Institutional Change and Experimentation

Centre de recherche interuniversitaire sur la mondialisation et le travail (CRIMT) Institutional Change and Experimentation: Shaping the Future of Work and Employment 21-23 May 2015 Author: Dr Andrés Feandeiro

HEC Montréal, Montréal [email protected]

Paper Title: Institutional change and experimentation in French regional economies - from hierarchy to collaborative coordination - Evidence from the Rennes-Bretagne ICT cluster Abstract This paper positions itself within recent debates in economic geography and the ‘comparative capitalisms’ literature interested in multi-scalar institutions and how, and the extent to which, institutional actors at lower spatial scales are able to engage in institutional experimentation in the context of persistent institutional constraints of ‘national business and innovation systems’ (NBIS). By way of a case-study of the ICT ‘competitiveness cluster’ in the city-region of Rennes-Bretagne in France, it explores multilevel governance mechanisms in France and interrogates the extent to which experimental forms of territorial and sectoral governance deviate from the ideal-type conceptions of the French NBIS. Findings suggest that while the ‘dominant-developmental state’ and French ‘integrated conglomerates’ (Whitley 2007) continue to be the dominant institutional actors in French regional economies, territorially ‘embedded’ governance actors have played an important role in mobilising regional collective action and in exploiting the institutional opportunities of the NBIS in order to reposition their local economies in the context of global competition and increasing economic uncertainty. While comparative data from other regions and sectors is needed, these findings suggest that it is possible for more dynamic modes of horizontal coordination; beyond large firms and hierarchy to become institutionalised in France. Developing and sustaining novel institutional mechanisms are, however, path-dependent and heavily shaped by both the industrial and innovation policies of the French state and on the corporate strategies of the leading regional French firms.

Literature and Theoretical Framework This paper acknowledges the recent interest that related, but largely tangential, literatures have shown in the governance of regional and local economies and institutional coordination mechanisms operating across multiple spatial scales. At an abstract level these literatures have begun to converge in conceptualising regional, local or territorial economies as economic and institutional spaces that are, to varying degrees, embedded in ‘national business and innovation systems’ (Whitley, 2007) or distinctive national ‘varieties of capitalisms’ (Hall and Soskice, 2001). Economic geographers and regional studies scholars have, for instance, begun to interrogate the dichotomy between societal and social institutions (Rodriguez-Posé and Storper, 2006). Peck and Theodore (2007) advocated a ‘variegated capitalisms’ approach which aims for a closer interrogation of the interaction between subnational economies and their macro-institutional environment. MacKinnon et al (2009) called for explicit theorising and empirical investigation of the ‘missing link’ between subnational and national institutions, while Gertler (2010) suggested that future research needs to be reinforced with ‘more geography’ in order to understand how institutions are produced, reproduced and interact at multiple spatial scales. These concerns have also begun to converge with

the work of ‘evolutionary’ geographers (Boschma, 2004; Maskell and Malmberg, 2007; Essletzbichler and Rigby, 2007; Ter Wal and Boschma, 2011) who have recognised the importance of micro and macro-level institutional drivers and constraints in evolutionary processes. For instance, MacKinnon et al (2009) suggested that combining traditional institutional insights with an evolutionary approach would better account for the role of power, governance and the important selection role played by state agencies in mobilising economic agents which establish new developmental paths and explain the clustering of technologies and industries in particular regions. The regional studies literature has therefore explicitly begun to engage with the comparative institutional literature. Holding NBIS (Whitley, 2007) or national ‘varieties of capitalisms’ (Hall and Soskice, 2001) as the main units of analysis, comparative institutional research has similarly begun to emphasise the importance of ‘multi-level’ governance coordination mechanisms (Deeg and Jackson, 2007; Lane and Wood, 2009; Crouch, 2005; Crouch et al, 2009) and have empirically investigated regional and/or sectoral modes of coordination within national models (Crouch et al, 2001, 2004; Crouch and Voelzkow, 2009; Trigilia and Burroni, 2009; Lange, 2009). While these authors have recognised that national models are important, they have suggested that they are not dominant. For instance, the comparative cross-sector and cross-national case-studies of Crouch et al (2001, 2004) highlighted that institutional environments across countries and sectors are increasingly characterised by multilevel interdependence. Crouch (2005) and Crouch and Voelzkow (2009), for instance, provided empirical evidence of ‘institutional entrepreneurship’ in a number of high-tech sectors and localities in the USA, Germany and the UK. The comparative institutional literature on the whole has therefore more recently evolved towards predicting that institutional experimentation and diversity at the regional, sectoral or local level would occur, but within the persistent constraints of the national institutional architecture and that such processes would not necessarily result in alternative local or regional institutional configurations (Lane and Wood, 2009). These theoretical and empirical insights support the more nuanced theoretical model developed by Whitley (2007) who argued that the national distinctiveness of NBIS was contingent on the strength of the national institutional regime vis-à-vis regional or sectoral international regimes, the extent to which national institutions are complementary in reinforcing the strategies and behaviour of leading firms, and the extent to which the organisation of societal and interest groups is standardized throughout the national economy. As suggested by recent studies on French capitalism (Amable and Hancké, 2001; Hancké, 2001, 2003; Levy, 1999, 2005; Aniello and Le Galès, 2001; Culpepper, 2001), the insights and analytical tools of comparative institutional scholars are particularly relevant and useful in explaining the coordination mechanisms governing French regional and local economies. This body of work has, for instance, advanced the ‘large firm-strong state’ thesis, supporting Whitley’s (2007) ideal type descriptors and suggesting a high degree of institutional complementarity between the ‘dominant-developmental state’ institutional regime, ‘integrated conglomerate’ business system and ‘state-led’ national innovation system. Whitley (2007) notably suggested that the institutional regime of the ‘dominant-developmental state’ has institutionalised a ‘state-led’ innovation system in France, resulting in a high degree of central coordination which encourages the integration of public research projects with the development of innovation in the leading French firms in technologically sophisticated sectors. Cohen (2007) argued that the leading French firms were all born or nurtured through the grand projects of French ‘high-tech Colbertism’ which provided them with discretionary power, the protection of the French state and enabled them to develop competitive advantages through mastering a particular technology, a high level of productivity or gaining market share. This

distinctive institutional configuration, as argued by Cohen (2007), has provided French capitalism with a comparative advantage in high-tech, low variety products and services which offer large economies of scale and are dependent on the long-term commitment of large firms and the state. The French innovation system is therefore characterised by a distinctive non-market, ‘state-led’ system where the risks involved in R&D are shared between the leading large firms and the state. The ‘dominant-developmental’ state therefore, according to Whitley (2007: 70) facilitates and coordinates the active involvement of firms in the public science system around mission-orientated innovation policy and is restricted to certain economic activities favoured by the state and not generalised throughout the economy. Knowledge production, transfer, and use, according to the typology, is monopolised by the state which, through the national education system and public research laboratories, is the most dominant producer of knowledge. Another key feature of French ‘high-tech Colbertism’ is what Storper (1993) termed the French ‘convention of quality’. French technological strength in ambitious, complex, high technology projects, suggested the author, was derived from ‘the convention of quality’ that industrial competence includes the entire commodity chain (filière). The development of a high technology product has therefore traditionally been based on the development of the filière rather than on final markets. Consequently, until recently, and as is still evident in a number of sectors, the leading French firms and ex-national champions have depended heavily on efficiently coordinated ‘local production systems’ for their competitive advantages which implicates a number of state and quasi-state actors. (Aniello and Le Galès, 2001) The institutional foundations of this model of capitalism is rooted in the elite French system of ‘grandes écoles’ which educate state technocrats, engineers and the top management of large French firms, creating a ‘closed-elite’ and encouraging a high degree of hierarchy, social solidarity and loyalty (Storper, 1993; Hancké, 2001). The French business system is therefore, according to Whitley (2007), dominated by ‘integrated conglomerates’ characterised by owners and top managers of the leading firms forming close network relationships with state elites. In such institutional environments, he suggested, state elites exercise authority over and provide resources to large firms which are involved in formulating and implementing the economic development and industrial policies of the state, which according to Cohen (2007), has enjoyed a recent resurgence in the absence of successful European industrial policy, and as French industrial competitiveness has waned. In the context of the ‘large firm-state led’ thesis highlighted above, French local and regional economies have been described as a ‘state-led Meccano production system’ (Aniello and Le Galès, 2001), where the decentralised activities of a limited number of leading French industrial firms accompanied by large scale investment in the skills and innovation infrastructure of provincial capital cities, have historically been the key feature of the French state’s post-war regional economic development strategy, particularly in under-industrialised or declining regions. While Levy (1999) and Hancké (2003) argued that local and regional economies in France were characterised by institutional incapacity and therefore dependent on the decentralised administration of the French state in the region (Levy, 1999) or ‘captured’ by large French firms (Hancké, 2003), Aniello and Le Galès (2001), presented a more optimistic picture of institutional capacity emerging in French regional and local economies, leading them to suggest that France may be facing a ‘silent revolution’ with a future beyond hierarchy, large firms and vertical coordination; towards a future where the comparative advantages of the French economy are derived from new local, regional and state public policies, the mobilisation of associations, horizontal modes of coordination and the dynamic capabilities of small and medium sized firms. This view was recently supported by Levy (2006) who suggested that regional and sector societal organisations in France have become strengthened.

Methodology and Data Collection The research formed part of a broader PhD research project exploring the role and importance of territorial institutional actors in ‘embedding’ multinational enterprises (MNEs) in peripheral or semiperipheral regions in France. The broader PhD research project adopted an exploratory, embedded case-study methodology and employed qualitative methods of data collection and analysis. An exploratory case-study research strategy is considered to be a useful way of overcoming challenges encountered when investigating a complex social phenomenon or problem (Yin, 2003), from multiple perspectives (Flyvbjerg, 2006) and when employing multiple levels of analysis (Eisenhardt, 1989). The case-study is also particularly useful when using critical case selection, as this enables exploratory research to arrive at potential theoretical generalisations (Flyvbjerg, 2006). The Bretagne region was critically selected as the single case to be studied as, although it is generally considered to be a peripheral technological region in France, it enjoys considerable engineering and scientific capabilities in information and telecommunications (ICT) technologies predominantly clustered around the provincial capital, Rennes. Rennes-Bretagne (as the region will be referred to), furthermore, has an active regional innovation policy directed at developing and sustaining territorial innovation networks and enhancing regional scientific capabilities (Crespy et al, 2007). The RennesBretagne region was therefore critically selected as a ‘most-likely’ case (Flyvbjerg, 2006) where bottom-up institutional change and experimentation were likely to be evident. The main units of analysis for this piece of research was the ICT ‘competitiveness cluster’, the pôle de compétitivité, ‘Images et Réseaux’ in the city-region of Rennes-Bretagne. Qualitative data was collected by way of 10 semi-structured interviews with subnational institutional actors concerned with innovation and local economic development policy; managers and governors of the ‘competiveness cluster’, including the CEO of a regionally important SME; managing directors of the leading French ICT firms responsible for setting up the cluster and actively involved in its governance; the managing directors of international corporate R&D units who are members of the cluster and senior management from the local university. Although, this is by no means an exhaustive sample of firms and actors in the Rennes-Bretagne ICT sector, the depth of qualitative data collected and the fact that views were triangulated (by conducting interviews with actors representing all relevant interest groups involved) makes it possible to arrive at corroborated generalisations. Interview data was furthermore supplemented by building a detailed case-study data base of documentary evidence. This is presented below. The Pôles de Compétitivité In 2004 the French state launched a call for regional actors to bid for the establishment of a pôle de compétitivité (‘competitiveness’ or innovation cluster) in a particular thematic area and covering a specifically defined geographical area. The ‘concours’ or contest initiated by the French state’s call took the form of a comparative selection process (‘beauty contest’) rather than a competitive bidding process (‘auction’) as the regions had no idea how many pôles the French state would eventually establish. In 2005 the French government accepted the majority of submissions made and created a total of 71 pôles, effectively covering all regional economies in France. The pôles de compétitivité can basically be viewed as a national innovation policy tool. However, as a policy tool which aims to territorially embed innovative capabilities in regions identified as specialising in a particular technological domain, it is as much a regional innovation policy tool, as a national one. In the city-region investigated, for instance, a senior official from the regional government, the Région Bretagne, highlighted that the region’s pôles were the main strategic tool used by the region to attract FDI and ‘embed’ foreign MNEs in the regional innovation system. Since

the launch of the pôles in 2005, however, all the French ‘competitiveness clusters’ have become inter-regional and can therefore be best understood as being neither national nor regional. The primary function of a pôle de compétitivité is to promote collaborative research and development (R&D) involving large firms, small and medium sized firms (SMEs) and the public scientific community, and fast-track funding available for innovative R&D projects, particularly to SMEs. Each pôle therefore validates and approves the viability and pertinence of the collaborative research projects which are submitted to them. An R&D project ‘labelled’ by a pôle de compétitivité is recognised for its technical and organisational quality, its cost efficiency and the credibility of the partners to complete predefined R&D tasks. The French state, however, maintains a fundamentally important supervisory role over the pôles; notably in agreeing to subsidise the collaborative R&D projects which each pôle accredits. The pôles, furthermore, are the only regional economic development organisation which signs a convention with the French state. The state therefore provides finance to the pôles according to their mission and objectives; their performance is audited on a regular basis and the extent to which they have successfully realised their mission objectives are evaluated using the ‘key performance indicator’ (KPI) measurement. The pôles therefore act as agents of the French state, implementing nationally defined innovation priorities at the level of a sector and region. The pôles accord no finance for the R&D projects it validates. Financial subsidies for the collaborative R&D projects ‘labelled’ by the pôles are primarily provided through the agencies and ministries of the French state: the ‘unique inter-ministerial fund’ (FUI), the public investment bank (BPI France-Oséo) and the national research agency (ANR). Between 2007 and 2011 these state funded agencies made respectively on average over 200 million, 187 million and 179 million euros available each year to all the pôles.1 In 2013, for instance, the French state’s national research agency (ANR) provided € 8 million to the pôle ‘Images et Réseaux’ for collaborative research projects. The national agencies and ministries of the French state therefore are the principal sources of finance for the R&D projects ‘labelled’ by the pôles and play a direct role in approving the type of research activities they are engaged in. The Rennes-Bretagne ICT cluster, ‘Images et Réseaux’ ‘Image et Réseaux’ is certified by the French state as a ‘worldwide’ cluster which counts the regional R&D units of the leading international French ICT firms; France-Telecom-Orange, ThomsonTechnicolor, Alcatel-Lucent, Thales and the R&D centres of foreign multinationals, such as Mitsubishi Electric and Canon, as its members. It was certified as a pôle by the French state in 2005. The city-region of Rennes-Bretagne has historically possessed strong capabilities and expertise in information and telecommunications technologies and has been heavily dependent on the mobile networks and ICT services industry. Over the years, these competencies have evolved; maintaining its core networks and digital video expertise, strongly related to its telecoms specialism and, more recently, developing, competencies in cyber security and cloud computing, around important engineering schools and the research facilities of the French Defence Ministry, which are located in the city-region. The region has a high concentration of digital services firms which provide a range of general services to the large French ICT firms and the Defence Ministry’s ICT research centre. When the national call to bid for the creation of a pôle de compétitivité was launched in 2004, competition between French high-tech regions bidding for the creation of a pôle in ICT related 1

Enquête annuelle de la DGCIS auprès des pôles

themes was expected to be fierce as each of the high-tech French districts; in the Paris region, in the Rhône-Alpes, in the south of France and in Bretagne, were expected to bid. In the city-region of Rennes-Bretagne, managers from local units of the leading French ICT firms, supported technically by the directors of the local grandes écoles and public research institutes specialising in ICT, as well as regional public policymakers from the regional government (Région Bretagne), who aided in the drafting of the dossier, were responsible for making a collective regional bid for the creation of a digital image and networks cluster, which they named ‘Image et Réseaux’. ‘Images et Réseaux’: A multi-stakeholder governance actor ‘Images et Réseaux’ is a multi-stakeholder associational organisation. Under French law it is an ‘association à but non-lucratif’ or a non-profit citizen’s organisation. It is therefore considered as a private entity with a third of its funding derived from members’ contributions. When the pôle was established, however, roughly two-thirds of its operating costs were shared equally between the French state and the regional government. It is therefore a public-private partnership, which is also reflected in its governance structure. The executive board consists of representatives from five ‘colleges’ representing the most important regional economic and institutional actors: large firms; SMEs; public research organisations and universities; economic development organisations and local authorities and other associations. The three most important ‘colleges’; large firms, SMEs and public research organisations and universities have the majority of the representatives on the executive board (18/21 members). The presidency of the pôle is held by a representative from the ‘college’ of large firms and its three vice presidents are nominated from each of the three most important ‘colleges’. Since its creation, the French state has reduced its subsidies and the financial short-fall has resulted in the territorial boundaries of the pôle being extended to cover software SMEs in the city-region of Nantes (Région Pays de la Loire.) ‘Images et Réseaux’ can therefore be accurately described as an inter-regional public-private partnership between the large French ICT firms, exclusively based in the Bretagne region, SMEs, universities and public research laboratories in the city-regions of Nantes and Bretagne, and the regional governments of the ‘Région Bretagne’ and the ‘Région Pays de la Loire’. The regional governments are not only important stakeholders in the pôle but are an important source of its operational funding. Each year, ‘Images et Réseaux’ signs a contractual agreement with both regional governments. They present a provisional budget and their existing finances to their public stakeholders, and based on this information each public authority agrees to subsidise them to a particular amount. The regional government of the ‘Région Bretagne’ played an important and active role in shaping the objectives of ‘Images et Réseaux’ and in financing it from the very beginning. The Région Bretagne played a critical role in ensuring that the pôle did not become a cluster of, and for, the large French ICT firms. The regional government invited regional SMEs to join the cluster and ensured that they had an important presence in the governance of the pôle and argued strongly that all stakeholders be included as members of the executive board. Regional governments are most significantly, an important source of funding for the collaborative research projects of the pôles which it shares on an equal basis with the French state. The regional government’s Department of Innovation and Technological Transfer has almost one-third of the region’s total budget for economic development, (30 million euros per annum) of which almost half goes towards financing collaborative R&D projects; the vast majority of which are research projects ‘labelled’ by one of the Région Bretagne’s four pôles. Between 2008 and 2011, French regions, for example, made on average over 181 million euros available to the pôles for collaborative research. Importantly, of this amount, over 54 million euros,

or almost one third of the funding provided by regional governments, went to collaborative projects not financed by the FUI and ANR, which evaluate the projects and decide whether to fund them or not.2 In the Région Bretagne this funding was used to finance collaborative research projects involving SMEs and academic institutions only (through a regional call for projects), and in which 50% of the budget and resources are in the hands of SMEs. The policy advanced by the regional government followed a developmental logic, giving SMEs an opportunity to be involved in a collaborative project with a small number of partners (2 SMEs and a research laboratory) in order to prepare them for larger project collaborations through the pôle or for collaborative European Union (EU) ICT projects. ‘Images et Réseaux’ was therefore the first ‘competitiveness cluster’ in France to provide exclusive support for SMEs and counts this as one of its major successes. SMEs have subsequently become the leading firms in certain collaborative research projects underway, even leading projects in which the large French ICT firms are participants. Compared to other pôles de compétitivité in France, there has therefore existed a strong symbiosis between the Région Bretagne and ‘Images et Réseaux’ because the regional government has been actively involved in co-constructing it with the ICT industry and the local scientific community. ‘Images et Réseaux’ has, furthermore, become an important collective actor at the regional level of governance. For instance, between 2009 and 2011, it piloted, on behalf of local managers of the leading firms of the Rennes-Bretagne ICT sector, the region’s bid for the creation of a regional ICT ‘Institute of Research and Technology’ (IRT - Instituts de recherche technologique) called ‘B-Com’. Extending the concept of territorially embedded collaborative modes of coordinating innovation, characteristic of the pôles de compétitivité, the IRTs were envisaged to be fully mutualised research centres where researchers from public laboratories and the leading French firms pool their human and physical resources to undertake experimental research and build innovative capabilities around interdisciplinary themes which would have the potential for application across various sectors. 3 The existence of a regional multi-stakeholder organisation, such as ‘Images et Réseaux’, which counted all the major actors from the region’s ICT ‘ecosystem’, and regional public policymakers on its governing board, provided a collective institutional space and a strong governance actor for the region to put together a comprehensive initial bid for a regional ICT technology institute and enabled local actors to project-manage its development and lobby the French state over an extended period of time. The large French ICT firms and ‘Images et Réseaux’ The leading French ICT firms present in the region invested a lot of time and effort in putting together a comprehensive bid for the pôle in 2005 and continue to invest in its governance, management and in the evaluation of its collaborative research projects. Notably, one of the most important technology advisors to the President of France Telecom and Innovation Director of Orange Labs, located in the city-region, was actively involved in the creation of ‘Images et Réseaux’ and became its first president, while its first managing director came from Thomson. The current (at the time of conducting the research) president is from France Telecom’s Orange Labs and one of its vicepresidents is from Thomson-Technicolor. France Telecom-Orange therefore invested a lot of time and resources in making the pôle a reality and a success. R&D managers of the locally based Orange-Labs took on a ‘project management’ role in the creation of the pôle, collaborating closely with local managers from Thomson-Technicolor. The thematic specialisations of the pôle (‘Images’ and 2 3

Enquête annuelle de la DGCIS auprès des pôles http://competitivite.gouv.fr/ ; http://investissement-avenir.gouvernement.fr/

‘Réseaux’) therefore reflect the core competencies of these two leading ICT firms in digital networks and audio-video technologies respectively, and they were the most important actors in initiating its activities from the beginning. The large French ICT firms in the pôle remain, at the time of research, important initiators of the projects accredited by the pôle. The creation of the pôle, furthermore, coincided with France Telecom-Orange actively pursuing a strategy of ‘open innovation’. France Telecom-Orange, has evolved its business model and supports using all standards, services and technologies which exist in the local innovation ‘ecosystem’. Of all the major ICT firms located in the region, France Telecom-Orange are the most open to accommodating engineers who want to develop a start-up project; participating in local innovation policy and making available their technological expertise and knowledge to the local ICT ‘ecosystem’. Its human resources policy, furthermore, facilitates the creation of innovative spin-offs, as their engineers have the possibility to leave the firm for up to three years to pursue a spin-off project with the job security of being able to return. The strategic importance of developing territorialised collaborative institutions in a region where it has an important presence is, nevertheless, clear for the now privatised French telecommunications ‘national champion’. It is a network operator which derives its competitive advantages from operating a platform and deploying applications across its network, but it does not have the capacity to invent all the internet services and applications imaginable in mobile communications. Consequently, the firm has a strategic motivation in actively developing collaborative and ‘open’ innovation structures and in nourishing the regional ‘ecosystem’ in which its R&D labs are located as the regional innovation system is the source of new ideas and potential innovative applications and network services which it can commercialise. The collaborative innovation occurring through the pôle is therefore potentially the source of its future competitive advantages. This contrasts with the Franco-American telecommunications equipment maker, Alcatel-Lucent, whose engagement in the pôle was constrained by the fact that its regional units undertake product ‘development’ rather than ‘technological research’ (which is located at the group’s Bell Labs outside Paris). This made it difficult for it to become easily embedded in a cluster dominated by France Telecom-Orange and Thomson-Technicolor whose principal ‘research and innovation’ centres are located in the region. Engineers at Alcatel working on relatively short-term product development therefore have relatively few common research interests with engineers and researchers from the other two leading firms involved in ‘upstream’ research on core technologies. The units of AlcatelLucent located in the region are also business units and required to be profit generating. R&D managers therefore need to ensure that there is a return on the investment of their R&D resources which makes participating in a collaborative R&D project without clear commercial gains more difficult. The active involvement of the leading French ICT firms in the creation and continued involvement, support and governance of the pôle is, nevertheless, motivated by a sense of civic responsibility as their participation in the cluster has more indirect than direct benefits for them. In an interview with one of the leading French ICT firms, local managers highlighted that it was important for them to be proactively involved in the regional ICT ‘ecosystem’ to make sure that it remains dynamic and attracts and keeps the best talent so that that value is created locally. The large French ICT firms viewed the creation of the pôle as an opportunity to develop the regional ICT ecosystem at a time when their firms and the telecommunications industry in the region were facing enormous challenges after the telecoms bubble burst in the early 2000s. Large firms such as Alcatel-Lucent and Thomson-Technicolor prior to, and in the period immediately after the creation of the pôle, were undergoing major restructuring of their activities and their investment in the development of the

cluster were, for the local managers of Alcatel, a way of showing that despite the negative social effects which had accompanied its restructuring, the firm was still actively engaged in developing and investing in the regional ICT sector. In the case of Thomson-Technicolor, the cluster provided local managers with a forum to be able to rationalise and explain to industry colleagues the new repositioning of their activities, and that the loss of jobs at the firm were also accompanied by the creation of new firms and start-ups which would nourish the region’s ICT ecosystem. On the other hand, the active involvement of the large firms in the creation and development of the pôle, was a way of mobilising the regional ICT community and giving the sector in the region a new vision; to a future beyond telecoms and towards audio-video and digital networks. The setting up and development of the pôle ‘Images et Réseaux’ in the period 2005 to 2008, was therefore a highly important and exciting phase of evolution, entrepreneurialism and dynamism in the regional ICT sector. The leading French ICT firms in the region were, and still are, therefore indeed motivated by a sense of civic responsibility, but view their active engagement in the activities and governance of the pôle as being essential to ensuring that the regional ICT ‘ecosystem’ remains performant as their success in highly competitive international markets depends on it. The large French ICT firms have furthermore remained heavily invested and involved in the collaborative activities and governance of ‘Images et Réseaux’ despite a substantial reduction in the subsidies they receive for involvement in the R&D projects of the pôle. This therefore, according to a senior manager of Thomson-Technicolor, highlights the extent to which they are committed to nourishing and developing the regional ICT sector. ‘Images et Réseaux’: vertical and horizontal modes of coordination The French pôles de compétitivité are essentially innovation clusters. The primary objective of the third phase of the pôles is to facilitate and develop collaborative or horizontal forms of innovation and economic coordination, particularly among SMEs and between SMEs and universities, research laboratories and the leading firms. The pôles are therefore primarily focused on developing the dynamic capabilities of SMEs and territorially embedding innovative capabilities through horizontal modes of coordination and collaboration between firms and research institutes. The interview data collected however suggested that the development of collaborative and horizontal forms of economic coordination in France has faced considerable institutional constraints. In the case of ‘Images et Réseaux’ local institutional actors involved in its creation and early development had considerable difficulty in getting the leading large firms, SMEs and research organisations to collaborate with each other as the highly entrenched traditions of either noncollaboration, on the one hand, or of ‘closed’ collaboration on the other, were difficult to overcome. The latter concerns the bilateral contracts between large firms and the research laboratories of the grandes écoles which was, and still continues to be, a significant method for large French firms to exploit the cutting edge research expertise of the French public innovation system, whilst the former, refers to the institutional environment which has traditionally characterised the relationship, or lack thereof, between small firms, on the one hand, and large firms and research laboratories and universities, on the other. Prior to the existence of ‘Images et Réseaux’ collaboration between SMEs and research centres and universities, for instance, was very rare. It was therefore necessary to find incentives to overcome firstly, the general lack of a collaborative culture, and secondly to overcome the considerable resource constraints that, particularly SMEs and research laboratories, faced in entering into a collaborative innovation project.

The early years of the pôle were, furthermore, dominated by the large regional ICT firms and characterised by highly hierarchical forms of collaboration. The leading French ICT firms located in the region had invested an enormous amount of time, resources and manpower into the pôle from its conception. Consequently, the early projects of the cluster were directed by the large French ICT firms and they required the necessary incentive to ensure that its early projects were successful. Therefore, in the first 2-3 years of the pôle’s existence, the large French ICT firms, received the highest amount of subsidies (40%) for participation in collaborative R&D projects. The large ICT firms also absorbed the majority of the total funding made available to the pôle for research projects as all of the early collaborative projects were ‘ready-made’ research projects that they had shelved over the years. The early projects of ‘Images et Réseaux’, for example, involved large-scale collaborations; with the regional units of one of the large French ICT firms leading a group of 10 to 15 SMEs. Furthermore, these early projects were characterised by hierarchical modes of coordination and a certain degree of condescension on the part of the large firms, where SMEs were treated like a subcontractor within the collaborative project. The existence of SMEs around the table for a collaborative R&D project, was less out of a necessity, but more out of ‘good conscience’ and SMEs were largely given the bits of the project which the large firms were not interested in developing. For instance, the experience of one SME during an early collaborative project was that the large ICT firm would ask an SME to complete a particular task and to deliver a technological ‘brick’ to them under specific instructions without giving them the opportunity to think about why they were developing it in that way. Later, however, ‘Images et Réseaux’ began to overcome some of these institutional constraints hindering collaboration and has since become less characterised by hierarchy and the dominance of large firms. Firstly, during the first 2-3 years of the pôle it became clear that the subsidies regime for R&D needed to be overhauled if entrenched institutions around collaboration were to be broken. The primary motive behind this change was to provide greater financial incentives for SMEs, which already had to surmount considerable difficulties in bidding for a collaborative project. In the first few years of the pôle’s activities, each partner involved in a collaborative project turned to its traditional method (and state agency) of securing finance for R&D, which ultimately meant that there was effectively no financial incentive for SMEs to be included in a collaborative research project. In fact, initially, this resulted in SMEs facing particular difficulties in securing the necessary subsidies as the ‘National Agency for the Valorisation of Research’ (ANVAR) which usually provided finance to SMEs for R&D would not finance collaborative research projects. Therefore, financial incentives had to be introduced in order to firstly, encourage SMEs to favour a collaborative R&D project launched by the pôle over an independent project, for which they were more certain to receive funding from ANVAR, and secondly, to encourage both large firms and research institutes to work with SMEs. Subsequently, under the new regime, an SME could be subsidised up to 50% for a collaborative research project and research institutes and large firms were also provided with additionally subsidies for collaborating with an SME. At the time of undertaking the research, the subsidy regime for participation in a collaborative project of the pôle was fixed at 25% for a large firm, 45% for an SME and 100% of any additional costs incurred by a public research institution (e.g. the financing of a doctoral or post-doctoral researcher).This indicates that through the pôles, SMEs are now provided with more financial support for R&D than large firms who are involved in collaborative research. The introduction of greater financial incentives to encourage collaboration between different actors were the first steps taken to develop more horizontal modes of coordination around R&D in the Rennes-Bretagne ICT cluster. Subsequent initiatives included, for instance, the creation in 2008 of a permanent Vice-President representing SMEs and the real chance that a CEO from a local SME,

rather than a manager from a large French ICT firm, could one day become president of the pôle. During the second phase of the pôles, with a leading presence on the governing council of ‘Image et Réseaux’, therefore SMEs began to push for a change to the hierarchical nature of collaborative R&D that characterised the first few years of the pôle. This involved encouraging ‘positive discrimination’, such as, concentrating 50% of the budget and manpower of a collaborative R&D project in the hands of SMEs and creating the conditions where SMEs could later lead large firms in a collaborative project. At the time of conducting the research, local SMEs were, for example, leading projects with large firms involved. These ‘regional calls’ for collaborative R&D projects involving SMEs only were financed by the regional government of Bretagne. While the large French ICT firms based in the region still generate a considerable amount of the collaborative R&D projects launched by the pôle, local SMEs have since developed their competencies in participating in collaborative R&D and have therefore begun to play an important role in initiating projects. The pôle has therefore achieved one of its most fundamental objectives which is to improve the competencies of the SME ‘ecosystem’ in order to be able to participate in collaborative R&D projects. The hierarchical and often condescending large firm-small firm relations which characterised the early years of the pôle have also evolved to a more horizontal relationship characterised by partnership and mutual respect. The large French ICT firms have therefore realised that the local SME ecosystem is capable of creativity and innovation and are now taken more seriously and consulted. While ‘Images et Réseaux’ is entirely credited with being the catalyst for developing a territorially embedded institutional space in which horizontal modes of coordination between large firms, SMEs and the regional scientific community can flourish, the limits of building territorially embedded, collaborative innovative capabilities should not be understated and remain heavily constrained. ‘Images et Réseaux’ has not managed to attain a ‘consortium agreement’, which effectively regulates which parties, and how, the results of a collaborative project will be communicated and/or commercially exploited. For reasons around intellectual property protection, there is no codevelopment of technology in the collaborative R&D projects of ‘Images et Réseaux’. Each partner brings a ‘technological brick’ in the value chain to a particular R&D project, but it and the patents it has generated around that particular technology, remain the intellectual property of the firm or laboratory which developed it. Therefore ‘collaboration’ in the pôle occurs in terms of the initial reflection on a common programme or theme and the joint problem-solving which occurs prior to the developmental stage, but actual development is completely intra-organisational. In the postdevelopment phase of R&D, there is furthermore, no communication between partners in the filing of patents. This process is undertaken completely independently. The collaborative R&D projects of the pôle do not, furthermore, normally involve technological transfer between partners. Where technology transfer has occurred, it has most notably emanated from research laboratories. There have been some instances of a technology being created in a research laboratory, during a collaborative project, which has then been integrated by an industrial partner. Since the creation of the pôle, seven start-ups have also been created in order to commercialise technological ‘bricks’ developed in a research laboratory during the course of a project. At the time of conducting the research, there was potentially an opportunity for a spin-off to commercialise a particular technology developed by a large firm during a collaborative project as it was not part of its core activities or business strategy. There were however at the time no concrete examples of technology transfers emanating from one of the large French ICT firms. In this type of situation part of the work of the pôle is to encourage the creation of spin-offs in order to avoid a technology with potential market success not becoming commercialised. According to the managing

director of ‘Images et Réseaux’, ensuring that the technology developed through collaborative research can find market access is one of the primary objectives of ‘phase three’ of the pôles Since the creation of the pôle, there have furthermore been instances where the large French ICT firms have used SMEs in the ‘valorisation’ of its technology, involving a more traditional clientsupplier type of relationships. There have also been instances of an SME developing particular technological tools while the large firm provides the user framework and if the large firm is interested in the tools developed it will enter into a commercial contract where the SME becomes its supplier. However, the large French ICT firms tend to use the collaborative projects of the pôle to acquire knowledge and add to their intellectual property and are not pursuing a strategy of ‘valorising’ and industrialising its patents and therefore bringing the local network of SMEs into its supply chain. For instance, despite attempts by ‘Images et Réseaux’ to develop commercial relationships between large firms and SMEs in the region, the vast majority of local SMEs in the cluster are not suppliers to any of the large French ICT firms. Consequently, the objectives of the collaborative research projects of the pôle have not been to industrialise and commercialise technology developed. Historically, the regional SME ‘ecosystem’ has developed as a result of spin-offs from the large telecoms and broadcast national champions located in the region, particularly France-Telecom / Orange and TDF around the city of Rennes. As these large conglomerates exited from particular business activities, technologists left with their patents and grafted themselves in the innovative ‘milieu’ of the large firms. The development of the optics cluster around the town of Lannion in Bretagne is one such example. Almost the entire SME optics sector in Lannion exists because when France Telecom decided to exit the fibre optic market, technologists working in their labs left the company to setup their own firms. SMEs in the region have therefore always traditionally developed around the periphery of the large French ICT firms and exploit the presence of the large firms who invest a substantial amount in developing the regional ICT sector. ‘Images et Réseaux’ has nevertheless been highly significant in developing horizontal ‘network’ modes of coordination, both formal and informal, in the region’s ICT sector. It has had a positive effect on the relations between ICT firms in the region, even between large firms with historically arms-length and hierarchical relations. A notable example of this is the relationship between AlcatelLucent and France Telecom-Orange. Relations between these two French conglomerates were historically hierarchical as Alcatel was a components supplier to France Telecom. There had existed research contracts between the two firms, but after the latter was privatised, for about ten years there was no interaction between the regional units of the two firms. Interaction between the R&D units responsible for product development at Alcatel and France Telecom was however reestablished upon the creation of the pôle. The pôle also initiated collaborative forms of economic coordination in the regional ICT sector that had not previously existed, notably between SMEs, on the one hand, and large firms and research laboratories, on the other. Prior to the creation of the pôle, there was a well-developed culture of collaboration between the state scientific community and the leading industrial firms in France. Collaboration between SMEs and public research laboratories and universities prior to the creation of the pôle were in contrast very rare. In an interview with one of the leading SMEs in the pôle, the CTO attested that he appreciated the networking opportunities that the pôle provided; being able to meet people from the local ‘ecosystem’. He highlighted, for example, that there were important universities and research institutions with whom his firm had extensively collaborated, which he would never have had the opportunity of meeting and appreciating, had it not been for the pôle. He also suggested that their firm has, through the collaborative activities of the pôle, collaborated

almost more extensively with local research laboratories and universities than the large French ICT firms. Perhaps most significantly, the creation of ‘Images et Réseaux’ was the catalyst for the regeneration of the city-region’s ICT sector after the bursting of the telecoms bubble and provided industry actors the opportunity to mobilise collectively, create a dynamic environment and shape a new future for the sector around digital audio-video and networks technologies. This therefore provided new startup firms, struggling SMEs and restructuring large firms with the possibility of integrating a ‘club’ which provided collective support and development. Furthermore, the collaborative forms of inter-firm relations evident in the pôle is characterised by the transfer of highly ‘tacit’ knowledge. While the SME ecosystem have generally not benefitted from the transfer of ‘hard’ technological knowledge, nor captured the markets of the leading French ICT firms, their participation in the collaborative projects of the pôle have provided them with invaluable tacit knowledge. SMEs have, for example, progressed in their ability to manage R&D processes through collaborating with the leading ICT firms which have well-developed methodologies in the management of R&D and have gained substantial tacit knowledge from the large firms in terms of business strategy, market analysis, intellectual property protection and the valorisation of their R&D. Through their participation and commitment to the pôle, the large French ICT firms have therefore played a fundamentally important role in developing the tacit knowledge and competencies of SMEs. In phase three of the pôles, one of the primary objectives of ‘Images et Réseaux’ is to develop the competencies of SMEs in the cluster in order for them to be able to participate in European R&D projects. The majority of local SMEs do not possess the capabilities to be able to enter into such projects on their own and the cluster is therefore counting on the leading French ICT firms to play an important role in supporting and developing a few SMEs in this area. This was confirmed in an interview with the CTO of a leading SME in the pôle. He particularly referred to the importance of the pôle in terms of the tacit forms of innovative activities which occur during a collaborative project, such as; the opportunity to mature innovative ideas, to gain market knowledge and knowledge in establishing partnerships. He also highlighted the importance of the subsidies accredited by the pôle, without which innovation (particularly in SMEs) would not occur as the risks associated with R&D in the ICT sector were so high, and the opportunities for valorisation and potential profits, so low. The collaborative projects which their firm had engaged in had enabled them to develop technological ‘bricks’ used in their own products. In a collaborative project of the pôle, the financial risks associated with technological development are therefore shared between partners, while its potential rewards and profits benefit individual firms and research organisations. Analysis and Discussion The case study on ‘Images et Réseaux’ presented in the previous section provides a more nuanced picture of the ‘large firm-strong state’ model of French capitalism that has traditionally been advanced in the ‘comparative capitalisms’ literature on France. This literature has collectively emphasised that French capitalism is constrained by the ‘strong stateweak society’ institutional environment where institutions between ‘market coordination’ and state ‘dirigisme’ at the level of an industry, sector or region have either been non-existent or weak. An interventionist ‘dominant-developmental state’, it has suggested, has consequently had to fill this institutional vacuum. This literature also argued that as the French state has retreated from direct economic planning, shifting to a ‘market-supporting’ mode of economic coordination based on ‘infrastructural’, as opposed to ‘authoritative’ power (Levy, 2006), and in the absence of strong associational and societal actors and ‘institutional incapacity’ at lower spatial scales, large French

firms have filled the institutional vacuum and effectively ‘captured’ the emerging institutions of the regionalised French political economy. In such an institutional environment, French regional economies have been characterised as hierarchical, vertically coordinated and integrated into the activities of ‘national champions’ and large French firms. This literature has therefore suggested that regional economic development has become ‘locked’ into the path-dependent trajectory of the ‘dominant-developmental’ objectives of the post-war ‘dirigiste’ French state which sought to develop certain industries in particular regions, and which, moulded regional economies specialising in one or two industries or sectors. The case of the pôle ‘Images et Réseaux’ does not challenge the ‘large firm-strong state-weaksociety’ thesis of the literature on French capitalism, however it highlights the extent to which the contemporary regional French political economy is in a state of transition and where significant new, territorially embedded, multi-stakeholder forms of associational governance are emerging and facilitating new forms of bottom-up institutional experimentation. The Statist French ‘National Innovation System’ In terms of the way in which innovation in firms are funded in France, the pôles de compétitivité are not a fundamental departure from the traditional state-led ‘national innovation system’. Simply put, the French state has always subsidised, in one form or another, the R&D activities of firms, both large and small, and the policy of the pôles is consistent with this. In fact during the early years of the pôle each partner involved in a collaborative R&D project validated by the pôle returned to its traditional methods of finding finance. For large French firms this was the Ministry of Industry, for public research laboratories and universities this was the Ministry of Higher Education and Research, while SMEs could apply to the ‘National Agency for the Valorisation of Research’ (ANVAR) for a reimbursable loan depending on the commercial success of the research undertaken. This posed substantial problems for SMEs, resulting in a significant disincentive for them participating in collaborative R&D projects during the early phase of the pôles. While the process of applying for funding for participation in a collaborative project of the pôles has been streamlined over the years, it does not depart from the state-led French ‘national innovation system’ in terms of the way in which firms finance innovation, and most notably in the powerful supervisory and technical role which French state agencies have in the definition and realisation of their objectives. Furthermore, the functioning, subsidies regime and operation of the pôles are standardised across all regions and sectors in France as predicted by Whitley’s (2007) ‘dominantdevelopmental state’ typology. The pôles however do represent a clear departure from the ideal-type conceptualisations of the French innovation regime in that state subsidies are only provided for research projects where two or more, of either large firms, SMEs or research institutions, are involved in collaborative research. In this way, the ‘innovation regime’ of the pôles contrasts considerably in both nature, and in terms of objectives, with the ‘grand projects’ regime which is most closely associated with the ideal-type ‘national innovation system’, often described in the literature as French ‘Colbertism’. As previously highlighted, the objectives of the policy of the pôles is to territorially embed the innovative capabilities of large and small firms, develop regional specialisms in a particular technological domain and encourage the development of an endogenously created ‘innovative milieu’ which enables, particularly the SME sector, to develop competitive advantages. As noted earlier, in the third phase of the pôles, the development of the innovative capabilities of SMEs is considered to be central to realising this objective, which represents a clear movement away from the ideal type ‘state-led’ French innovation system which was constructed around the innovation

needs of large French firms and ‘national champions’. In this regard, ‘Images et Réseaux’ has already begun to show some positive results, which will be further discussed later. However, as highlighted by Cohen (2007), the pôles de compétitivité is one of several industrial and innovation policy tools. The French state has not entirely abandoned the national ‘grand projects’ mode of coordinating innovation. Actors from the Rennes-Bretagne ICT were also involved in two innovation projects launched in 2006 by the new ‘Agence de l’innovation industrielle’ (AII), the national agency set up in 2005 to promote experimental innovation across French industry. Project ‘Quaero’ led by Thomson-Technicolor and France Telecom, and involving German partners, envisaged developing software to create a French multimedia search engine to rival Google and Yahoo!, while a project known as ‘TVMSL’ (mobile television without limits) piloted by Alcatel aimed to develop a European standard for mobile TV. The development of regional IRTs is a further example of state-led French industrial and innovation policy, and even though the regional technological institutes are envisaged as an extension and a deepening of the collaborative innovation characteristic of the pôles de compétitivité, it is a territorialised form of ‘Colbertism’. In 2010 the French state made available 35 billion euros (the grand emprunt) for investment in higher education, research and training; the development of small and medium sized enterprises; sustainable development, green energy and the digital economy, of which €2 billion was set aside for investment over a period of ten years in the IRTs. This represents a major and long-term financial commitment by the French state. The institutes are also financed by the leading firms in a particular region who have joined the IRTs and which have also made substantial long-term investments. Local and regional governments are furthermore responsible for financing the development of the infrastructure. Similarly to the launching of the pôles, a competitive call was made for regions to bid for the awarding of an IRT. The French government awarded 8 projects IRT status and these fall under the responsibility of the ANR, the French state’s national research agency who operate the ‘Investing in the Future’ programmes in the fields of higher education and research. Similarly to the pôles, the IRTs are governed by regional public and private actors; including regional and local governments and regional universities and research laboratories who are also primary investors in the project, however, the aim is for them to become completely independent organisations with their own business strategy. The IRTs therefore could potentially become an interesting experimental form of regional coordination, and an important public-private regional economic governance actor around innovation. However, at the time of conducting the research, it was unclear how they would develop and to what extent the independent interests and divergent strategies of member firms could be accommodated with the idea of collaboration and shared intellectual property and moulded into a single collective organisational entity with its own strategy. In the case of the Rennes-Bretagne ICT cluster, the IRT B-com, has already faced a significant setback, the decision by three out of the four most important French ICT firms present in the region not to join due to concerns over its shared intellectual property regime. This could well have important consequences for the organisation’s development and success as the leadership, commitment, involvement and participation of local managers from the large French ICT firms was, and is, one of the main reasons for the successful setting up and development of ‘Images et Réseaux’. On the other hand, the fact that ‘B-Com’ has not become a club for the large French ICT firms may potentially be one of the key factors in ensuring that a genuinely experimental mode of governance or ‘regional business system’ emerges, upon which the region’s dynamism and future economic development can be built.

The pôles de compétitivité and the IRTs therefore suggest that France’s ‘dominant-developmental’ institutional regime and ‘state-led’ national innovation system persist despite the ‘retreat’ of the French state from direct economic planning and coordination. The pôles de compétitivité, and more recently, the IRTs are the most significant industrial, innovation and regional development policy tools of the French state and their implementation particularly draws on the ‘infrastructural’ resources of the state. These policy tools have also reinforced the French state’s position as the primary actor in moulding an institutional environment where ‘bottom-up’ institutional experimentation and renewal can begin to emerge. One of the most significant findings from data collected, was the unanimous view of all respondents interviewed that the pôles were entirely responsible for developing collaboration between actors which had not previously existed. This evidence is highly significant and largely supports the ‘strong state-weak society’ conceptualisation of French capitalism. Without the ‘infrastructural’ power of the French state; including the appropriate policy tools, institutional incentives, close supervision and financial subsidies, the horizontal and collaborative modes of coordination which have become characteristic of the pôles and the IRTs, would not have been able to emerge. Furthermore, the extent to which these regional clusters are territorially embedded is an important point of analysis. As highlighted in the precious section, in terms of their geographical boundaries and governance structure, most of the pôles have become inter-regional multi-stakeholder associations. This is either for strategic reasons, or as finance from the French state has been withdrawn, the pôles have been forced to expand their geographical boundaries in order to increase its membership, as in the case of ‘Images et Réseaux’. The presence, nevertheless, of the corporate R&D units of the leading French ICT firms, Technicolor and France Telecom’s Orange Labs in the cityregion of Rennes-Bretagne, has ensured that, ‘Image et Réseaux’, remains territorially concentrated in Rennes-Bretagne. Key actors involved in its governance have commented on the highly focused nature of the pôle in terms of its activities in comparison to the other major ICT cluster in the Paris region. Institutional capacity at the regional and local level The case study of ‘Images et Réseaux’ most notably shows that important territorial institutional actors, outside of large firms and the French state, have emerged in the governance of French regional and local economies since the 1990s and that there exists considerable territorial institutional capacity in France. Genuinely horizontal modes of coordination would not have emerged regionally without the important role played, most notably, by the regional government of Bretagne. As presented in the in the previous section, the Région Bretagne played a fundamental role in coconstructing the pôle together with the large French ICT firms and other actors from the ICT ‘ecosystem’. In this context, the policy of the pôles can be contrasted, to some extent, to the policy of the IRT’s which has been a more ‘top-down’ innovation policy, more consistent with the institutions of French ‘Colbertism’ and the French NBIS. The Région Bretagne, for instance, has played a less active institutional role in terms of shaping, financing, and governing the ‘B-Com’ as it had done in the creation and development of ‘Images et Réseaux’. This could be explained by the fact that former is in the very early stages of development and as the French state will withdraw financially from these institutes, regional governments may become more important actors in terms of both funding its operations and in terms of shaping how technology produced is used. The regional government of Bretagne has also demonstrated considerable ‘institutional entrepreneurialism’. It moulded the pôles in line with its own strategic developmental objectives and created novel institutions for the development of SMEs through ‘regional calls’ for collaborative R&D

projects involving only SMEs, which it exclusively finances. These initiatives were institutionalised prior to the French state’s ‘third phase’ of the pôles, which sees the development of the innovative capabilities of SMEs as the primary objective. For example, almost half of the region’s expenditure for innovation and technology transfer is devoted to funding the collaborative R&D projects of all the region’s pôles and much of this finance is directed at funding collaborative R&D not funded by the French state agencies. The pôles have therefore become as much a regional innovation, industrial and economic development policy tool, as a national one. The pôles in the region are also considered as the foremost strategic tool in attracting FDI and embedding foreign MNEs in the region. This institutional ‘activism’ on the part of the Région Bretagne has therefore played a significant role in ensuring that the activities of the collaborative projects of the pôles are no longer ‘captured’ by the large French firms and that their governance is not dominated by them. In short, that the pôles have emerged as genuinely multi-stakeholder regional governance actors is in large part due to the active role played by regional public policymakers. Even though the governance of French regional economies continues to be characterised by ‘institutional layering’ in terms of the existence of important institutional actors operating at multiple spatial scales, the case of ‘Images et Réseaux’ suggests that the creation of the pôles have significantly entrenched regional governments as the lead institutional actor in terms of strategic territorial economic development and innovation. Most significantly, the pôles de compétitivité is not just a policy tool, neither are they simply loose associational networks. The pôles have become important regional economic governance organisations and therefore important regional and inter-regional institutional actors in their own right in terms of implementing regional economic development strategy. The fact that they are the only regional economic development organisation which signs a contract with the French state and are governed by both regional public actors and the leading regional firms provides them with a considerable level of credibility in long-term regional economic development strategy. The leading role played by ‘Images et Réseaux’ in developing their region’s bid for the IRT ‘B-Com’ is clear evidence of the important institutional role it plays in terms of federating all the key public and private actors implicated in the region’s ICT economic development strategy. As an open, associational, multi-stakeholder, network actor representing the interests of firms and public policymakers in a particular sector and across a particularly defined geographical region and contracting with the French state, they have become one of the most efficient regional development organisations. The pôles maintain numerous skills and competencies in a wide range of areas; from networking, economic intelligence, and consultancy in intellectual property to specialist expertise in particular technologies, markets and industries and are therefore important institutional support actors for ICT firms located in both regions. Large Firms and French Regions The case of ‘Images et Réseaux’ also confirms the literature on French capitalism arguing that large French firms are the dominant institutional actors in French regions. The leading French firms and exnational champions in each region were responsible for setting up the pôles. For instance, without the weight and the investment in human time and effort on the part of the large French ICT firms, neither ‘Images et Réseaux,’ nor the IRT ‘B-Com’ would have seen the light of day. These kinds of projects require a lot of lobbying and a federating force. Without the leading institutional role played, particularly by local management of France Telecom-Orange, the IRT B-Com, for example, would not have happened. The region almost lost out on the IRTs, but thanks to a major push by local management of Orange, the region was finally awarded one of the eight IRTs. The large French firms, furthermore, dominated their early activities with ‘off-the-shelf’ projects, received most of the

subsidies for collaborative R&D and exploited them to realise their strategic objectives; such as boosting their own intellectual property. The operation of the pôles in the early years is, therefore, most consistent with Whitley’s (2007) ‘state-led’ and ‘integrated conglomerate’ NBIS typologies and suggests that, from the beginning, its activities were heavily shaped and constrained by the ideal-type institutions most commonly associated with French capitalism. The case also suggests that in the early years of the pôle, it was considerably difficult to overcome the ‘institutional inertia’ of non-collaboration and hierarchy, particularly involving SMEs. In ‘Images et Réseaux’, despite the presence of SMEs in collaborative R&D projects led by the large ICT firms, their participation was characterised by hierarchal rather than open, equal and collaborative alliance modes of coordination. The case study evidence presented has however made a considerable contribution to the literature, in highlighting that distinctive territorially embedded modes of coordination are emerging in France. Large firms and SMEs in ‘Images et Réseaux’ are not vertically integrated. The regional ICT sector has therefore been fairly successful in developing a performant tissue of dynamic SMEs and spin-offs which are nourished though their participation in the collaborative R&D projects of the pôle and which are able to develop without necessarily becoming suppliers to the large French ICT firms. In fact, that an SME or start-up will become a major long-term supplier to one of the large French ICT firms after collaborating with them in an R&D project, it was suggested, is very difficult to imagine. The development of horizontal modes of coordination in the Rennes-Bretagne ICT sector is, however, strongly linked to the evolving strategies of the large French ICT firms. Thus, large diversified conglomerates, such as France Telecom-Orange and Thomson-Technicolor, have restructured and exited from particular activities and markets, and in the process, small and medium-sized spin-offs have emerged, nourishing the regional ICT ‘ecosystem’ and making it less dependent on the activities of ‘integrated conglomerates’. This has contributed to developing a regional economy that is more dynamic and almost entirely not vertically integrated at all. The case of ‘Images et Réseaux’ highlights that since the early years of the pôles, genuinely horizontal modes of coordination, which did not exist at all prior to the creation of the pôles, have to a large extent replaced the more hierarchical relationships of the past. The collaborative modes of coordination which have come to characterise the pôles are, however, largely of the tacit type, and highlight the institutional constraints of a French NBIS which, as predicted by Whitley (2007), displays low levels of trust and alliance integration. For instance, as highlighted previously, there have been no concrete instances of technology transfer (during a collaborative R&D project of the pôle) from one of the large ICT firms which has resulted in the creation of an innovative start-up. The large French ICT firms use the collaborative R&D projects of the pôle in order to develop their own intellectual property and, for instance, in the case of the network operator, France Telecom-Orange, to gain access to new knowledge and ideas for new applications. The pôles de compétitivité is therefore an instance of collaborative coordination within the low-trust institutional constraints of the French NBIS. The financial risks associated with innovation are shared between the state and firms, while individual firms ‘internalise’ the potential rewards of the knowledge they generate through intellectual property protection. Despite the efforts by institutional actors in the city-region to develop and institutionalise collaborative relationships between the leading French ICT firms and local SMEs and start-ups, this has not been possible due to the constraints that these firms have in terms intellectual property protection. Therefore, in terms of support for innovative start-ups, encouraging innovative spin-offs, engaging in open collaboration and making their firms available to start-ups, the leading firms are

highly possessive of their intellectual property and careful to avoid unnecessary spill-overs. As already stated, the French telecommunications leader, France Telecom-Orange, is an exception, they have embraced ‘open innovation’ and have evolved their business model to use all standards, services and technologies. Their open innovation strategy is therefore aimed at exploiting the local ICT ‘ecosystem’ for the novel technologies and services it can potentially provide. The tacit type of collaborative knowledge generation apparent in the pôles suggests that horizontal forms of coordination around innovation have begun to emerge, however, when analysed in terms of Whitley’s (2007) ideal-type NBIS typologies, the nature and extent of collaboration in the pôles de compétitivité do not represent a significant departure from the archetypal ‘state-led’ national innovation system and a business system dominated by ‘integrated conglomerate’ and ‘compartmentalised’ firm-types. The mode of inter-firm coordination that has emerged through the pôles can be described as a formal, network mode of coordination. Thus, for SMEs, collaborating with the large French firms has given them access to highly tacit knowledge, such as in the management of R&D, the development of their business strategies, intellectual property protection, the commercialisation of their R&D and gaining market knowledge. However, the pôles have not unleashed a revolution in terms of technology transfer and the transfer of expertise between large and small firms, or in terms of the creation of innovative start-ups and spin-offs. Neither have they transformed the ideal-type institutions associated with France’s ‘state-led’ national innovation system in terms of how knowledge is produced, coordinated and transferred. For instance, ‘Image et Réseaux’ has had some success in the creation of start-ups emanating from university research laboratories, but no concrete examples of spin-offs or start-ups emanating from a collaborative research project involving a large ICT firm, as in the case of other high-tech regions such as Silicon Valley. The few number of start-ups emanating from research laboratories, also suggest that deeply embedded institutions governing the relationship between science and industry act as persistent constraints on technology and knowledge transfers. Generally, the movement of scientists and engineers from academia to industry is rare in France. In the case of the city-region of RennesBretagne, while the local university has put into place a number of measures encouraging researchers to strengthen their links with local industry, there have been few cases of academics leaving the university to join industry, or of researchers who have started their own start-up projects. Despite this being encouraged by the university, this phenomenon is rare and not at all planned in terms of the professional career development of researchers and technologists in France. The creation of regional technology research institutes (IRTs) which mutualise the R&D expertise of firms and public laboratories and which in the long-term ‘privatises’ the competencies and expertise of researchers from the public scientific realm, may, therefore, be viewed as an attempt to overcome some of the deeply embedded institutional constraints of the French innovation system. The institutions of the French ‘state-led’ innovation system, are more particularly, still evident in the close relationships that, for instance, the leading French ICT firms (in which the French state is an active shareholder) have with particular public research laboratories and the institutions around intellectual property protection. As suggested by the CEO of a leading Japanese R&D unit in the cityregion, this ‘non-market’ relationship between ex-national champions and the public research sector has created a particular institutional environment around intellectual property protection that a fully private, foreign firm, for example, which requires a more ‘arms-length’ and highly regulated intellectual property regime, would not accept. While the case of ‘Images et Réseaux’ confirms the dominant institutional role played by large French firms in their regional economies and the persistence of the institutions of the ‘state-led’, ‘integrated conglomerate’ NBIS, it also suggests that the relationship between large French firms and the

regional economies in which they have a historical presence has become increasingly complex, as former ‘national champions’ have been transformed into global multinationals; responding to competitive pressures by restructuring, changing the strategic focus of their business, attracting foreign institutional investors or merging with foreign MNEs, whilst simultaneously remaining institutionally ‘embedded’ in the French NBIS and playing their part as ‘good citizens’. This has significant implications for the nature of their territorial ‘embeddedness’, their participation in regional collective governance structures, such as the pôle, and their capacity for collaboration and alliance forms of coordination. The example of the Franco-American multinational Alcatel-Lucent is a case in point. Immediately after ‘Images et Réseaux was created, the French telecommunications equipment supplier, Alcatel merged with the American technology firm, Lucent Technologies. For regional actors, communication and dialogue with the new entity, Alcatel-Lucent, during this period became more complex, as the new firm became embroiled in a long-term restructuring of its activities. During this time it was difficult to explain to American managers what a pôle de compétitivité was and why it should invest in particular collaborative projects. Consequently, the business units of Alcatel-Lucent present in the region were not able to invest much in the research projects of the pôle in its early years. The newly created MNE group, however, had several units located in the region and the level of the firm’s involvement in the collaborative projects of the pôle was largely dependent on where managerial decision-making at each unit was located. Therefore, for instance, the subsidiary unit of the firm, Alcatel Lucent Enterprise (ALE) located in the city of Brest was able to participate in and lead important projects of the pôle as managerial decisions for this unit were made on site. This was not the case for the other units of the firm located in the region, where managers were based at divisional headquarters in Chicago or New Jersey. The example of Alcatel-Lucent and its involvement in ‘Images et Réseaux’ therefore highlights that certain leading ICT firms in the region face greater organisational and institutional constraints in developing collaborative, alliance forms of business relationships, particularly when the principal corporate R&D units are located outside of the region. These constraints become even more complex when a large, indigenous firm merges with a foreign multinational to become an institutional and organisational hybrid. Since the merger in 2006, Alcatel-Lucent has closed or sold off all, but one, of its business units in the region, however, the MNE has nevertheless remained an integral part of the regional ICT ‘ecosystem’. Over the years, its participation in the collaborative projects of ’Images et Réseaux’ have been non-negligible, have enriched the firm’s local R&D teams, and demonstrated that the group is committed to, and continues to view the region as an important source of its dynamic capabilities. The argument, most convincingly advanced by Hancké (2001, 2003), that the regional units of large French firms were the dominant institutional actors in the regionalisation of the French economy is therefore still relevant and applicable to firms in the ICT sector. The author, for instance, suggested that in the 1980s and 1990s regional units of the large French automakers, exploited institutional resources related to training, innovation and SME development in order to shift to post-Fordist mass production. Similarly, in the period leading up to, and immediately after, the creation of the pôles de compétitivité, international and increasingly ‘compartmentalised’ French conglomerates, operating in highly competitive and strategic global markets, have responded to increasing uncertainty and global competitive pressures by exploiting the institutional resources of the French NBIS in order to reinforce the regional dynamic capabilities upon which the competitive advantages of their regional units have been forged.

Whilst it can be argued that the leading international French firms are today less ‘embedded’ in their regional economies than the diversified ‘integrated conglomerates’ of the 1980s and 1990s; and in some cases have become considerably ‘dis-embedded’ by exiting certain activities and closing down certain territorial business units, the changing nature of their ‘embeddedness’ has contributed to the emergence of a territorialised, tacit, ‘network’ form of horizontal coordination in which multistakeholder regional economic governance structures, notably the pôles de compétitivité, play an increasingly important institutional and technological role. These new territorial governance structures suggest that in the coordination of innovation, French regional economies have indeed begun to forge modes of coordination which, as optimistically suggested by Aniello and Le Galès (2001), are beyond hierarchy, national champions, large firms and vertical coordination, and where their comparative advantages are derived from; experimental public policies at multiple spatial scales, the mobilisation of regional associations and public policymakers, horizontal modes of coordination and the dynamic capabilities of small and medium sized firms. The case-study of ‘Images et Réseaux’ however suggests that this phenomenon is a silent ‘evolution’, rather than the ‘revolution’ that the authors had expected. On the one hand, the leading French conglomerates have increasingly internationalised their innovative capabilities and restructured their activities to be more ‘compartmentalised’ and competitive. This has opened up some institutional opportunities for other firm and non-firm actors to shape governance mechanisms and forms of economic coordination and alliance integration in the regional cluster. The French state has furthermore maintained its active hold on innovation policy and regional development from afar, providing the institutional framework for experimental governance mechanisms to emerge while simultaneously reinforcing regional technological specialisms and regulating and standardising the form and nature of ‘institutional entrepreneurship’. Following the typologies of institutional change and experimentation advanced by Streeck and Thelen (2005), institutional transformation in the French ICT sector is therefore characterised by a form of ‘institutional layering’; creating and placing new institutional mechanisms on top of existing ones. Conclusions and Future Research The pôles de compétitivité are therefore clear instances of experimental forms of territorialised multilevel governance. The pôles are an experimental form of governance in that they include both ‘exogenous’ and ‘endogenous’ governance mechanisms (Crouch et al, 2004) and have formalised, institutionalised and reinforced the territorial nature of social and economic relations between and within the local scientific and professional communities. The descriptive case of ‘Images et Réseaux’ in the city-region of Rennes-Bretagne has highlighted that the pôles are important providers of what Crouch et al (2001, 2004) termed ‘local collective competition goods’ (LCCGs) and have reinforced the institutional infrastructure at the territorial spatial scale in France. The tangible and intangible LCCGs of the pôles have been instrumental in making French regional and local economies more dynamic and in the case of ‘Images et Réseaux’, have provided actors in the Rennes-Bretagne ICT sector with the opportunity to collectively construct a new future after a major external economic shock. Some caution does, however, need to be taken not to over-state the extent to which multi-level institutional experimentation, as is evident in the Rennes-Bretagne ICT cluster, is capable of radically renewing the region’s high-tech economy. Interviews with the executive manager of a major Japanese ICT conglomerate highlighted that whilst the regional ICT ‘ecosystem’ remains performant, it is losing steam. He also suggested that the region is indeed ‘locked into’ the technological specialism of digital video, networks and telecommunications, a domain that is fast approaching technological saturation and that the region has neither the skills, expertise or strategic foresight to

take advantage of new technological trajectories in order to withstand the next external economic shock. Thus, the development of the sector is following a highly path-dependent trajectory, which while producing a performant cluster around telecommunications and digital video and networks has not delivered the broader ICT revolution upon which other regional and national economies have recently capitalised. This is also particularly evident when examining the low level of FDI into the region’s ICT sector since the creation of the pôle. The policy tool of the pôles de compétitivité is therefore a clear instance of how the ‘national business system’ is fundamental in influencing, not only the diffusion of local agglomeration economies in France; in line with the redistributive and path-dependent logics of the ‘dominantdevelopmental’ French state, but in terms of the extent to which these new governance mechanisms in French regional economies can produce the radical transformation needed to stay abreast of global market developments, particularly in sectors characterised by radical forms of innovation and frequent technological ruptures. The ‘silent evolution’ that is indeed underway in the coordination of regional economies in France, as reflected in this emergent territorialised, multilevel mode of governance, has only tentatively taken root and is still highly dependent on the evolving strategies and ‘ownership coordination’ of the large French ICT firms, which over the years have moulded the ‘cluster’ and the regional ICT sector in line with their core competencies and which have reinforced, to a certain degree, narrow technological ‘lock-in’. The institutional experimentation that has been unleashed as a result of the pôles has furthermore taken the path of least institutional resistance; confined to the coordination of the innovative activities of firms and in the context of persistent national institutions shaping the way in which innovation and R&D are incentivised in France. Future research can consequently take two possible paths. The first, an in depth comparative study of another pôle in a similarly peripheral or semi-peripheral region, covering another technological domain and sector. The second, a more in depth analysis of the Rennes-Bretagne ICT cluster with a greater focus on human resources management, employment systems and the institutions governing labour markets in order to explore the extent to which firms in a more ‘coordinated market economy’ (CME) can excel in sectors characterised by radical innovation. Building on the work of Whitley (2007) and Allen (2013), an important empirical research agenda would therefore be to examine the extent to which territorial governance actors in CMEs are able to build novel institutions at the level of a region or sector in the development of a regionally distinctive ‘project network’ type business system. Whilst, there has been considerable, in depth comparative institutional research into high-tech sectors in Germany, in depth empirical studies of French high-tech local economies is sparse. The ICT sector in the city-region of Rennes-Bretagne is a particularly interesting regional case to study further as it suggests that ‘community’ or ‘cultural’ modes of governance may be an important factor in the coordination of the sector in the city-region which may not be found in Grenoble in the Rhône Alpes or Sophia Antipolis in the south of France. The emerging horizontal modes of coordination in the Rennes-Bretagne ICT cluster, for instance, suggest that a ‘project network’ business system type could emerge over time and it would be interesting to explore in greater depth, the extent to which macro-institutional constraints, particularly human resource management, labour market and employment institutions, prevail and how more ‘project-based’ modes of coordination could be reinforced. References Aniello, V. and Le Galès, P. (2001) ‘Between Large Firms and Marginal Local Economies: The Making of Systems of Local Governance in France’ in Crouch, C., Le Galés, P., Trigilia, C. and Voelzkow, H.

(2001) Local Production Systems in Europe: Rise or Demise? New York: Oxford University Press, 117153. Allen, M (2013) ‘Comparative capitalisms and the institutional embeddedness of innovative capabilities’, Socio-Economic Review, 11, 771–794. Amable, B. and Hancké, B. (2001) ‘Innovation and Industrial Renewal in France in Comparative Perspective’, Industry and Innovation, 8:2, 113–133. Birch, K., MacKinnon, D. and Cumbers, A. (2010) ‘Old Industrial Regions in Europe: A Comparative Assessment of Economic Performance’, Regional Studies, 44:1, 35-53. Boschma, R. (2004) ‘Competitiveness of Regions from an Evolutionary Perspective’, Regional Studies, 38:9, 1001-1014. Cohen, E. (2007) ‘Industrial Policies in France: The Old and the New’, Journal of Industry, Competition and Trade, 7: 3-4, 213-227. Crouch, C. (2005) Capitalist Diversity and Change: Recombinant Governance and Institutional Entrepreneurs, Oxford: Oxford University Press Crouch, C., Le Galés, P., Trigilia, C. and Voelzkow, H. (2001) Local Production Systems in Europe: Rise or Demise? New York: Oxford University Press Crouch, C., Le Galés, P., Trigilia, C. and Voelzkow, H. (2004) Changing Governance of Local Economies: Responses of European Local Production Systems, New York: Oxford University Press Crouch, C. and Voelzkow, H. (2009) Innovation in Local Economies: Germany in Comparative Context, Oxford: Oxford University Press Crouch, C., Schröder, M. and Voelzkow, H. (2009) ‘Regional and sectoral varieties of capitalism’, Economy and Society, 38:4, 654-678. Culpepper, D. (2001) ‘Employers, Public Policy, and the Politics of Decentralized Cooperation in Germany and France’ in Hall, P.A. and Soskice, D. (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press, 275-304. Deeg, R. and Jackson, G. (2007) ‘Towards a more dynamic theory of capitalist variety’, SocioEconomic Review, 5, 149–179. Essletzbichler, J. and Rigby, D.L. (2007) ‘Exploring evolutionary economic geographies’, Journal of Economic Geography 7, 549–571. Gertler, M.S. (2010) ‘Rules of the Game: The Place of Institutions in Regional Economic Change’, Regional Studies, 44:1, 1-15. Hall, P.A. and Soskice, D. (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press. Hancké, B. (2001) ‘Revisiting the French Model: Coordination and Restructuring in French Industry’ in Hall, P.A. and Soskice, D. (2001) Varieties of Capitalism: The Institutional Foundations of Comparative Advantage, Oxford: Oxford University Press, 306-334. Hancké, B. (2003) ‘Many Roads to Flexibility: How Large Firms Built Autarchic Regional Production Systems in France’, International Journal of Urban and Regional Research, 27.3, 510-26.

Lane, C and Wood, G (2009) ‘Capitalist diversity and diversity within capitalism’, Economy and Society, 38:4, 531-551. Lange, K. (2009) ‘Institutional embeddedness and the strategic leeway of actors: the case of the German therapeutical biotech industry’, Socio-Economic Review 7, 181–207. Levy, J.D. (1999) Tocqueville’s Revenge: State, Society and the Economy in Contemporary France, Harvard University Press. Levy, J.D. (2005) ‘Redeploying the State: Liberalization and Social Policy in France’ in Streeck, W. and Thelen, K. (2005) Beyond Continuity Institutional Change in Advanced Political Economies, Oxford: Oxford University Press, 103-124. Levy, J.D. (2006) ‘The State After Statism: From Market Direction to Market Support’ in Levy, J.D. (2006) The State After Statism: New State Activities in an Age of Liberalisation, Harvard University Press MacKinnon, D., Cumbers, A., Pike, A., Birch, K. and McMaster, R. (2009) ‘Evolution in Economic Geography: Institutions, Political Economy and Adaptation’, Economic Geography, 85: 2, 129-150. Maskell, P. and Malmberg, A. (2007) ‘Myopia, knowledge development and cluster evolution’, Journal of Economic Geography 7, 603–618. Peck, J. and Theodore, N. (2007) ‘Variegated capitalism’, Progress in Human Geography, 31:6, 731– 772. Rodriguez-Pose, A. and Storper, M. (2006) ‘Better Rules or Stronger Communities? On the Social Foundations of Institutional Change and Its Economic Effects’, Economic Geography 82:1: 1–25. Storper, M. (1993) ‘Regional “Worlds” of Production: Learning and Innovation in the Technology Districts of France, Italy and the USA’, Regional Studies, 27:5, 433-455. Streeck, W. and Thelen, K. (2005). ‘Introduction: Institutional Change in Advanced Political Economies’, in Streeck, W. and Thelen, K. (2005) Beyond Continuity Institutional Change in Advanced Political Economies, Oxford: Oxford University Press, 1-40. Ter Wal, A. and Boschma, R. (2011) ‘Co-evolution of Firms, Industries and Networks in Space’, Regional Studies, 45:7, 919-933. Trigilia, C. and Burroni, L. (2009) ‘Italy: rise, decline and restructuring of a regionalized capitalism’, Economy and Society, 38:4, 630-653. Whitley, R. (2007) Business Systems and Organizational Capabilities: The Institutional Structuring of Competitive Competences, Oxford: Oxford University Press.

Acknowledgements De Montfort University Scholarship 2012 (BAL FB1) Philip Almond, Professor of Comparative Employment Relations, Faculty of Business & Law, De Montfort University