energy production and consumption is (historically) a typical example of a highly institutionalised domain: perceived problems usually have a high economic and social importance and are dealt with in ‘powerful’ commissions where the government, employers and employees are represented. Traditionally, the Control Committee for Electricity and Gas (CCEG) controlled the electricity and gas markets. The International Energy Agency (IEA 1994, pp. 175-177) observes that: “…The committee, unlike US public utility commissions, has very limited resources (...). The market power of Electrabel is enormous, and the regulatory response has not been correspondingly strong. (...) The significant degree of cross-ownership in the industry – together with a high degree of horizontal and vertical integration, a light form of organisation, and a mixture of formal and informal relationships among production, transmission and distribution companies – has produced a centralised and monopolistic power sector…”. Why then bother to include other problem definitions or other types of interaction? We see two advantages:

508

Annex 1

• Other actors can bring into the discussion a greater variety on perspectives, based on scientific knowledge, but also based on more intuitive knowledge, experience, personal involvement and value positions. In particular, by framing our question on the future of nuclear energy in a sustainable development perspective, we encourage participants to reason from internationally accepted principles such as precaution, a perspective of global solidarity, interests of future generations and broad participation by stakeholders – an altogether different framing of the problem than strictly based on near-term socio-economic considerations prevalent in the past; • Participants in the interview sessions are encouraged to think about the roles of other actors. More traditional forms of interaction apply a piecemeal approach: scientists try to give objective input to the discussion; interest groups formulate an advice based on their particular interests in the issue at hand, and government decides. In our approach, the reflection is broader: participants were asked to reflect on the value positions of scientists and the role of different groups in energy governance; interest groups are encouraged to reflect not only from predefined positions but also explicitly referring to underlying values and the validity of different argumentations was tested. We were of the opinion that a selection of member organisations of the FRDO could serve the purposes of wider problem framing we envisaged in our research project. The FRDO was set up in 1997, in succession to the National Council for Sustainable Development which functioned as from 1993. Its main aim is the regulating the coordination of the federal policy on sustainable development. The FRDO advises the federal authorities about that policy, at the federal government’s and parliament’s request, as well as on its own initiative. In addition to its advisory duties the FRDO acts as a forum to encourage the sustainable development debate, for instance by means of organising symposia. Experts in the area, representatives of government and civil society, and a wider public have the opportunity to explain their point of view and to dialogue. The Council makes use of the results when formulating advices. The Council was also given the task of sensitising organisations and citizens on the subject of sustainable development. The members of the FRDO represent various social organisations: environmental organisations, development organisations, consumers’ unions, trade unions, employers’ federations, energy producers and the world of science. Federal and regional government representatives and delegates of environmental and socio-economic advisory bodies only have an advisory vote in the meetings. Selection of FRDO member organisations was thus guided by the following logic: • We did not strive for statistical representativeness, but rather for problem representativeness. An important number of organisations which have been active in the field of energy are represented in the council: electricity utilities, NGO’s, employers’ and employees’ associations, scientists, and to a lesser degree, development organisations. Each of these actors thus represents a ‘microcosmos’ with regard to the question of sustainable energy (policy); our aim was to bring together these perspectives in order to show the patchwork of possibly overlapping or conflicting elements mobilised by these actors in order to support their point of view. Official representatives of these organisations in the council were invited by letter, but the possibility to appoint a replacement person was left open at their discretion; • We expected that members of this council would be interested in long-term explorations of societal problems;

Case-study methodology

509

• We also expected that members of this council would have experience and insight in the process of political decision-making, without becoming too involved in policy-making. The FRDO is, after all, still positioned at a safe distance from the centre of government. In the end, the groups represented in the table below were found willing to participate in our research project (interviews + scenario workshop + multi-criteria exercise). Representatives of the ministries and administrations involved in nuclear policy were also invited but did not want to participate. Participation of representatives from advisory bodies (chairman or scientific staff) might also present some difficulties. Of course, advisory bodies have no ‘official’ opinion on the questions of interest to us (since they only act as an outlet for the collective opinions of the organisations represented in the advisory body), but nonetheless we chose to include representatives of these bodies because we were more interested in their professional experience in the field (with the aim of getting as much information as possible on problem structuring) rather than any ‘official’ position. Furthermore, all participants in the research project were invited to voice their personal opinion (which of course for representatives of interest groups will coincide largely with the ‘official’ view of the organisation in question) and interview results were treated confidentially. Nevertheless, the representative of the FRDO (member of the scientific staff) chose to no longer participate in our project after the interview phase because of the possible confusion between the personal and official point of view. Representatives of academia had different academic backgrounds (human ecology, environmental sciences, climatology, law and energy policy) and thus also contributed to a broadly encompassing view on our research topic. The electricity producer SPE declined further participation after the interview because of the ‘sensitivity’ of the issue of nuclear power. Both the representative of the employers’ organisation VBO and the labour union ACV changed their professional position between the scenario workshop and the multi-criteria exercise, and hence no longer expressed an interest in further participation. We chose not to invite the people replacing them in their former organisation because they would be unfamiliar with the logical succession of the different research steps. The representative of the development NGO Oxfam declined further participation after the interview because in his view, his organisation was not really involved in the energy and/or nuclear debate at the Belgian level. Other absences are the result of other agenda commitments at the time when the research step in question was organised.

510

Letter code A1

Annex 1

Description

Environmental NGO

A2 B C1

Development NGO Labour union

C2 D1

Employers’ organisation

D2 D3 E1 E2

Energy producers

F1 F2

Academia

F3 F4 F5 G1 G2 G3

Advisory bodies

Organisation

Interviews

Greenpeace

x

Bond Beter Leefmilieu (BBL) Oxfam Fédération Générale du Travail Belge (FGTB) Algemeen Christelijk Vakverbond (ACV) Unie van Zelfstandige Ondernemers (UNIZO) Federatie van Chemische Nijverheid (FEDICHEM) Verbond van Belgische Ondernemers (VBO) Electrabel Samenwerkende Vennootschap voor Productie van Elektriciteit (SPE) Université de Liège (ULg) Vrije Universiteit Brussel (VUB) Universiteit Gent (UG) Université Catholique de Louvain (UCL) Université Libre de Bruxelles (ULB) Federaal Planbureau (FPB) Milieu- en Natuurraad Vlaanderen (MinaRaad) Federale Raad Duurzame Ontwikkeling (FRDO)

x

Scenario workshop

Multi-criteria mapping x

x

x x

x

x

x x

x

x

x

x

x

x x

x

x

x

x

x x

x

x

x

x

x x

x

x x

x

x x

x

Table A-1. The participants in the interview sessions, scenario workshop and multi-criteria mapping exercise

2. Interview protocol The interview approach was developed based on the work of Fischer (1980) and Grin et al. (1997) on action theory. In describing the rationality informing a particular actor’s political judgement, these authors distinguish between first order notions, which are specific for a situation, and the more generic second order convictions underlying them. First order notions include two layers: assessment of the means to achieve given objectives (solution assessments), and problem definitions that contextually vindicate these objectives. Second order notions consist in the background theories that an actor prefers on the one hand, and the deeper preferences that he wishes to realise on the other (dependent on his value system, preferred social order, etc.). The ensemble of first and second order notions is called an ‘action theory’ (Grin et al. 1997, pp. 38-39).


511

To summarise: First order notions • How does the actor evaluate costs, effects, and side effects of different solutions for a given problem – as he defines it? • What is for the actor, in the given situation, precisely the problem (or opportunity)? Second order notions • Which background theories (habitual ways to reflect on the world) does the actor employ? • What are the deeper preferences that the actor wishes to realise? At first glance, the concept might appear very abstract. However, it can be made operational by drawing up an interview scheme in order to probe the link between a person’s assumptions and their judgement in a concrete situation. First order notions can be identified relatively easily. One may directly inquire into solution assessments (what are, for that actor, the costs, effects and side effects of a particular solution?). Similarly, one can directly inquire into problem definitions (what does the actor consider to be the problem or ‘challenge’, what criteria does he use to judge the situation and how are they weighted?). Although one may also be able to inquire directly into second order convictions, this may be more difficult, if only because such underlying assumptions may be hidden from the actor himself. Herein lies a major challenge for the analyst, through his hermeneutic effort. In Grin et al. (1997, pp. 38-44; 66-71), it is argued that one way to tackle this problem is to complement direct inquiry about the specific content of a first order notion with asking ‘why’ questions. By asking why somebody considers a particular aspect an advantage, or why he gives more weight to the claimed advantages than to the admitted disadvantages, his problem definition may be revealed. Similarly, by asking why the problem is defined in a particular way, one may identify the relevant elements of the worldviews and value systems underlying them. And asking why these values and causalities are considered important is likely to yield an answer in terms of ‘because they guide actions so that we get closer to the world I prefer.’ How these considerations influenced our interview protocol is explained in the following table. In general, the questions probed into the role of the respondent or the respondent’s organisation in the debate on sustainable energy; the goals pursued in this debate; the general interpretation of the principle of sustainable development in the context of energy production and consumption; the view on some (broadly conceived) suggested policy options; the perception of other actors’ role; and the (possible) contribution of government, experts and laypeople to the debate. Interviews were halfstructured, meaning that the interview protocol was only used as a broad guideline (ensuring that at least the same questions were posed in each interview), and not as a scenario that should be followed slavishly. For instance, particular interview subjects could be developed in more detail in the course of an interview, or the order of questioning could be adapted to the line of argumentation of a particular respondent.

512

Annex 1

Question

Comment

1.

Focuses directly on the preferences of the organisation.

How is your organisation involved in the issue of ‘sustainable energy’? According to you, why does your organisation play an active role in this issue? What is its motivation? Which goals does it pursue?

[participants were given an introductory text about energy use in Belgium]

2.

a) b)

3.

a)

b)

Demand Side Management (DSM) is perceived by a large number of actors as an important option to include in a strategy for the sustainable provision of energy services (SPES). According to you, what are the most important opportunities and obstacles for this option on a short term? Is a limitation of the demand for primary energy a goal or a means to an end? According to you, which factors (cultural, social, and economic) will play an important role in the future demand for energy services? According to you, what are the most essential characteristics of a sustainable provision of energy services? Can you describe this in a few sentences? Indicate the most important criteria to evaluate a sustainable provision of energy services? How do these criteria relate to each other? According to you, what are the most important costs, effects and side-effects for Belgium of a policy aimed at : Production of electricity without nuclear energy (closure of nuclear reactors when they have reached the age of 40 years); Production of electricity with a constant and important contribution of nuclear production (5060 %); A maximal development of electricity production based on renewable energy sources (biomass, solar, wind, hydro, etc.); Production of electricity from fossil energy sources (coal, gas, oil); A maximal development of decentralised energy production (cogeneration, solar energy,...); An extension of the possibility to exchange electricity with neighbouring countries.

[participants were given introductory texts about the involvement of experts in energy policy (the AMPERE-report (2000) was used as an example) and about the involvement of laypersons (the large scale public consultation in the Netherlands was used as an example – see Stuurgroep maatschappelijke discussie energiebeleid (1983)) ]

4. a)

Science and energy policy - What are the main advantages and disadvantages of an involvement of experts in energy policy according to the approach of the AMPERE-commission? - What do you expect of the role of scientific expertise in an issue such as the sustainable

This question searches for a judgement on a (broadly defined) solution to a commonly perceived problem. Further questions seek to explore how the actor constructs the problem definition.

First questions inquire directly into the deeper preferences and values of the actor. Further specifications reveal the theories used by the actor in the evaluation of different (broadly defined) alternative scenarios.

Again, participants were confronted with some concrete approaches related to science and energy policy, government and energy policy & involvement of laypersons in energy policy (in order to discuss the institutional component of sustainable development). Further questions inquire into the 2nd order notions. Last question probes into the position of other actors, as the participant perceives it.


513

provision of energy services? Can the experts formulate decisive arguments? What is the role of scientific analysis? Government and energy policy - How do you judge the government decision to phase out nuclear energy in the perspective of sustainable development? - Which arguments does the government have to include in their reasoning? Laypersons and energy policy - Is an approach such as the Dutch societal discussion on energy policy useful in a Belgian context? - According to you, is this a good example of participation? What are the main advantages and disadvantages? - Do you think that laypersons are capable of sound judgement on a complicated subject? - If so, how can laypersons contribute to a problem such as the sustainable provision of energy services? Electricity producers and distributors - What is the role and responsibility of electricity producers and distributors with regard to their environment (social, economic, environmental, etc.)? According to you, which actors have had the largest impact on the shaping of the electricity sector such as we know it now? What are the goals of these actors?

-

b)

c)

d)

e)

[participants were given an introductory text about the management of high-level radioactive waste ]

5.

a) b) c)

6.

a)

b)

According to you, what are the main advantages and disadvantages of a geological disposal of HLW within the context of sustainable development? What are the options for the existing waste? What are the options for the prevention of the generation of future waste? According to you, what are the most relevant arguments to underpin your view on the management of HLW (ethics, economics, politics, science / technology, risk perception, society, etc.)? According to you, why has nuclear energy been such a controversial issue over the years? Why is nuclear energy irreconcilable with the value system of some other groups in society? From your point of view, can future designs for nuclear power plants contribute to a sustainable provision of energy services? Can public confidence be gained?

Participants were confronted with a concrete solution to the problem of HLW management (commonly perceived as a major challenge for the nuclear energy sector). Further questions inquire into the construction of the problem definition and the values at stake.

Questions the participant’s view and the view of other actors on the future of nuclear power.

Table A-2. Interview protocol

514

Annex 1

3. Data collection and analysis In general, data collection in a case-study research can happen in many different ways, as a researcher typically relies on multiple sources of evidence. In order to ensure that the derivation of the conclusions from the data seems logical to the reader, the researcher has to maintain a chain of evidence. In our research, we relied on several different information sources. Public information (documentation on the parliamentary debates on the nuclear phase-out proposal, the AMPERE report, academic literature, etc.) was used, together with information distributed (or published on the websites) by some of the stakeholders (leaflets, in-depth reports, newsletters, etc.), which was sometimes handed over during an interview session. Besides that, in a side project, we were also able to follow the evolution of the main actors’ argumentation in the debate over the years, based on archival records and interviews (see Laes et al. 2004c). Lastly, the interview sessions provided an occasion to ask questions on some topics (which were perhaps not covered in ‘official’ information spread by the organisation in question) in more depth. Throughout the research (not limited to the interview phase), data collection was alternated with data analysis. Schematically reconstructed, we can identify at least five moments of analysis:

1. In the ‘pre-interview’ phase, a document analysis gave a first impression of the possible

2.

3.

4.

5.

boundaries and themes within which the debate could move. These themes were translated into questions and used as a checklist for the handling of interview information; During the interview sessions, the interaction between the researcher and the interviewee assured that the analysis was anchored in the dialogue, e.g. by asking for clarifications in terms of the checklist items or asking ‘why’ questions; After each interview, the checklist was used in order to summarise the information obtained during the interview. Each interview was taped in order to enable structuring of the information in a personal interview summary form (these forms are on file with the author). Furthermore, such analysis gave the opportunity to identify further themes which should be taken up in subsequent interviews; A comparison and confrontation of the individual interview summaries in terms of the argumentations produced with regard to the checklist items enabled us to identify some common perspectives on the question of sustainable energy (and the role of nuclear power therein). These perspectives were analysed in a report (Laes and Meskens 2002b) and circulated to the interview participants for validation; The interview results were used for the construction of a list of criteria and indicators in order to judge the merits of different energy scenarios in light of overarching sustainability principles (through a multi-criteria exercise). In doing so, participants were confronted with other actors’ views, and the discussion could be oriented more explicitly towards the long-term perspective. Results of this exercise were also analysed and reported in the final chapter of this thesis, though validation by the participants in referred to a later stage.


515

4. Reliability, validity and limitations of the case-study approach Quantitative researchers often make derogatory comments on qualitative (case-study) research in the sense that it lacks criteria for judging the quality of the research set-up. The challenge to qualitative researchers is then to provide such criteria. According to Vandenabeele (1999), the overarching criterion for qualitative research is ‘intersubjective traceability’ – i.e. a researcher has the responsibility to document his activities and choices so that his/her colleagues can (in principle) reconstruct the research step by step. In the preceding sections, we have already explained our choices with regard to the selection of participants, the design of the interview protocol and strategy, and the method of analysis. This (we hope) enables readers to form an opinion on the steps in our research and to judge its strengths and weaknesses. The striving for ‘intersubjective traceability’ is generally made operational in two quality criteria, namely ‘reliability’ and ‘validity’. Reliability means that the influence of chance factors on the empirical results should be minimised. Reliability can be checked either by repeating the research (i.e. external reliability) and assuring that the results still stay the same, or by controlling the consistency of empirical results. Validity means that the research results should be a faithful representation of the ‘reality’ (in our case, a ‘pre-interpreted’ reality) studied. In general, quantitative research (e.g. surveys) has developed different technical measures for assessing reliability and validity (e.g. factor analysis, correlation coefficients, etc.). Such technical measures of course cannot be applied to qualitative research. However, this does not mean that qualitative (or interpretative) research is not scientific. For qualitative research, reliability and validity are made operational through other measures (Rowe and Frewer 2004).

4.1. Reliability and validity For qualitative research, reliability can be made operational through the concept of argumentative reliability. In the first chapter of the present thesis, we already described our research as an iterative process of constant to-and-fro reasoning between theoretical concepts and empirical observations on the way sustainable energy acquired its meaning in specific socio-political settings. Of course, this process makes reproduction by other researchers a very difficult matter. Hence the importance of argumentative reliability. By giving a clear and transparent representation of the chain of evidence and the general reasoning, readers can, at least in thought, follow the general argumentation. This argumentative reliability was enhanced in our research through a continuous dialogue with the research group STEM (University of Antwerp), specialised in analysing interactions between society and technology. This obliged us to make explicit the research choices and account for every new research step. The notes taken after meetings (meeting reports) as well as the preparatory notes and reports form a kind of logbook giving an idea of the different questions that arose during the research. In particular, we refer to the different preparatory versions of the interview protocol, letters of invitation (giving details on the project background) sent to participants, the individual interview forms and the summary report of the interviews (Laes and Meskens 2002b), entries in the scientific report of SCK-CEN (the institute which financed and guided this research project), and the evaluation report made by STEM (Keune et al. 2004) based also on feedback of participants in the research project. Furthermore, results were also communicated to the larger scientific community through presentations on international scientific congresses (Laes et al. 2002a; Laes and Meskens 2003) and publication in international peer-reviewed journals (Laes et al.2004a&b). Also contributing to argumentative reliability is the principle of triangulation we applied to our research:

516

Annex 1

interviews were combined with other data sources (archive material, published reports, leaflets, etc.) and other methods of analysis (multi-criteria mapping). With regard to validity, a first important element is problem coverage. In the selection of participants, we have attempted to include the relevant range of perspectives on the issue of sustainable energy (governance). Although not every participant could contribute to all research phases, we still attempted to include at least one representative of each category (unions, NGO’s, academia, etc.) in each research phase. Only in the scenario workshop this goal was not reached (no representatives of labour unions were present). A possible negative element affecting validity is the fact that the whole research project (starting from the interviews and ending with the multi-criteria exercise) took about two years to complete (spring 2002 – fall 2003). This is not so much a problem in the sense that it could be expected that actors would change their argumentation over this period (our historical analysis of the nuclear controversy has revealed a remarkable consistency of arguments over the years due to the very slow changes in energy policy), but the long time period involved made it more difficult for the participants to keep in mind the logic of the process (an oftrepeated critique in the evaluation carried out by STEM). Qualitative research can also be validated through communicative processes – i.e. feedback by the participants themselves. Interview participants were offered the opportunity to review the summary we made of each separate interview as well as the report covering all interview results. Feedback however was limited, possibly due to the busy schedules of the participants. Therefore, at the end of the entire research process, STEM researchers individually contacted all participants by telephone for feedback. At that time, results from the last phase (the multi-criteria mapping) were however not yet communicated to the participants in the exercise. Due to time limitations, this feedback is planned after the reporting of the findings to the research community (i.e. after finalisation of the PhD thesis). An oft-asked question with regard to validity is the extent to which the results obtained can be generalised to other contexts – i.e. a question pertaining external validity: ‘Based on the research results, can we make valid judgments regarding a reality outside of the case-study?’ Again, an important element in assuring the transferability of our research findings to other contexts consists in the choice of representative actors. As explained, these were selected on the basis of problem representativeness rather than statistical representativeness. All of these actors, due to their particular involvement in the FRDO, are looking for ways to position their organisation’s activities in a logic of sustainable development. Thus, they were the most suitable candidates for thematising and analysing the range of possibilities for how this positioning could take place. The analytical categories used to describe these different positioning perspectives result from our interpretation of key themes in the interviews. They consist of heterogeneous entities such as ‘the economic structure’, ‘opening up of electricity markets’, or even ‘the role of technology in society’, that are in a sense typical of the unfolding of the (nuclear) energy controversy. However, it would be delusive to maintain that the constitution of these categories is a process which takes place from scratch in each controversy. It uses as a resource categories which have been previously constituted (Cambrosio et al. 1991). For instance, a sociological analysis of the controversy surrounding the political decision to build nuclear power plants in the seventies (Leroy 1979; Laes et al. 2004c) shows a remarkable 1 consistency between the arguments used at that time by opponents of the nuclear option and most of the anti-nuclear arguments used during the interviews. Furthermore, we were to some extent able to make a check for validity with non-nuclear related exercises in such

1

Taking into account the changed policy context, i.e. the Kyoto obligations, liberalisation of energy markets, the decision to phase out nuclear power.


517

diverse fields as the long-term explorations of the economic impact of development on ecosystems (Van Asselt 2000), or reflections on visions in the area of agriculture and world food production that might “... serve as a ‘scale model’ for other technological visions…” (Grin et al. 2000, pp. 23-24). All of these categories were then connected in such way as to provide a more or less consistent argumentation scheme on sustainable energy and the role of nuclear power. In making this connection, we were guided by the different ways our respondents reflected on the meaning of a ‘sustainable energy system’. This sometimes proved to be a difficult exercise ‘on the spot’, but most participants were able to give a definition or to sum up important themes after some time of reflection. These responses, even if given by individuals, also reflected the wider ‘social worlds’ in which they are engaged. In particular, the typology used by Boltanski and Thévenot proved to be useful as a structuring principle for schematising the different points of view. From the interviews, we were thus able to deduce three distinct perspectives on the concept of sustainable (energy) development. As explained in chapter 1, our choice for Boltanski and Thévenot’s framework was not the point of departure for the analysis of interview results, but rather a result of the interview findings. Furthermore, it is not simply a matter of showing that Boltanski and Thévenot’s framework ‘applies’ to the (nuclear) energy debate, since this framework in itself is underdetermined (the framework cannot be used a priori for deciding which ‘objects’ will be mobilised to count as valid prove in the construction of a common world). Thus, the interview results provide the necessary background information for showing how the move from the ‘commonwealths’ to ‘common worlds’ is actually performed. In addition, the empirical findings also clearly reflected some of the wider key themes of ‘risk society’ as analysed by Beck. From the interviews and our analysis of the policy development cycle in the case of the Belgian nuclear phase out, it became clear that actors in the debate often ‘wrestled’ with the problem that ‘classical’ solutions – e.g. mechanisms of social consultation, the role of experts in defining consensual energy scenarios, political lobbying, regulation in protected market settings, etc. – more and more prove to be ineffective in a ‘risk society’. In a sense, most of the actors participating in our case-study still had one leg in the ‘classical’ world, while also seeking to connect with the new realities of ‘risk society’.

4.2. Limitations It is useful to contrast our interview approach with a socio-political analysis in the classic sense: we do not systematically connect argumentations with the societal actors actually involved in the nuclear debate or explain why this particular actor adheres to this particular socio-political vision; we do not trace the evolution of the debate over time; we do not analyse how and why certain aspects of the debate lose their ‘controversial status’ or how new elements enter the controversy in the course of time 2 . Certain inherent features limit the usefulness in a socio-political sense: 1.

2

We explicitly asked our respondents to reveal their personal opinion. Thus we were able to collect a more profound information than a simple restatement of the official discourse employed by the participant’s organisation. As a consequence, the responses cannot be interpreted as directly representative of a societal interest;

This type of socio-political argumentation analysis is undertaken in Laes et al. (2004c). Of course, this type of analysis can be seen as complementary to the interview approach, giving more background on the origin of certain argumentation patterns.

518

Annex 1

2.

3.

The identified categories and dimensions are of a cognitive nature: they are meant to structure an argumentation. They do not describe how the actors are actually implied in the controversy (actors will generally adapt their communication strategy to an implied audience, e.g. communication in the media will differ from communication in an advisory council, etc.) – a limitation from a socio-political point of view. In particular, for the purpose of socio-political analysis, the positioning of the media and political parties should have been taken up in the analysis (as both provide important mediating channels between the societal debate and government initiatives); The identified categories and dimensions are (to a certain extent) a reconstruction made by the analysts – again a disadvantage from a socio-political point of view: it is unclear which dimensions will gain or lose importance in the future.

However, these limitations prove to be acceptable for our specific targets. The point is that we try to surface some points regarding the content of the debate. We aimed to identify the boundaries within which the debate on sustainable development and nuclear power might evolve in the future, without actually making predictions on the forcefulness of these arguments. The categories cannot be used to predict the level of support or rejection of nuclear power or to predict how certain actors will be involved in the controversy (e.g. which communication strategy they will follow); but rather, to present the arguments that could be used to claim support or dejection. We chose this approach because it provides a stimulating basis to discuss the long-term perspective, e.g. by using different perspectives as the basis for scenario development.

ANNEX 2 – RELATIVE (BASE YEAR 1990) ENERGY DEMAND ACTIVITY LEVELS (BASE CASE + REDUCED DEMAND) Demand Cement kton/year Construction Materials - other Glas - flat Glas - hollow Glas - insulation Lime - limestone Elektrische motoren vermogenklasse A Elektrische motoren vermogenklasse B Elektrische motoren vermogenklasse C Elektrische motoren vermogenklasse D Industry intersect.- Electrothermal Industry - Lighting Chemical Industry - Ammonia Synthesis Chemical Industry - Steam (interm.temp) Chemical Industry - Process Heat (high temp) Chemical Industry - Chlorine Elec. Iron & Steel Production Iron & Steel Procession Other industry low temp. steam & heat Other Industry Residential - open Residential - urban (halfopen and closed) Residential - flats Small service sector Large service sector Warm water (R1, open) Warm water (R2, halfopen and closed) Warm water (R3, flats) Warm water (small service sector) Warm water (large service sector) Food preparation (resident.& service sector) Electricity Use Residential Electricity Use Small Service Sector Electricity Use Large Service Sector Public Lighting Short Distance Transport by Car Long Distance Transport by Car Transport by Bus Transport by Truck Transport by Rail, Passenger Transport by Rail, Goods Transport by Ship Aviation Bunker Marine Bunker

1990

2000

2010

2030

2050

100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100

133 133 133 133 132 133 121 121 121 121 130 106 142 130 132 109 102 109 133 133 107 108 112 115 143 107 112 119 128 129 114 120 151 151 117 121 124 105 120 104 127 115 122 111

138 139 139 139 137 139 133 133 133 133 143 117 149 136 138 114 104 111 145 145 112 113 123 126 156 110 122 136 141 141 125 145 171 171 145 150 164 118 141 124 149 132 161 111

150 150 150 150 149 150 157 157 157 157 167 137 166 153 155 128 106 114 173 173 126 127 147 152 187 124 144 174 170 171 152 217 219 219 225 197 265 140 187 175 201 173 267 111

165 161 164 165 163 165 181 181 181 181 190 160 180 169 172 143 106 114 205 205 126 127 147 180 218 124 144 174 200 202 179 294 270 270 316 197 265 162 237 230 256 217 385 111

520

Annex 2

Demand Cement kton/year Construction Materials - other Glas - flat kton/year Glas - hollow kton /year Glas - insulation kton/year Lime - limestone kton/year Elektrische motoren vermogenklasse A Elektrische motoren vermogenklasse B Elektrische motoren vermogenklasse C Elektrische motoren vermogenklasse D Industry intersect.- Electrothermal Industry - Lighting Chemical Industry - Ammonia Synthesis Chemical Industry - Steam (interm.temp) Chemical Industry - Process Heat (high temp) Chemical Industry - Chlorine Elec. Iron & Steel Production : kton/year Iron & Steel Procession Other industry low temp. steam & heat Other Industry Residential - open Residential - urban (halfopen and closed) Residential - flats Small service sector Large service sector Warm water (R1, open) Warm water (R2, halfopen and closed) Warm water (R3, flats) Warm water (small service sector) Warm water (large service sector) Food preparation (resident.& service sector) Electricity Use Residential Electricity Use Small Service Sector Electricity Use Large Service Sector Public Lighting Short Distance Transport by Car Long Distance Transport by Car Transport by Bus Transport by Truck Transport by Rail, Passenger Transport by Rail, Goods Transport by Ship Aviation Bunker Marine Bunker

1990

2000

2010

2030

2050

100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100

103 102 103 103 102 103 104 104 104 104 106 102 104 103 103 100 96 97 107 107 107 108 112 115 143 107 112 119 128 129 114 120 151 151 117 106 114 102 111 110 113 109 122 111

105 104 105 105 105 105 109 109 109 109 111 104 109 106 107 100 91 93 115 114 112 113 123 126 156 110 122 136 141 141 125 145 171 171 145 113 129 104 122 120 126 117 161 111

110 108 110 110 109 110 118 118 118 118 122 108 117 112 114 100 83 86 129 129 126 127 147 152 187 124 144 174 170 171 152 217 219 219 225 125 157 109 144 141 153 135 267 111

116 113 115 116 114 115 127 126 127 127 133 112 126 118 120 100 74 80 144 143 126 127 147 180 218 124 144 174 200 202 179 294 270 270 316 138 186 113 166 161 179 152 385 111

ANNEX 3 – MARKAL DATA OF POWER GENERATION OPTIONS The following tables contain technical-economic characteristic data of power generation options used in the MARKAL database. Not all generation options are included, only the key technologies.

Centralised technologies Pulverised Coal Power Plant (ASC/USC) Life [year] Start year [year] Annual availability [%] Efficiency [%] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJe]

1990

Pulverised Coal Power Plant (ASC/USC) de-CO2 Life [year] 30 Start year [year] 2010 Annual availability [%] Efficiency [%] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJe]

1990

Coal Gasification Combined Cycle (IGCC) Life [year] Start year [year] Annual availability [%] Efficiency [%] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJe]

1990

PWR Nuclear Power Plant Life Start year Annual availability Material input Material output Investment bound : up Residual installed capacity Investment cost Annual fixed O&M cost Annual variable O&M cost

2000

2010

2020

2030

2040

2050

84 47.0 1054 32.72 0.69

84 52.0 1044 32.72 0.69

84 52.0 1034 32.72 0.69

84 52.0 1023 32.72 0.69

84 52.0 1013 32.72 0.69

2010

2020

2030

2040

2050

84 42.6 1718 45.81 38.92

84 47.1 1701 45.81 38.92

84 47.1 1685 45.81 38.92

84 47.1 1668 45.81 38.92

84 47.1 1652 45.81 38.92

2010

2020

2030

2040

2050

83 41.0 1358 42.14 0.48

83 44.2 1287 40.65 0.41

83 48.0 1216 40.65 0.41

83 52.5 1145 40.65 0.41

83 52.5 1145 40.65 0.41

2000

2010

2020

2030

2040

2050

85 0.6442 0.6442 0 4.613 1922 36.94 0.50

85 0.6442 0.6442 0 5.713 1801 36.94 0.50

85 0.6442 0.6442 0 3.88 1679 36.94 0.50

85 0.6442 0.6442 0

85 0.6442 0.6442 0

85 0.6442 0.6442 0

1557 36.94 0.50

1436 36.94 0.50

1314 36.94 0.50

30 2010

2000

2000

25 2010

1990 [year] 40 [year] 1990 [%] 85 [ton/PJe] UROX1 0.6442 [ton/PJe] UROXS 0.6442 [GWe] 0 [GWe] 4.613 [EURO/kWe] 1983 [EURO/kWe] 36.94 [EURO/GJe] 0.50

522

Annex 3

PWR Nuclear Power Plant - new First core load : UROX1 Life Start year Annual availability Material input Material output Investment cost Annual fixed O&M cost Annual variable O&M cost

[ton/GWe] 43 [year] 60 [year] 2010 [%] [ton/PJe] UROX1 [ton/PJe] UROXS1 [EURO/kWe] [EURO/kWe] [EURO/GJe]

MHTGR Nuclear Power Plant First core load : UROX2 Life Start year Annual availability Material input Material output Investment cost Annual fixed O&M cost

[ton/GWe] 28 [year] 30 [year] 2010 [%] [ton/PJe] UROX2 [ton/PJe] UROXS2 [EURO/kWe] [EURO/kWe]

STAG Power Plant Life Start year Annual availability Efficiency Residual installed capacity Investment cost Annual fixed O&M cost Annual variable O&M cost

[year] [year] [%] [%] [GWe] [EURO/kWe] [EURO/kWe] [EURO/GJ]

20 1990

STAG Power Plant de-CO2 Life Start year Annual availability Efficiency Investment cost Annual fixed O&M cost Annual variable O&M cost

[year] [year] [%] [%] [EURO/kWe] [EURO/kWe] [EURO/GJ]

20 2010

Hydro pumped storage Life Start year Annual availability Efficiency Bound on capacity : up Residual installed capacity Investment cost Annual fixed O&M cost

[year] [year] [%] [%] [GWe] [GWe] [EURO/kWe] [EURO/kWe]

40 1990

1990

1990

2000

2000

2010

2020

2030

2040

2050

85 0.534 0.534 1801 36.94 0.50

85 0.534 0.534 1679 36.94 0.50

85 0.534 0.534 1557 36.94 0.50

85 0.534 0.534 1436 36.94 0.50

85 0.534 0.534 1314 36.94 0.50

2010

2020

2030

2040

2050

85 0.3446 0.3446 2452 52.78

85 0.3446 0.3446 1935 52.78

85 0.3446 0.3446 1419 52.78

85 0.3446 0.3446 1419 52.78

85 0.3446 0.3446 1419 52.78

1990

2000

2010

2020

2030

2040

2050

85.5 55.0 0.158 620 24.54 0.41

85.5 58.0 0.158 410 24.54 0.41

85.5 60.2 0 410 24.54 0.41

85.5 62.5 0 410 24.54 0.41

85.5 65.0 0 410 24.54 0.41

85.5 65.0 0 410 24.54 0.41

85.5 65.0 0 410 24.54 0.41

1990

2000

2010

2020

2030

2040

2050

85.5 54.5 893 53.50 0.90

85.5 56.6 893 53.50 0.90

85.5 58.9 893 53.50 0.90

85.5 58.9 893 53.50 0.90

85.5 58.9 893 53.50 0.90

1990

2000

2010

2020

2030

2040

2050

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

92 75.8 1.304 0.03 1236 39.54

MARKAL data of power generation options 523

Hydro Power Plant Life Start year Annual availability Bound on capacity : up Residual installed capacity Investment cost Annual variable O&M cost

[year] [year] [%] [GWe] [GWe] [EURO/kWe] [EURO/GJ]

Wind turbine onshore, seaside Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Bound on capacity : up [GWe] Residual installed capacity [GWe] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe]

Wind turbine onshore, polders Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Bound on capacity : up [GWe] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Wind turbine offshore Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Bound on capacity : up [GWe] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe]

1990

2000

2010

2020

2030

2040

2050

95 0.11 0.0867 2471 3.47

95 0.11 0.0868 1983 3.47

95 0.11 0.0868 1983 3.47

95 0.11 0.0868 1983 3.47

95 0.11 0.0868 1983 3.47

95 0.11 0.0868 1983 3.47

95 0.11 0.0868 1983 3.47

1990

2000

2010

2020

2030

2040

2050

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0.0042 1804 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0.0042 811 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0 811 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0 811 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0 811 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0 811 29.65

28.0 24.6 28.0 24.6 34.1 30.3 0.15 0 811 29.65

1990

2000

2010

2020

2030

2040

2050

19.37 17.01 19.37 17.01 23.62 20.91 0.25 1804 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

0.19 0.17 0.19 0.17 0.24 0.21 0.25 811 29.65

1990

2000

2010

2020

2030

2040

2050

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

37.7 33.1 37.7 33.1 45.8 40.8 1 1419 37.80

40 1990

25 1990 29.7

25 1990 20.5

25 2000 40

524

Annex 3

Wind turbine onshore, inland Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Bound on capacity : up [GWe] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Photovoltaics Life Start year Availability

[year] [year] [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Bound on activity : up [PJe] Investment cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] Wood gasification STAG Life Start year Annual availability Efficiency Investment cost Annual fixed O&M cost Annual variable O&M cost

1990

2000

2010

2020

2030

2040

2050

16.1 14.2 16.1 14.2 19.7 17.4 0 1804 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

16.1 14.2 16.1 14.2 19.7 17.4 0.1 811 29.65

1990

2000

2010

2020

2030

2040

2050

20.0 0.0 30.4 0.0 8.8 0.0 36 5000 20.50

20.0 0.0 30.4 0.0 8.8 0.0 36 1981 14.80

20.0 0.0 30.4 0.0 8.8 0.0 36 1525 11.45

20.0 0.0 30.4 0.0 8.8 0.0 36 1071 8.10

20.0 0.0 30.4 0.0 8.8 0.0 36 930 7.60

20.0 0.0 30.4 0.0 8.8 0.0 36 930 7.60

2000

2010

2020

2030

2040

2050

80.0 37.8 1140 51.31 0.54

81.7 39.9 1140 51.31 0.54

83.3 42.2 1140 51.31 0.54

85.0 44.8 1140 51.31 0.54

85.0 47.7 1140 51.31 0.54

85.0 51.0 1140 51.31 0.54

2000

2010

2020

2030

2040

2050

83 55.1 1411 40.90 3.04

83 56.3 1392 40.90 3.04

83 57.5 1373 40.90 3.04

25 1990 17.1 -

30 2000 10.6

1990 [year] [year] [%] [%] [EURO/kWe] [EURO/kWe] [EURO/GJ]

Fuel cell (SOFC) integrated coal gasification Life [year] Start year [year] Annual availability [%] Efficiency [%] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

25 2000

1990 25 2030


Fuel cell (SOFC) ICG de-CO2 Life Start year Annual availability Efficiency Investment cost Annual fixed O&M cost Annual variable O&M cost

[year] [year] [%] [%] [EURO/kWe] [EURO/kWe] [EURO/GJ]

25 2030

Municipal Waste Incinerator Life Start year Annual availability Efficiency Bound on activity : up Investment cost Annual fixed O&M cost Annual variable O&M cost

[year] [year] [%] [%] [PJe] [EURO/kWe] [EURO/kWe] [EURO/GJ]

30 1990

1990

2000

2010

2020

2030

2040

2050

83 51.8 1740 56.92 3.04

83 52.9 1725 56.92 3.04

83 54.0 1710 56.92 3.04

1990

2000

2010

2020

2030

2040

2050

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

84 30.0 3.6 672 31.58 0.48

1990

2000

2010

2020

2030

2040

2050

80

80

80

80

80

80

80

23.0 50.7 2.2 914 4.09 1.54

24.1 53.0 2.2 914 4.09 1.54

25.3 55.6 2.2 914 4.09 1.54

26.6 58.5 2.2 914 4.09 1.54

28.0 61.6 2.2 914 4.09 1.54

28.0 61.6 2.2 914 4.09 1.54

28.0 61.6 2.2 914 4.09 1.54

1990

2000

2010

2020

2030

2040

2050

80

80

80

80

80

80

38.5 38.5 1 684 26.03

39.0 39.0 1 684 26.03

39.5 39.5 1 684 26.03

40.0 40.0 1 684 26.03

40.5 40.5 1 684 26.03

41.1 41.1 1 684 26.03

Decentralised technologies

Gas Turbine (4 MWe) PSH Life Start year Annual availability Efficiency Electric Thermal Heat to power ratio Investment cost Annual fixed O&M cost Annual variable O&M cost Gas turbine (40 MWe) PSH Life Start year Annual availability Efficiency Electric Thermal Heat to power ratio Investment cost Annual fixed O&M cost

[year] [year] [%]

15 1990

[%] [%] [-] [EURO/kWe] [EURO/kWe] [EURO/GJ]

[year] [year] [%] [%] [%] [-] [EURO/kWe] [EURO/kWe]

20 1995

526

Annex 3

Gas Engine (1 MW) PSL Life [year] Start year [year] Annual availability [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] Gas Engine (650 kW) TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] BTF Gasification gas engine TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

50

50

50

50

50

50

50

34.1 51.2 1.5 446 4.46 2.60

36.0 54.0 1.5 446 4.46 2.60

37.5 56.3 1.5 446 4.46 2.60

38.0 57.0 1.5 446 4.46 2.60

38.0 57.0 1.5 446 4.46 2.60

38.0 57.0 1.5 446 4.46 2.60

38.0 57.0 1.5 446 4.46 2.60

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

33.0 52.8 1.6 527 5.21 2.75

33.7 53.9 1.6 527 5.21 2.75

34.4 55.1 1.6 527 5.21 2.75

35.2 56.3 1.6 527 5.21 2.75

36.0 57.6 1.6 527 5.21 2.75

36.0 57.6 1.6 527 5.21 2.75

36.0 57.6 1.6 527 5.21 2.75

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

22.3 40.2 1.8 1696 157.86 1.25

23.1 41.6 1.8 1616 140.41 1.11

24.0 43.1 1.8 1537 122.96 0.97

24.9 44.8 1.8 1458 105.50 0.84

25.8 46.5 1.8 1378 88.05 0.70

15 1990

20 1995 50

15 2005 50


Wood Gasification gas engine TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] Diesel Engine TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

Fuel cells (MC/SO/PA) hydrogen PSH Life [year] Start year [year] Annual availability [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.0 45.0 1.8 1537 115.77 0.48

25.5 45.8 1.8 1537 115.77 0.48

25.9 46.7 1.8 1537 115.77 0.48

26.5 47.6 1.8 1537 115.77 0.48

27.0 48.6 1.8 1537 115.77 0.48

27.6 49.6 1.8 1537 115.77 0.48

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

37.0 42.6 1.15 404 4.04 1.24

38.2 43.9 1.15 404 4.04 1.24

39.4 45.3 1.15 404 4.04 1.24

40.7 46.7 1.15 404 4.04 1.24

42.0 48.3 1.15 404 4.04 1.24

42.0 48.3 1.15 404 4.04 1.24

42.0 48.3 1.15 404 4.04 1.24

1990

2000

2010

2020

2030

2040

2050

80

80

80

80

53.3 28.8 0.54 1185 32.90 3.04

54.0 29.2 0.54 1185 32.90 3.04

54.7 29.5 0.54 1185 32.90 3.04

55.4 29.9 0.54 1185 32.90 3.04

15 2000 50

20 1990 50

10 2015

528

Annex 3

Fuel cells (MC/SO/PA) hydrogen TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] Fuel cells (MC/SO/PA) hydrogen TER2 Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

Fuel cells (MC/SO/PA) natural gas PSH Life [year] Start year [year] Annual availability [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

51.4 27.7 0.54 1334 37.06 2.58

52.1 28.2 0.54 1334 37.06 2.58

52.9 28.6 0.54 1334 37.06 2.58

53.7 29.0 0.54 1334 37.06 2.58

54.5 29.4 0.54 1334 37.06 2.58

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

51.4 27.7 0.54 1334 37.06 2.58

52.1 28.2 0.54 1334 37.06 2.58

52.9 28.6 0.54 1334 37.06 2.58

53.7 29.0 0.54 1334 37.06 2.58

54.5 29.4 0.54 1334 37.06 2.58

2010

2020

2030

2040

2050

80

80

80

80

50.3 27.7 0.55 1212 0.82 3.04

51.0 28.1 0.55 1212 0.82 3.04

51.7 28.4 0.55 1212 0.82 3.04

52.4 28.8 0.55 1212 0.82 3.04

10 2005 50

1990

2000

10 2005 50

1990

2000

10 2015


Fuel cells (MC/SO/PA) natural gas TER Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ] Fuel cells (MC/SO/PA) natural gas TER2 Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

40.0 20.0 0.5 1363 37.06 2.58

40.0 20.0 0.5 1363 37.06 2.58

40.0 20.0 0.5 1363 37.06 2.58

40.0 20.0 0.5 1363 37.06 2.58

40.0 20.0 0.5 1363 37.06 2.58

2010

2020

2030

2040

2050

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

25.5 13.1 5.4 1.5 90.0 78.2

40 20 0.5 1363 37.06 2.58

40 20 0.5 1363 37.06 2.58

40 20 0.5 1363 37.06 2.58

40 20 0.5 1363 37.06 2.58

40 20 0.5 1363 37.06 2.58

10 2005 50

1990

2000

10 2005 50

530

Annex 3

Combined heat and power with heat to grid

STAG for cogeneration LTH Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] STAG for cogeneration LTH de-CO2 Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Fuel cells cogeneration plant (hydrogen) LTH Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

43.8 33.1 0.76 684 27.27

45.1 34.2 0.76 684 27.27

46.5 35.3 0.76 684 27.27

48.1 36.4 0.76 684 27.27

48.1 36.4 0.76 684 27.27

48.1 36.4 0.76 684 27.27

2000

2010

2020

2030

2040

2050

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

40.9 31.0 0.76 1492 59.44

42.2 32.0 0.76 1492 59.44

43.6 33.0 0.76 1492 59.44

43.6 33.0 0.76 1492 59.44

43.6 33.0 0.76 1492 59.44

2010

2020

2030

2040

2050

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.1 29.2 0.54 1190 33.04 1.67

54.9 29.7 0.54 1190 33.04 1.67

55.9 30.2 0.54 1190 33.04 1.67

56.8 30.7 0.54 1190 33.04 1.67

20 1995 50

1990 20 2010 50

1990

2000

15 2015 50


Fuel cells cogeneration plant (gas) LTH Life [year] Start year [year] Availability [%] Intermediate day [%] Intermediate night [%] Summer day [%] Summer night [%] Winter day [%] Winter night [%] Efficiency Electric [%] Thermal [%] Heat to power ratio [-] Investment cost [EURO/kWe] Annual fixed O&M cost [EURO/kWe] Annual variable O&M cost [EURO/GJ]

1990

2000

2010

2020

2030

2040

2050

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

54.0 23.0 22.0 0.0 90.0 51.0

52.0

55.2

55.2

55.2

0.55 1190 33.04 1.67

0.55 1190 33.04 1.67

0.55 1190 33.04 1.67

0.55 1190 33.04 1.67

15 2015 50

ANNEX 4 – A TECHNICAL NOTE ON THE MULTI-CRITERIA MAPPING METHODOLOGY The technique employed in chapter 7 is a ‘heuristic multi-criteria mapping’ (MCM) exercise. This approach was developed by Stirling and Mayer (1999, 2000) and used for the first time in the context of genetically modified foodstuff. This approach has since then been applied in a variety of sectors, including public health, energy, radioactive waste, environmental remediation and agricultural strategies. The present annex draws heavily upon the methodological explanations as found in Stirling and Mayer (1999). The technique developed by Stirling and Mayer is called heuristic because the principal aim is to explore issues and come to a better understanding of the way problems are structured through discourse, rather than to find a definite single ‘optimal’ solution. It is a mapping exercise because the results are expressed systematically in terms of the perspectives taken by different participants, with conclusions being drawn only conditionally with regard to these perspectives. In both these respects, MCM differs from other multi-criteria methods which rely on more complex techniques in an attempt to identify a unique and ‘optimal’ resolution to a certain policy problem. The MCM approach used in chapter 7 relies upon a linear additive weighting aggregation model, which is based on a simple weighted average of option performance:

ri = ∑c sic ⋅ wc In other words, this equation tells us that the overall performance rank obtained for the ith policy option (ri) is found as the sum of the performance scores determined for that option under the cth appraisal criterion (sic) each multiplied by the importance weighting on that criterion (wc). The scores are normalised so that

sic = (mic − mc ,min ) /(mc ,max − mc ,min ) In other words, this equation tells us that the performance score for the ith policy option under the cth appraisal criterion (sic) is found as the ratio of the difference between the performance measure determined for that option (mic) and that for the lowest-performing option (mc,min) with the difference between the performance measures determined for the highest (mc,max) and lowest (mc,min) performing options under that criterion. A non-linear scale might apply where there is some threshold value or non-linear function for the relationship between the range of scores and the performance of the scenarios. This aggregation method represents one of the simplest of all theoretically valid approaches. Following Stirling and Mayer (1999, 2000), we adopted this method out of a concern for not allowing the quantification procedure to obscure the (for the time being) more important qualitative features of the appraisal. This point of view is furthermore strengthened in a formal sense by Arrow’s impossibility theorem (stating that there can be no formal solution to the problem of finding a definitive social preference ordering), mirrored in a sociological sense by Boltanski and Thévenot’s observation that there is no overarching ‘meta-commonwealth’ able to transcend conflicting views on which commonwealth logic applies to a situation at hand. It therefore remains

534

Annex 4

an open question, and therefore fundamentally a matter of reasoned judgment, whether the loss of simplicity and transparency in the more complex applications of multi-criteria aggregation is worth the sometimes marginal improvements in fidelity. For instance, one main difference with ‘traditional’ multi-criteria exercises is that we have not used our combined value tree as a basis for the aggregation of different stakeholder views into one ‘collective’ opinion. Instead, the combined value tree was used in our case as a ‘probe’, which enabled us to reveal the important differences between the ways of framing apparently similar criteria from different stakeholder perspectives. In fact, these different ways of framing are an important outcome of the exercise in its own right. Likewise, the assigning of ‘technical performance scores’ by the participants themselves for some of the criteria used in our case is also a point of contrast with many multi-criteria exercises, which often rely on a separate panel of specialists to determine a single set of scores under each criterion. One reason for this approach is that most participants in this exercise were themselves professionals with respect to different aspects of the broad energy policy field. The more important reason however is that each participant, due to the different problem framings, generated a (somewhat) different scoring scheme. The choice of one particular scoring scheme for a particular criterion (i.e. the choice of a ‘spokesperson’ for a particular ‘actant’ in Latourian parlance) would thus clearly be a point of contention in its own right. This question can thus only be addressed after a clear picture is drawn up on the different perspectives. However, in contrast with Stirling and Mayer (1999, 2000), we have chosen to represent particular scores by physical metrics or established indicators. Leaving everything open to ‘arbitrary’ rating scales (as Stirling and Mayer propose to do) in our view concedes too much to the ‘free framing’ of the problem and is unlikely to lead to the establishment of at least some common ground. However, important framing assumptions such as discounting of future impacts were left open to the participants’ discretion (by explicitly leaving open the possibility to value mid-term and long-term impacts separately). Furthermore, the open individual interview approach adapted in our exercise left open the possibility for contesting the quantitative measures employed (and/or suggesting others). Another important point to consider in multi-criteria assessments concerns weightings. The simple scalar weightings used in the MCM approach do not seek to model any non-linearities which there may be in the relationship between the performance measures and subjective values. These factors are sometimes addressed formally through the use of so-called ‘value functions’. However, these ‘complex’ relationships are addressed in the MCM practice by the unconstrained and reflexive character of the weighting procedure (with respect to the rankings of the individual options). Participants in the exercise could at any time revise the weightings they attributed to lower-level criteria on the basis of higher-level criteria or even the graphic representation of the different rankings. In our exercise, weightings were determined by participants on a ‘holistic’ appreciation of the relationship between all criteria. The only ‘technical’ aid employed was the recalculation of the weighting values as they were entered and their display in percentage terms. This was done in order to fit an ‘intuitive’ description of 100 ‘importance points’ that had to be allocated between the criteria in question. Again, more complex or elaborate procedures exist and are applied in the context of energy policy, e.g. ‘swing weighting’, ‘analytic hierarchy’ (see e.g. Hämäläinen 1990) or ‘electre’ methods (see e.g. Haldi et al. 2002) – for an overview, see Hobbs and Meier (2000, pp. 7599). These methods systematically build up an overall weighting scheme on the basis of pair wise trade-offs between criteria. However, this is again achieved only at the expense of a great increase in complexity. Thus the rationale for the more straightforward approach adopted in MCM rests on

A technical note on the Multi-Criteria Mapping methodology

535

the qualified role of the weightings (compared with framing assumptions) and the importance of iterative and reflexive deliberation over rankings. Finally, the treatment of uncertainties (which is also a key methodological variable in multi-criteria assessment) is done by means of a deterministic sensitivity analysis in MCM – i.e. by reference to an ‘optimistic’ and a ‘pessimistic’ score for a valuation criterion. Other possibilities include the stochastic approach employed in ‘utility functions’. However, as Stirling and Mayer (1999) point out, the information requirements of such approaches are potentially enormous and always subject to queries over the applicability of the chosen statistically or theoretically-derived probabilities. Moreover, probabilistic methods cannot offer a valid means to characterise conditions of strict uncertainty and ignorance, which possibly dominate many of the aspects of the future performance of energy systems in hypothetical scenario exercises. All in all, we believe that Stirling and Mayer (1999, 2000) have succeeded in giving a convincing answer to some of the most common criticisms directed at the use of multi-criteria methods in policy option appraisals, e.g. the lack of a well-defined procedure for criteria choice, the potential for gaps and overlaps between criteria, the intrinsic subjectivity of weighting assumptions, etc. Stirling and Mayer’s answer to these critiques is that most of these points should be regarded as advantages rather than shortcomings in the case of MCM applied in the context of ‘complex’ socio-technical questions (surrounded by a diversity of existing views and difficult political, social and scientific factors). The reason is that MCM features such as the lack of constraint on the type of criteria that can be included, the openness to different weighting schemes and the ability to combine quantitative and qualitative factors allows a careful deliberation and open discussion of particular perspectives in appraisal, rather than concealing these choices in the presentation of ‘definite’ results of risk or costbenefit analyses.

ANNEX 5 – SELECTION OF ATTRIBUTES (BOLD) AND SCORES FOR EACH PARTICIPANT IN THE MCM EXERCISE

538

Annex 5

Selection of attributes and scores for each participant in the MCM exercise 539

540

Annex 5


542

Annex 5


544

Annex 5


546

Annex 5

ANNEX 6 – SENSITIVITY TESTS ON MCM WEIGHTINGS Weighting of ‘Environmental and Human Health & Safety’ divided by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

LCS

LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

LCS

2.

P.LCS.I

P.LCS.I

P.CS

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

P.LCS

3.

P.CS

P.LCS

P.LCS

P.CS

P.LCS.I

P.LCS

P.LCS.I

P.CS

P.CS

P.LCS.I

4.

LCS

P.CS

P.LCS.I

LCS

P.CS

P.CS

LCS

P.LCS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.CS

LCS

P.LCS

P.LCS

2.

P.LCS.I

LCS

LCS

P.LCS

P.LCS

LCS

P.LCS

P.CS

P.LCS.I

LCS

3.

P.CS

P.LCS.I

P.CS

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS.I

P.CS

P.CS

4.

LCS

P.CS

P.LCS.I

LCS

P.LCS.I

P.CS

LCS

P.LCS

LCS

P.LCS.I

Market

Rational

Weighting of ‘Environmental and Human Health & Safety’ multiplied by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

LCS

P.LCS

P.LCS.I

P.LCS

P.LCS.I

P.CS

LCS

P.LCS

P.LCS

2.

P.LCS.I

P.LCS

P.LCS.I

P.LCS

LCS

LCS

P.LCS

P.LCS.I

P.CS

P.LCS.I

3.

P.CS

P.LCS.I

P.CS

LCS

P.CS

P.LCS

P.LCS.I

P.LCS

P.LCS.I

LCS

4.

LCS

P.CS

LCS

P.CS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS

LCS

LCS

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

LCS

P.LCS.I

P.LCS.I

P.LCS

P.LCS.I

P.CS

P.LCS.I

P.LCS.I

P.LCS.I

3.

P.CS

P.LCS.I

P.CS

LCS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

LCS

4.

LCS

P.CS

LCS

P.CS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

Market

Rational

548

Annex 6

Weighting of ‘Economic Welfare’ divided by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

1.

P.LCS.I

2.

P.LCS

P.LCS.I

P.LCS

P.LCS

P.LCS.I

3.

LCS

LCS

4.

P.CS

1.

F3

G1

G2

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

P.LCS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS.I

P.LCS.I

LCS

P.CS

P.CS

P.LCS

P.LCS.I

P.LCS

P.CS

LCS

P.CS

P.CS

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

P.LCS.I

P.LCS.I

P.LCS

P.LCS

P.LCS

P.CS

P.LCS.I

P.LCS.I

P.LCS.I

3.

P.CS

LCS

P.CS

LCS

P.CS

LCS

P.LCS.I

P.LCS

P.CS

LCS

4.

LCS

P.CS

LCS

P.CS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

Market

Rational

Weighting of ‘Economic Welfare’ multiplied by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

LCS

LCS

P.LCS.I

LCS

LCS

P.CS

LCS

P.LCS

LCS

2.

P.CS

P.LCS.I

P.CS

P.LCS

P.LCS.I

P.LCS.I

P.LCS

P.LCS.I

P.LCS.I

P.LCS.I

3.

P.LCS.I

P.LCS

P.LCS

P.CS

P.LCS

P.LCS

P.LCS.I

P.CS

P.CS

P.LCS

4.

LCS

P.CS

P.LCS.I

LCS

P.CS

P.CS

LCS

P.LCS

LCS

P.CS

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

LCS

P.CS

LCS

P.LCS

P.LCS

Market

Rational 1. 2.

P.CS

LCS

P.CS

P.LCS

P.CS

P.LCS.I

P.LCS

P.CS

P.LCS.I

LCS

3.

P.LCS.I

P.LCS.I

LCS

P.CS

P.LCS

P.LCS

P.LCS.I

P.LCS.I

P.CS

P.CS

4.

LCS

P.CS

P.LCS.I

LCS

P.LCS.I

P.CS

LCS

P.LCS

LCS

P.LCS.I

Sensitivity tests on MCM weightings

549

Weighting of ‘Social, Political, Cultural and Ethical Needs’ divided by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

Market 1.

P.LCS

LCS

P.LCS

P.LCS.I

LCS

LCS

P.CS

LCS

P.LCS

P.LCS.I

2.

P.LCS.I

P.LCS.I

P.LCS.I

P.LCS

P.CS

P.LCS.I

P.LCS

P.LCS.I

P.LCS.I

P.LCS

3.

P.CS

P.LCS

LCS

P.CS

P.LCS.I

P.LCS

P.LCS.I

P.CS

P.CS

LCS

4.

LCS

P.CS

P.CS

LCS

P.LCS

P.CS

LCS

P.LCS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS

LCS

LCS

P.CS

LCS

P.LCS

P.LCS

2.

P.LCS.I

LCS

P.LCS.I

P.LCS.I

P.LCS

P.LCS.I

P.LCS

P.LCS.I

P.LCS.I

LCS

3.

P.CS

P.LCS.I

P.CS

P.CS

P.CS

P.LCS

P.LCS.I

P.CS

P.CS

P.LCS.I

4.

LCS

P.CS

LCS

LCS

P.LCS.I

P.CS

LCS

P.LCS

LCS

P.CS

Rational

Weighting of ‘Social, Political, Cultural and Ethical Needs’ multiplied by 3 Ranking

A1

A2

C1

D1

D2

E1

F2

F3

G1

G2

1.

P.LCS

P.LCS

LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

P.LCS.I

P.LCS

P.LCS

P.LCS

P.LCS

P.CS

P.LCS.I

P.LCS.I

LCS

3.

P.CS

LCS

P.CS

P.CS

P.LCS.I

LCS

P.LCS.I

P.LCS

P.CS

P.LCS.I

4.

LCS

P.CS

P.LCS.I

LCS

P.CS

P.CS

LCS

P.CS

LCS

P.CS

1.

P.LCS

P.LCS

P.LCS

P.LCS.I

LCS

P.LCS.I

P.LCS

LCS

P.LCS

P.LCS

2.

P.LCS.I

LCS

LCS

P.LCS

P.LCS

P.LCS

P.LCS.I

P.LCS

P.LCS.I

LCS

3.

P.CS

P.LCS.I

P.CS

P.CS

P.CS

LCS

P.CS

P.LCS.I

P.CS

P.LCS.I

4.

LCS

P.CS

P.LCS.I

LCS

P.LCS.I

P.CS

LCS

P.CS

LCS

P.CS

Market

Rational

KATHOLIEKE UNIVERSITEIT LEUVEN FACULTEIT TOEGEPASTE WETENSCHAPPEN DEPARTEMENT WERKTUIGKUNDE AFDELING TOEGEPASTE MECHANICA EN ENERGIECONVERSIE Celestijnenlaan 300 A, 3001 Heverlee

NUCLEAR ENERGY AND SUSTAINABLE DEVELOPMENT Theoretical reflections and critical-interpretative research towards a better support for decision making

Jury : Prof. Dr. ir. P. Van Houtte, voorzitter Prof. Dr. ir. W. D’haeseleer, promotor Prof. Dr. ir. R. Weiler, promotor Prof. Dr. ir. R. Belmans Prof. Dr. ir. P. De Meester Prof. Dr. ir. B. De Moor Prof. Dr. B. Raymaekers Prof. Dr. P. Kroes (T.U.Delft) Dr. F. Hardeman (SCK•CEN)

Proefschrift voorgedragen tot het behalen van het doctoraat in de toegepaste wetenschappen door Erik LAES

U.D.C. 621.039 Wettelijk depot D/2006/7515/67 ISBN 90-5682-733-2

Oktober 2006