Multinational Corporations and the Location of Technological ...

Regional Studies, Vol. 34.4 pp. 317± 332, 2000

Multinational Corporations and the Location of Technological Innovation in the UK Regions J O H N C A N T W E L L * and S I M ON A I AM M AR I N O ² *Department of Economics, University of Reading, PO Box 218, Whiteknights, Reading RG6 6AA, UK ² National Institute of Statistics (ISTAT) and Institute for International Affairs (IAI), Via Cesare Balbo 16, 00184 Rome, Italy (Received December 1998; in revised form April and August 1999) C A N TWEL L J. and I AM M A RI NO S. (2000) Multinational corporations and the location of technological innovation in the UK regions, Reg. Studies 34, 317± 332. In a rapidly globalizing economy, and particularly in the face of a process of economic integration such as that occurring in the European Union, regions forge an increasing number of linkages with other locations within and across national boundaries through the local technological development eVorts of multinational corporations (M NCs). By using patents granted to the largest industrial ® rms ± arranged by the region host to the research facility responsible ± the paper explores the location of innovative activities in MNCs, and the relationship between the pro® les of technological specialization of foreign-owned and indigenous companies in the U K regions. The results are consistent with the hypothesis that the pattern of MNC networks for innovation conforms to a hierarchy of regional centres, and that the pattern of technological specialization of foreign-owned aYliates in diVerent regional locations depends upon the position of the region in the locational hierarchy. Multinational corporations

Technological innovation

C A NT W EL L J. et I A M MA RI N O S. (2000) Les socieÂteÂs multinationales et la localisation de l’innovation technologique dans les reÂ gions du Royaume-Uni, Reg. Studies 34, 317± 332. Etant donneÂ la rapiditeÂ de la mondialisation de l’eÂconomie et face aÁ un processus d’inteÂgration eÂconomique, tel celui qui se produit au sein de l’Ue, les reÂgions eÂtablissent un nombre croissant de liens avec des emplacements situeÂs dans l’enceinte et au-delaÁ des frontieÁres nationales par les eVorts de deÂ veloppement technologique locaux des socieÂteÂs multinationales. A partir des brevets accordeÂs aux plus grandes entreprises industrielles ± classeÂs en fonction de la reÂ gion qui accueillit l’eÂtablissement de recherche en question ± cet article cherche aÁ examiner la localisation des activiteÂs aÁ caracteÁre innovatif des socieÂteÂs multinationales et le rapport entre les pro® ls de speÂcialisation technologique des socieÂteÂs Áa capital eÂtranger et domestique dans les reÂ gions du Royaume-Uni. Les reÂsultats correspondent aÁ l’hypotheÁse que la distribution des reÂ seaux d’ innovation conforme aÁ une hieÂrachie de centres reÂgionaux, et que la distribution de la speÂcialisation technologique des ® liales aÁ capital eÂtranger situeÂes dans diverses zones deÂ pend du positionnement de la reÂgion dans la hieÂrarchie des emplacements. SocieÂteÂs multinationales SysteÁmes reÂgionaux

Innovation technologique

Regional systems C AN T W EL L J. und I AM M A RIN O S. (2000) Multinationale Firmenkorporationen und der Standort technologischer Innovation in den Regionen des UK, Reg. Studies 34, 317± 332. In einer zunehmend globale Ausmaû e annehmenden Wirtschaft, und besonders angesichts des Vorgangs wirtschaftlischen Zusammenschlusses wie er z.Zt. in der EU statt® ndet, gehen Regionen mit Hilfe der technologischen Entwicklungsanstrengungen multinationaler Firmenkorporationen (multinational corporations5 M NCs) eine wachsende Anzahl von Verbindungen mit anderen Standorten innerhalb und uÈber Staatsgrenzen hinweg ein. Mit Hilfe von Patenten, die sich im Besitz der groÈû ten Industrie® rmen be® nden, untersucht dieser Aufsatz ± dank Vereinbarung mit der Gastregion, die fuÈr die Forschungsregion verantwortlich ist, ± den Standort innovativer Unternehmungen in M NCs, sowie die Beziehung zwischen den Pro® len technologischer Spezialisierung einheimischer und in auslaÈndischem Besitz be® ndlichen Gesellschaften in den oben erwaÈhnten Regionen des U K. Die Ergebnisse stimmen mit der Hypothese uÈberein, daû das Muster der MNCs Netze fuÈr Innovation einer Hierarchie regionaler Zentren entsprechen, und daû das Muster technologischer Spezialisierung an Auslandsbesitz be® ndlicher Schwesterbetriebe in verschiedenen Regionalstandorten von der Stellung der Region in der Standorthierarchie abhaÈngt. Multinationale Firmenkorporation Technologische Innovation Regionalsysteme

0034-3404 print/1360-0591 online/00/040317± 16 ©2000 Regional Studies Association

318

John Cantwell and Simona Iammarino I N T RO D UC T I O N

Claims about the existence of regional diVerences have provided grounds for many recent analyses of economic development in the European Union (EU), e.g. Q UAH , 1996; F AG ER B ER G and V E R S PAGE N , 1996; M A L M B ER G et al., 1996; F AGER BE R G et al., 1997; C A NI EÈ L S , 1997; B R E S C H I , 1997. Somehow surprisingly, however, too little is yet known at the regional level with respect to the geographical dispersion of technological innovation in the EU, despite its undoubted in¯ uence upon interregional growth rates across sectors. The ® rst aim of this paper is to look at the geographical distribution of innovative activities carried out by large industrial multinational corporations (M NCs), both local and foreign, in the UK regions. This might help identify the basic factors underlying regional diVerences which have characterized the country for over half a century, and particularly in the most recent decades. The second goal of this paper is to examine whether the concentration of technological activity by foreign aYliates is correlated to the concentration of the same activity carried out by local ® rms. The assumption is that regional agglomerations of knowledge and capabilities attract foreign direct investment (FDI) in R&D to a diVerent extent and with a diVerent sectoral spread, depending upon the position of the region in a locational hierarchy, which can be established both within and across national boundaries. The paper is divided into ® ve sections. The ® rst section addresses the issue of the globalization of technology and its implications for the competitiveness of multinational ® rms, regions and national states. Some previous evidence on how and to what extent innovative activity by M NCs is internationally dispersed and regionally concentrated ± with reference to a similar investigation for Italy ± is provided in the second section. The third section explores the general features of the location of innovation of both indigenous and foreign large ® rms in the UK regions, by using a database of patents granted to the largest ® rms by the US Patent and Trademark OYce (US PTO) as an indicator of advanced technological capacity. Our attempt to test the relationship between foreign-owned and UK multinationals’ technological pro® les, in order to clarify the concept of `regional hierarchy’ across E U regions, is reported upon in section four. Finally, a ® fth section presents some concluding remarks and highlights some further questions for future research. G L O B A L I Z AT I O N O F T E C H N O L O G Y A ND COMP E T IT IV E NES S Attention has been increasingly focused on the emergence of the trend for M NCs to establish internal and external networks1 for innovation (e.g. C A N T WE L L ,

1995; D U NN I NG , 1995; H AGED OO R N and N A RU LA , 1996; S AC H WAL D , 1998; C A NT W EL L and J A N NE , 1999; K U EM M ER L E , 1999; Z A N D ER , 1999), which are characterized by diVerent levels of territorial and social embeddedness with reference to the location which hosts them. Thus, the development of cross-border corporate integration and intra-border inter-company sectoral integration, as new forms of global governance, makes it increasingly important to examine where and how innovative activity by MNCs is internationally dispersed and regionally concentrated. The globalization of innovation refers not only to the ùse’ of M NC technological competencies but, even more, to the way they are generated. As argued in our previous work (C A N TW EL L and I A M M A R I N O, 1998, 1999), by `globalization’ we refer to a high degree of interdependency among units which constitute the M NC.2 By taking into account the geographical dimension of the globalization, therefore, it should be clear that this de® nition implicitly assumes also an interdependence between the global and the local. As accurately emphasized by D I C K EN , 1994, `global and local are not ® xed scales; rather, they represent the extreme points of a dialectical continuum of complex mutual interactions’ (p. 103). Thus, structure and behaviour of the two èxtreme points’ need to be considered within the context of their increasing interdependence, including both endogenous and exogenous variables relevant to the analysis. What is the meaning of the nexus between `global’ and `local’ for ® rms, states and regions? Although the viewpoint changes substantially according to the unit of analysis, competitiveness and competition seem to underlie all three levels. To the multinational ® rm, the global± local question with regard to innovative activities revolves mainly around the choice of location, a choice that has important consequences for the access to new technological competences. The choice clearly depends upon both the strategy followed by the MNC ± i.e. the type of network through which it develops technological specialization (upgrading or diversifying it) ± and on the characteristics of alternative innovative contexts where research may be located. In line with evolutionary approaches to innovation and industrial organization, we have shown elsewhere (C A N T WE L L and I A M M A R I N O, 1998) that MNC networks for innovation and location choices conform, particularly within the EU, to a geographical hierarchy of centres established across and within national boundaries, and which appears to be extremely sensitive to the characteristics of regional systems. Accordingly, the technological specialization of foreign-owned aYliates in diVerent regional locations depends upon the position of the region in the hierarchy ± i.e. whether the regional system (and its technological system) is at the top of the hierarchy (higher order location) or is a lower order regional centre. From the viewpoint of the ® rm, the aim is to

Multinational Corporations and the Location of Technological Innovation in the UK Regions increase its global technological advantage from selected foreign sources (C A N T WE L L , 1989, 1992b, 1994; C A NT WE L L and H O D S O N , 1991). Traditional location advantages (static endowments) are increasingly substituted by intangible assets, i.e. localized accumulated knowledge, skills and expertise, on which the M NC builds and reinforces its competitiveness on global markets. Therefore, in the economic geography of innovation, growing attention has to be due to the interactions between the `knowledge-seeking’ M NC (F O R S , 1998) and the local host context, rather than to the two entities considered separately. To the state, the problem of global versus local is rather diVerent, as the multinational ® rm is only one of the major channels through which the forces of globalization are transferred and disseminated (D I C K EN , 1994). The new conditions for competitiveness imply a stronger role of the state in sustaining the competitive advantage of the country and increasing its share of value added by encouraging the most appropriate kinds of local technological specialization, depending upon the particular kinds of local skills and institutions. In fact, the degree of geographical diVerentiation is likely to augment, due to the `focused’ orbit of M NCs’ networks for innovation, which set in motion (or reinforce) virtuous and vicious circles. The distributive side of globalization processes is thus concerned, since the state has to deal with diVerent economic and technological paths of development within its own boundaries, while pursuing the objective of overall country competitiveness (see, among others, H I R S T and T H O M PS O N , 1996; D I C K EN , 1994; A R CH I BU G I and I A M M A R I N O, 1999). However, insofar as inter-country competition for inward FDI in innovation is concerned, the best means of attracting foreign-owned research is an already strong local tradition in innovation in the sector in question (D I C K EN , 1990; C A N T WE L L , 1992a; D U NN I N G , 1996). Thus, the role of governments is to foster the development of those sectors in which the local potential for innovation is greatest, and which are in the best position to take advantage of international knowledge ¯ ows within and between MNCs, to facilitate the codevelopment of locally specialized expertise and that of foreign centres with complementary ® elds of technological specialization. Therefore, it can be argued that in the global± local interaction, the key-role of the state should be that of intermediary. In a policy perspective, only at the country level is it feasible to devise new forms of international knowledge exchange that complement local capability and hence enhance competitiveness, or create it in the case of backward regions, i.e. in cases in which the lack of likelihood of entirely domestically-driven start-ups calls for intervention capable of reversing the negative causation that would otherwise be ampli® ed by globalization processes, by helping to ® nd the most

319

appropriate outlet for unexploited potential skills and expertise.3 To the region, the issue global versus local manifests itself in the ability of local actors to construct appropriate linkages with complementary but geographically distant sources of knowledge and technological competence, through which interchange they can enhance their own knowledge base.4 We use the word ènhancing’ and not `building up’ because the existence of such a base ± in other words, that the region possesses a minimum threshold stock of technological competence ± is a necessary, although not suYcient, prerequisite behind the capacity to become part of an international network of specialized technological centres, linked by knowledge ¯ ows especially within M NCs. The suYcient condition to take part, as a principal actor, in the `dialectical continuum of complex mutual interactions’ is, instead, the dynamics of regional technological advantage as specialization increases, and the ability to upgrade the quality of local competence. In the regional system, i.e. which satis® es the prerequisite of having a suYcient technological threshold, transformations to cope with a greater intensity of international knowledge ¯ ows between centres of excellence mean, ® rst of all, an awareness of the change in MNC strategy towards greater cross-border integration. Innovative flows are driven both by foreignowned ® rms attracted to the region for some speci® c competence or for more general expertise, and by locally-owned ® rms which try and tap into knowledge sources abroad. The impact of these ¯ ows on both the ® rm and the region clearly depends upon the technological pro® les and strategies of MNCs, as well as upon the characteristics of the regional system of innovation and its local areas of specialization. M N C s A N D RE G I O N A L C E N T RE S I N T H E E U: M E T H O D O L O G I C A L I S S UE S A N D S O M E P RE V I O US E V I D E N C E Until recently, comparatively little research has investigated the subnational concentration of technological activities in Europe, mainly because of the lack of regional data on innovation. Several studies exist for the US (among others, M A L EC K I , 1980; J A F FE et al., 1993; A UD R ET S C H and F EL D M A N , 1996; F E L D M A N , 1994), which observe the concentration of innovative and economic variables both across regions and sectors. Using European Patent OYce (EPO) data, B R ES C H I , 1995, 1997, ® nds evidence of considerable diVerences in the spatial concentration of innovation across sectors, while C A N I EÈ L S , 1997, shows that the top 12 regions account for 70% of the total number of patents in the period 1986± 90. However, for the most part, both theory and empirical analysis have paid only scant attention to the nationality of ownership of innovative activity or, in other words, to the particular characteristics of foreign-

320

John Cantwell and Simona Iammarino

owned as opposed to domestically-owned technological activities.5 The completion of the Single European Market and the pursuit of economic and monetary union have spurred the reorganization of operations of MNC aYliates located in the E U to a much greater extent than in the case of aYliates based outside the area (A M I N , 1992; C A N TW EL L , 1987, 1992c; D I C K EN , 1994). C A NT W EL L , 1992a, has shown that the degree of interdependency among the geographically distinct units which constitute the M NC is relatively higher in Europe, in which intra-® rm networks and linkages between foreign-owned aYliates and local ® rms turn out to be far more entrenched than in other areas of the world. The intra-area rationalization of economic activities boosted by the integration process has thus given a strong impetus to the globalization process, thereby strengthening the link between global and local through the formation of cross-border corporate networks. To examine the regional level in Europe we can make use of the Eurostat scheme of classi® cation ± the Nomenclature of Territorial Units for Statistics (NUTS). The NUTS classi® cation is based on the institutional divisions currently in force in the Member States, according to the tasks allocated to territorial communities, to the sizes of population necessary to carry out these tasks eYciently and economically, and to historical, cultural and other factors (E U R O S TAT , 1995). However, by choosing the appropriate NUTS level in each case, inter-country diVerences can be compensated, at least in part. The empirical investigation uses patents granted to the world’s largest industrial ® rms for inventions achieved in their European-located operations, classi® ed by the host European region in which the research facility responsible is located. The use of corporate patents as an indicator of advanced technological capacity is one of the most established and reliable methods of estimating the cross-sectional patterns of innovative activities. The advantages and disadvantages of using patent statistics are well known in the literature (see, among others, P AVI T T , 1988; G R I L I C H ES , 1990; A R CH I BU G I , 1992) and will not be rehearsed here. For our purposes, the use of patent records provides information both on the owner of the invention (from which we have derived the country of location of the ultimate parent ® rm through a consolidation of patents at the level of international corporate group) and, separately, the address of the inventor, thus allowing the identi® cation of where the research and development underlying the invention was carried out in geographical terms. The regionalization of our patent database consists of attributing a revised, although still compatible, NUTS code (up to level 3) to each patent recorded, according to the location of inventors in the EU countries, with reference to the period 1969± 95. Moreover, patents can be classi® ed by detailed technological ® elds (grouped into 56 sectors in the database,

see Appendix), which would not be otherwise possible by using indicators such as, for example, R& D expenditure (Z A ND E R , 1999). This is particularly appropriate here since the aim is to look at the regional concentration of the innovative activity of MNCs located in the EU, and to examine its sectoral distribution at a subnational level. Finally, the choice of US patenting is convenient, since large ® rms are especially prone to patent their best quality inventions in the US market, the largest and the most technologically advanced, following more extensive testing in their respective home markets. It is therefore more likely that our data re¯ ect the patenting of inventions that have a signi® cant commercial importance, as well as allowing for a meaningful analysis of the territorial distribution of the technological operations of MNCs in the European Union. M N C R& D L O C A T I O N I N T H E U K RE G I O NS The regionalization of the University of Reading patent database has been extended to cover the case of the UK.6 The latter diVers substantially from the Italian case, both in terms of the magnitude of the phenomenon under investigation and in terms of the kinds of government policies adopted to support regional attractiveness.7 Figs 1± 3 report, with reference to the year 1995 (the last year of the period considered in the Reading database) some selected indicators by region. Not surprisingly, all the indicators point rather clearly to the strong polarization of innovative activities in the South East of England,8 even relative to the demographic size of the region (almost 30% of total population) and its economic weight (35% of national value added in 1995). In terms of R& D personnel employed in the business sector (Fig. 1), the South East represents 42% of the total, followed by the North West, with 13%, and the South West (9%), while the concentration of R &D expenditure of the business sector (Fig. 2) in the South East reaches 49% of the national expense. The indicator of technological output (Fig. 3) presents a more evenly spread territorial distribution with regard to all E PO requests, of which the South East accounts for 34% and the North West over 15%, but the share of patents granted in 1995 by the US PTO to large corporate inventors in the South East is over 52%, followed by the North West (11%) and the West Midlands (8%).9 Regional gaps arise not only in existing technological stocks and relative size, but also in dynamic terms, endorsing the view of a widening of regional diVerences within the country. Indeed, C H AT T ER JI and D EWH U R S T , 1996, estimating a model using nominal GDP per capita by NUTS 2 region over the period 1977± 91, have shown that the UK counties (and regions) tend to diverge relative to the GDP per capita of Greater London (South East).10 It has been

Multinational Corporations and the Location of Technological Innovation in the UK Regions

321

Fig. 2. R&D expenditure: business sector, 1995 Source: Eurostat, 1997.

Fig. 1. Population (3

1,000) and R& D personnel (absolute nos.), 1995 Source: Eurostat, 1997.

shown that diVerences in innovative capabilities across European regions ± which are even more pronounced than at a purely country level ± account for a good deal in explaining divergent trends in economic growth (F AGER B E R G and V ER S PAGEN , 1996). Insofar as globalization spurs agglomeration in regional centres, the tendency towards divergence might thus be reinforced. It should be noted, however, that the fact that regional inequalities increase with the level of geographical breakdown is not surprising. This is the consequence of the tendency of economic and, to an even greater extent, innovative activities to agglomerate, and of forces which concentrate key-functions in core centres, such as Greater London in the UK, driving out other activities with lower value added content. At a high level of geographical aggregation (namely the country level), such diVerences often average out (D U NF OR D , 1996). Turning more speci® cally to the geographical distributions of innovation generated by large ® rms, Table 1 reports, for the whole period 1969± 95, the shares of US patents (relative to the UK as a whole) of both the largest UK-owned ® rms and the largest UK-located foreign-owned ® rms, attributable to the research of each located in the UK regions.

Fig. 3. EPO requests and US PTO patents, 1995 Source: Eurostat, 1997; University of Reading database.

322


Table 1. Shares of US patents of both the largest UK ® rms and the UK-located foreign-owned ® rms, attributable to research in UK regions relative to the UK as a whole, 1969± 95 (%) Regions North Yorkshire and Humberside East Midlands East Anglia South East South West West Midlands North West Wales Scotland Northern Ireland Total UK Total UK (absolute nos.)

UK ® rms 4´0 4´4 6´3 1´5 40´2 5´7 16´4 17´0 2´4 2´0 0´1 100 23,404

Foreign ® rms 2´3 3´9 4´5 5´3 60´8 4´2 3´6 7´8 3´4 4´0 0´3 100 11,815

Total 3´4 4´3 5´7 2´7 47´1 5´2 12´1 13´9 2´7 2´7 0´1 100 35,219

As already highlighted, the geographical agglomeration of innovation is remarkable: more than 73% of the total research activity carried out by large ® rms in the UK is concentrated in three regions, namely the South East, with the 47´1% of the total, the North West (13´9%) and the West Midlands (12´1%). Foreignowned research appears to be even more singularly concentrated than that of UK ® rms, with 60´8% of total research undertaken by foreign-owned aYliates located in the South East (well above the indigenous company share of 40´2%). Both the North West and West Midlands, though, exhibit regional shares de® nitely higher for indigenous companies (17% and 16´4% of the total respectively) than for foreign-owned ® rms (7´8% and 3´6% of the total UK-located research of foreign-owned aYliates). However, it is interesting to note that, contrary to what is observed in the Italian case, the capacity to attract externally-owned innovation activity seems to be relatively higher in a number of regions which show the lowest overall shares. This is particularly the case of East Anglia, with a share of foreign-owned patents almost quadruple the equivalent for UK-owned ® rms (the regional share of research of foreign-owned aYliates is 5´3%, compared to 1´5% for indigenous ® rms), but a similar proportionately greater attraction of foreign-owned research also occurs in the most peripheral regions of the UK outside England, i.e. Wales, Scotland and even Northern Ireland (whose regional share of the overall total is anyway negligible). It is worth mentioning here the role played by region-speci® c public science externalities, which may be proxied by the Research Council Grants to higher education institutions (HEIs), funded through the Government’s OYce of Science and Technology (OST). Universities and other HEIs are the main providers of basic research and much of the strategic research carried out in the UK. The role played by the

Fig. 4. Government research grants to higher education institutions by region 1994± 97 (% of U K total) Source: Higher Education Statistics Agency.

government in strengthening the regional science base and maximizing its contribution to economic performance ± by providing the core general funding ± is therefore crucial in shaping the pattern of local attractiveness towards external resources. Fig. 4 shows the regional shares of research grants for the period 1994± 97, con® rming once more the primacy of the South East, with 40% of total grants, followed by Scotland (12´4%), the North West (8´6%) and East Anglia (8´4%). The attractions of the science base in Scotland and East Anglia may help to explain how the technological eVorts of foreign-owned ® rms are relatively more drawn to these regions than the spatial distribution of indigenous corporate R& D would suggest. Particularly in the case of Scotland, government eVorts to improve wealth-creating potential of basic research, giving rise to positive spillovers, may help explain the relative higher share of foreign-owned technological development located in the region. The high geographical concentration of innovation generated in MNCs compels us to restrict our analysis at a more detailed level to the three regions above mentioned, i.e. the South East, the North West and the West Midlands, since the absolute numbers of patents granted in the other regions are too low for meaningful statistical analysis. It is necessary to bear in mind, however, the diVerences in the degree of attractiveness of external resources, already highlighted, which mark out these three regional systems as against

Multinational Corporations and the Location of Technological Innovation in the UK Regions Table 2. Foreign shares (foreign-owned ® rms’ percentage of total patents granted to large ® rms for local research in the U K), by sector and region, 1969± 95 Regions Sectors 1 3 5 6 7 9 11 12 13 14 16 17 20 23 25 28 29 33 34 35 36 37 38 39 40 41 42 49 50 51 53 56 Total, 56 sectors

South East 14´6 35´1 40´1 87´9 48´2 27´6 42´4 43´3 39´0 58´3 33´3 45´4 49´5 65´2 25´9 45´7 27´6 51´4 59´3 34´6 61´8 33´1 50´3 46´7 74´4 56´7 24´4 17´6 35´4 25´0 41´9 44´5 43´3

West Midlands North West 13´3 18´5 12´6 60´0 13´2 6´9 11´1 47´2 23´6 18´3 27´4 16´2 20´8 6´8 0´0 18´3 3´9 4´5 10´1 16´7 28´6 4´9 6´2 3´6 11´1 8´8 3´6 4´8 13´2 0´0 15´0 2´6 10´0

35´7 32´8 19´9 54´3 22´3 18´4 18´9 26´2 22´5 11´2 25´0 6´1 21´6 33´3 38´5 31´4 20´2 15´4 14´3 0´0 50´0 14´3 9´7 6´0 57´9 20´0 33´3 6´5 4´8 3´0 15´9 8´1 18´8

UK 20´7 23´2 31´5 80´0 38´5 28´8 34´4 41´5 28´5 40´1 29´6 32´1 44´4 50´0 37´6 38´2 16´0 44´8 47´8 31´5 58´4 28´3 37´3 27´7 62´1 49´9 17´6 15´9 23´1 17´9 33´3 28´4 33´5

Notes: The sectoral codes are given in the Appendix.

all others in the UK, and thus lend support, at least at ® rst glance, to our hypothesis of the existence of an internal geographical hierarchy. The sectoral distribution of patents demonstrates interregional diVerences to an even greater extent. Table 2 reports the foreign shares by sector and region (foreign-owned ® rms’ percentage of total patents granted to large ® rms for research located in the region).11 As expected, Table 2 shows that, for the period 1969± 95, the highest contribution of foreign research to the regional total is recorded in the South East (43´3%), which is much above the national average (33´5%). The other two regions are far less relatively attractive, with foreign shares accounting for almost 19% of regional research in the North West and 10% in the West Midlands. The most pronounced and consistent contribution of foreign-owned research to the regional total is found, in the case of the South East, in sectors identi® ed as various categories of electrical equipment, such as

323

semiconductors (40), image and sound equipment (36), other electrical communication systems (34), telecommunications (33) and electrical devices and systems (38). As for the country as a whole, the foreign share is also above average in all these sectors, but in the same macro-sector the other two regions show foreign shares signi® cantly higher than the average only in image and sound equipment (36) and, for the North West, in semiconductors (40). Foreign shares of regional and national patents are signi® cantly below average in all cases in synthetic resins and ® bres (9), rubber and plastic products (49), and coal and petroleum products (51). In contrast, some examples of relatively large foreign shares in all regions, and in the country as a whole, are found in photographic chemistry (6) ± in which the contribution of foreign research to the sectoral patenting activity is particularly remarkable in the South East (88%) and in the UK total (80%) ± and in cleaning agents and other compositions (7). As expected, and in line with what was observed for Lombardia in Italy (C A NT WE L L and I A M M A R I N O, 1998), the sectoral distribution of foreign-owned companies’ shares of regional and national research activity shows a greater degree of correspondence between the South East and the UK than between the UK as a whole and any other individual region. Such higher order regions have a greater overall volume of activity spread across a wider representation of the ® elds of innovation that characterize a country as a whole. This might be illustrated with reference to the cases of mining equipment (23), oYce equipment and data processing systems (41) ± both of which show a foreign share over 50% of either regional or national total research ± and miscellaneous metal products (14), in which the core region and the country as a whole have foreign shares well above the average. The other two regions, as in the case of Piemonte in Italy, seem to indicate more focused and locally speci® c sectoral patterns in the distribution of foreign participation in local research. In particular, in the West Midlands foreign-owned research makes a substantial contribution in pharmaceuticals and biotechnology (12), in which ® eld the foreign share for both the country as a whole and the North West is above average as well, and the broad industrial groups of metals and metallurgy ± i.e. metallurgical processes (13) and miscellaneous metal products (14) ± and mechanical engineering ± i.e. chemical and allied equipment (16), metal working equipment (17) and assembly and material handling equipment (20). Technological innovation in mechanical engineering also receives a particularly high contribution by foreignowned aYliates in the North West, although it appears principally directed to sectors, such as mining equipment (23), textile and clothing machinery (25) and other specialized machinery (28) ± a diVerent set of engineering ® elds from those in which foreign-owned

324


® rms are most active in the West Midlands. However, it is worth noting that these two regions manifest foreign shares that are well above average in some sectors for which the South East and the UK show instead relatively low values, such as (for the North West) food and tobacco products (1) and inorganic chemicals (3). A noteworthy discrepancy between the North West and the other cases illustrated is in internal combustion engines (42), in which foreign-owned aYliates account for one-third of regional patenting activity, substantially above foreign participation in research in this ® eld in other UK regions. Finally, Figs. 5 and 6 report the evolution over time

Fig. 5. Patents by year: UK total, 1969± 95

Fig. 6. Foreign shares (% of total patents) by year and region, 1969± 95

of, respectively, the UK-owned and foreign-owned ® rms’ total numbers of patents and the foreign shares for the 56 sectors as a whole by region. From Fig. 5, it emerges clearly that the number of patents granted in the US to UK-owned ® rms has been on a continually declining trend, and has seen a sharp downturn since the end of the 1980s. By contrast, the number of patents granted to foreign-owned ® rms for research carried out in the UK has remained fairly stable, which has led foreign-owned aYliates to overtake indigenous ® rms, by the end of the period considered, in the absolute magnitude of technological eVort that they undertake in the country. The foreign proportion of total research located by large ® rms in the South East (Fig. 6), having been stable for most of the period, has increased markedly in the most recent years, reaching the highest values (over 70%) at the end of the period. This pattern is also evidently driving that of the country as a whole, which recorded a foreign share of approximately 60% in 1995, more than double the equivalent share in 1969. This is due mainly to the decline of indigenous innovative activities, which is observed more markedly at the country level than for any of the three particular regions considered. The foreign share of regional patents in the West Midlands and the North West shows, in both cases, a greater ¯ uctuation, increasing moderately over the period 1969± 95 as a whole, with some sharp decreases in the mid-1980s for the West Midlands and at the beginning of the 1990s for the North West. However, both regions end with shares in the ® nal year that are about three times bigger than those registered in 1969. In general terms, the above picture seems to con® rm, in the UK case as in Italy, that the location of technological activities of foreign-owned MNCs tends to be strongly agglomerated at a subnational level. In a pattern that also matches the dominant role of Lombardia in Italy, foreign-owned technological development in the UK is concentrated in the same major region (the South East) that has served as the main pole of attraction for the equivalent activity of indigenous ® rms. Indeed, foreign-owned innovative activities are even more locationally concentrated in the South East than is domestically-owned innovation, as demonstrated by the high foreign share of total patents in that region. Moreover, it might be possible to interpret the remarkable decline in the patenting activity of UK-owned ® rms from their home base either by a narrowing of UK corporate technological specialization, or by the renewed increase in the internationalization of research by UK-owned companies in the early 1990s from already high levels (C A NT WE L L and H A R D I NG , 1998). Allied to a further relocation of UK-owned corporate research abroad as an explanation of why foreignowned ® rms have overtaken local ® rms as creators of technological innovation might be the form taken by newer inward asset-seeking investment, namely the

Multinational Corporations and the Location of Technological Innovation in the UK Regions acquisition of UK ® rms by foreign M NCs. In other words, some of the relative shift in the patenting activity from UK to foreign M NCs may simply re¯ ect a change in the ownership of R &D facilities through acquisition. However, this would need a closer analysis of the sectoral composition of the fall in UK-owned company patenting, by comparison with the recent rise in technological development abroad by British-owned ® rms, which lies beyond the scope of this paper. P RO F I L E S O F T E C H N O L O G I C A L S P E C I A L I Z A T I O N O F F O RE I G NOWN E D A N D I ND I G E NO US F I RM S : T H E RE G I O N A L H I E RA RC H Y To investigate the characteristics of the pro® les of technological specialization of foreign-owned and local ® rms we have used the revealed technological advantage index (RTA), a proxy for technological specialization ® rst applied in the country context by S O ET E , 1987, and subsequently developed and extended to the analysis of company patterns by C A NT WE L L , 1989, 1991, 1993; and P AT EL and P AV I T T , 1991. The RTA index of a region in a particular technological sector is given by its share of total world patenting. Therefore the RTA index is de® ned as follows: RT Ai j 5

(Pi j /Pwj )/S

j

Pi j /S

j

Pwj )

where i5 j5

1, 2, 312 (1) 1, . . . , 56

where: Pij 5 the number of patents of region i in sector j Pwj 5 the number of world patents in the same sector. The index has been calculated for both UK-owned and foreign-owned ® rms in the three regions ± the South East, the North West and West Midlands ± and in the UK as a whole for the period 1969± 95. It gives a measure of the performance of (nationally-owned or foreign-owned) ® rms in a region in one particular sector relative to all large ® rms in the world as a whole. Since the R TA index varies around unity, an index greater than one suggests a comparative technological advantage, whilst an R TA below unity indicates a position of comparative disadvantage. Table 3 reports the R TA values by region and sector and it can be compared with Table 4, which shows the revealed scienti® c advantage (RSA) index by scienti® c sector and region. The latter index has been built on an equivalent basis using the values of research grants to HEIs awarded by the Government in 1994± 97 classi® ed by the location of each HEI and by the scienti® c ® eld of activity, and it gives an account of the scienti® c specialization of publicly-funded research in each region relative to the UK as a whole. First of all, the overall picture seems to indicate that the technological advantages of both the nationallyowned and foreign-owned ® rms in the South East and in the country as a whole are more widely dispersed at

325

the sectoral level than in the other two regional cases (the cross-sectoral variance in the R TA index is lower for the South East or the UK than it is for either the West Midlands or the North West). In both the South East and the UK as a whole, the technological specialization of foreign-owned aYliates overlaps most with that of local ® rms in the areas of pharmaceuticals and biotechnology (12) and special radio systems (35). In the latter sector, the fact that foreign shares appear to be well below the average in both cases suggests a point of strength of local technological innovation. Looking at scienti® c advantages (Table 4), the South East turns out to be rather specialized in medical sciences and electrical, electronic and computer engineering, whilst regional public research is relatively weak both in pharmacy, pharmacology and biosciences and in information technology and system sciences. This helps to suggest the form of local linkages between science and the development of corporate technological competences. In other words, foreign MNCs are probably much more attracted by the university general expertise in some broader and related scienti® c areas ± such as medical sciences with respect to pharmaceuticals and biotechnology ± rather than by speci® c capabilities in their own immediate ® eld of expertise. The latter may well be preferably developed and strengthened ìn house’ in large ® rms, and thus funded to a much greater extent by the private sector through commercial research, as well as through the international linkages of the MNC (cross-border corporate integration, which also provides indirect links to the best university research in the biosciences themselves on a worldwide basis). Other examples of a comparative advantage of local ® rms in the core region and in the country as a whole are found in other organic compounds (11) and other specialized machinery (28). The South East alone attracts foreign research also in sectors such as illumination devices (37), other instruments and controls (53) and other manufacturing and non-industrial (56), whilst the country as a whole shows a greater matching between indigenous and foreign ® rms in cleaning agents and other compositions (7), in all of which the foreign technological contribution turns out to be substantial. As far as the other two regions are concerned, the technological specialization of both local and foreign ® rms is more concentrated within broad industrial groups. In the West Midlands, local ® rms show a strong comparative advantage in metals and metallurgy (i.e. metallurgical processes (13) and miscellaneous metal products (14)) and mechanical engineering (i.e. other specialized machinery (28) and, particularly, other general industrial equipment (29)) in which foreign-owned aYliates are also relatively specialized. The latter sector, with the most remarkable value of the R TA, is indeed one of the traditional points of strength in the regional production structure, and despite the attraction of

326

John Cantwell and Simona Iammarino Table 3. RTA index (U K and foreign ® rms) relative to the world, by sector and region, 1969± 95 Regions South East

Sectors 1 3 5 6 7 9 11 12 13 14 16 17 20 23 25 28 29 33 34 35 36 37 38 39 40 41 42 49 50 51 53 56 Total, 56 sectors Standard deviation

West Midlands

North West

UK

UK ® rms

Foreign ® rms

UK ® rms

Foreign ® rms

UK ® rms

Foreign ® rms

UK ® rms

Foreign ® rms

2´742 0´648 0´870 0´230 6´659 0´747 1´306 3´480 0´751 0´678 0´899 0´617 0´541 0´184 0´342 1´089 1´279 0´978 0´930 3´464 0´575 1´956 0´852 0´575 0´343 0´482 1´354 1´185 0´689 1´230 1´123 1´032 1´000 0´807

0´616 0´458 0´765 2´198 0´804 0´373 1´262 3´482 0´630 1´245 0´587 0´671 0´694 0´452 0´157 1´200 0´639 1´354 1´778 2´406 1´219 1´267 1´131 0´661 1´307 0´826 0´573 0´332 0´495 0´538 1´064 1´085 1´000 0´685

0´416 0´555 0´867 0´024 0´368 0´372 0´187 0´115 1´424 1´573 0´541 1´540 0´606 1´154 2´220 1´398 5´209 0´471 1´080 0´230 0´057 1´066 1´167 1´349 0´421 0´323 2´301 3´550 0´636 0´159 0´647 0´975 1´000 1´080

0´576 1´134 1´122 0´323 0´501 0´248 0´210 0´925 3´948 3´172 1´833 2´671 1´426 0´759 0´000 2´824 1´883 0´202 1´096 0´413 0´206 0´496 0´691 0´457 0´473 0´279 0´778 1´596 0´868 0´000 1´024 0´237 1´000 0´974

0´556 0´949 1´299 0´185 4´159 1´452 2´459 2´084 0´641 0´768 0´749 0´599 0´403 0´163 0´647 0´727 0´527 0´238 0´706 0´177 0´055 0´213 0´729 0´762 0´101 0´239 0´100 1´541 2´330 0´545 0´549 0´865 1´000 0´868

1´336 1´999 1´397 0´950 5´159 1´417 2´484 3´206 0´804 0´421 1´080 0´169 0´481 0´352 1´748 1´442 0´575 0´187 0´509 0´000 0´239 0´153 0´340 0´212 0´603 0´259 0´217 0´463 0´504 0´074 0´450 0´330 1´000 1´078

1´707 0´947 0´998 0´205 1´297 0´823 1´196 1´997 1´061 0´966 0´911 0´944 0´603 0´596 1´539 1´206 2´080 0´732 0´806 1´740 0´321 1´130 0´808 0´762 0´310 0´367 1´141 1´736 1´010 0´903 0´946 1´267 1´000 0´470

0´884 0´565 0´909 1´620 1´610 0´659 1´240 2´807 0´839 1´281 0´759 0´882 0´956 1´180 1´837 1´475 0´783 1´177 1´460 1´582 0´894 0´884 0´952 0´579 1´008 0´723 0´483 0´649 0´602 0´390 0´935 0´994 1´000 0´485

foreign-owned research in this ® eld (an R TA value above 1), the foreign share of the total regional research carried out in the sector has remained low. The same comment applies to another sector conventionally indicated as strong in the regional pattern of production specialization linked to traditional engineering and motor vehicles, namely rubber and plastic products (49), in which foreign-owned ® rms also show a technological advantage but have failed to achieve a high foreign share of local activity. Moreover, it is interesting to note that, in the case of the West Midlands, the high values of the RSA index in both general and chemical engineering and mineral, metallurgy and materials engineering, con® rm a remarkable specialized engineering expertise in publicly funded research, which however is not found for mechanical engineering as such. This supports our earlier hypothesis that the relationship between private (foreign) and public basic research is not as direct as might be expected, and operates more through scienti® c ® elds related and complementary to the speci® c sector of technological

specialization of the M NC (the most immediate knowledge for which is likely to be generated directly from within large ® rms). Turning to the North West, the sectoral concentration of comparative advantages turns out to be the greatest for both indigenous and foreign-owned ® rms. The region shows a clear focus of strength in chemicals, i.e. chemical processes (5), cleaning agents and other compositions (7), synthetic resins and ® bres (9), other organic compounds (11), and in pharmaceuticals and biotechnology (12), the latter being an area of technological advantage in the UK as a whole.13 Also in this case, the RSA index shows a regional scienti® c specialization in chemistry and physics, whilst indicating a relative despecialization in pharmacy, pharmacology and biosciences, which shows once more that the kind of matching between privately- and publicly-funded research occurs more at the level of complementary sectors of science and technology rather than between directly similar ® elds of expertise, with clear implications for regional policy aimed at encouraging linkages


327

Table 4. RSA index relative to the UK, by scienti® c ® eld and region, 1994± 97 Regions Sectors

South East

West Midlands

North West

Medical sciences Pharmacy, pharmacology and biosciences Chemistry Physics Agriculture and forestry Earth, marine and environmental sciences General sciences General engineering Chemical engineering Mineral, metallurgy and materials engineering Civil engineering, architecture and planning Electrical, electronic and computer engineering Mechanical, aero and production engineering Other technologies Mathematics Information technology and systems sciences Catering and hospitality management Business and management studies Geography and social studies Librarianship, communication and media studies Language studies, humanities and creative arts

1´365 0´880 0´751 0´951 0´513 1´032 0´005 0´949 1´072 0´885 0´780 1´047 0´923 0´588 0´724 0´659 0´653 0´804 1´217 0´198 0´969

0´557 0´763 1´180 1´307 0´000 0´336 0´000 2´233 2´840 2´271 0´833 0´772 0´820 0´000 0´959 0´634 0´000 2´396 1´473 0´076 0´510

0´487 0´797 1´153 1´577 0´000 1´168 0´458 0´212 0´607 1´422 1´574 1´094 1´405 3´456 1´111 1´032 0´000 2´054 0´738 1´093 1´160

Total

1´000

1´000

1´000

Source: Higher Education Statistics Agency (H ESA).

and spillover eVects. It should be recalled that the narrower sectoral spread of foreign-owned and locallyowned research in the North West has to be understood also in relation to its historically high degree of productive specialization in textiles and clothing, from which the region’s specialization has shifted over time towards chemicals through the development of synthetic ® bres (it is worth noting that the largest research facilities of ICI are based in the region). This seems to be reaYrmed by the small foreign share of patents in synthetic resins and ® bres (9) despite a high foreign-owned R TA value, while in textile and clothing machinery (25) and other specialized machinery (28) foreign-owned aYliates show both high contributions to regional research and comparative technological advantages (whereas large local ® rms appear to have since undergone a process of despecialization). It is worth mentioning that the correlation coeYcients between the R TA indices of diVerent regions provide some additional insights on the characteristics of the regional and national systems of innovation. The relevant point which emerges is that the specialization patterns of both UK-owned and foreign-owned ® rms in the three regions are not signi® cantly correlated, i.e. the regional hierarchy, also in this case, turns out to be associated with distinctive patterns of regional specialization. As expected, the UK national system as a whole can be understood as a composite of the three regional systems, as con® rmed by the signi® cant correlation found between the RTA of UK-owned ® rms in each of the regions and in the country as a whole

(the highest coeYcient, as expected, is found for the South East). In order to test our hypothesis that the composition of technological specialization of foreign-owned aYliates in each region follows more closely the equivalent pattern of specialization in indigenous companies in lower order than in higher order regions, a simple regression analysis was carried out for the period 1969± 95. In accordance with the two territorial units considered in the analysis (regional and national), two geographical levels were taken into consideration in the cross-section regression analysis. First of all, the hypothesis was tested at the regional level for the three regions (the relevant subscript i5 1, 2 or 3), and then for the UK as a whole. In the former case, a logarithmic transformation of the index was used, since the distribution of the R TA index in each regional case is skewed because of the smaller numbers of patents at the regional level, creating a pattern closer to lognormality than to normality, unlike for the country as a whole.14 The regression was run across all the 56 technological sectors (subscript j ): lnRTAFOR1j 5

a1 b lnRTAUK1j 1 «

1j

lnRTAFOR2j 5

a1 b lnRTAUK2j 1 «

2j

lnRTAFOR3j 5

a1 b lnRTAUK3j 1 «

3j

RTAFORj 5

a1 b RTAUKj 1 «

j

(2)

(3)

First of all, in both the core region, i.e. the South East, and the UK as a whole the specialization of

328


Table 5. Results of the cross-sectoral regressions of foreignowned RTA on indigenous RTA in the U K as a whole, and the principal UK regions UK Intercept RTA UK 2

South East

West Midlands

North West

1´088** 2 0´524* 2 2´125** 2 1´239* (6´241)1 (2 2´589) (2 3´520) (2 2´646) 0´154 0´447* 0´791** 2 0´100 (2 0´691) (1´518) (2´200) (5´061)

LM diagnostic statistics Serial correlation 0´104 Heteroscedasticity 0´014 No. of observations 56

0´388 0´038 56

0´109 0´026 56

0´588 1´231 56

Notes: 1. Figures in brackets are t-statistics. 2. For the three individual regions the regression was run in logarithmic form, the independent variable being lnR TAUKij . *signi® cantly diVerent from zero at 5%; **signi® cantly diVerent from zero at 1%.

foreign-owned ® rms does not depend on the technological advantage of indigenous ® rms (see Table 5). For the South East, this result is analogous to our ® ndings for Lombardia in the Italian case (C A NT WE L L and I A MM A R I NO , 1998), suggesting that regions at the top of the geographical hierarchy, such as these two in the UK and Italy respectively, attract the development eVorts of a broad range of foreign-owned ® rms, generally extending their own lines of domestic specialization by drawing on the general capabilities to be found locally in the leading centre of all-round excellence. This is all the more true in the case of Greater London, whose dynamism in terms of technological and productive activities (particularly services), general infrastructure, ® nancial facilities, etc., need not be rehearsed. Moreover, it has been shown that local ® rms based in the South East region are much more open to external networks, whether through competition or through collaboration, not only in comparison with other UK regions, but also to other European locations (K EEB L E , 1997). Therefore, it might be argued that the spillover eVects due to a high concentration of productive and innovative activities apply over a broader range in this region, which may thus be labelled as a higher order location in the geographical hierarchy, both within the UK and even in a European context. On the other hand, the overall UK model, unlike the Italian case, resembles that of its core region, suggesting a kind of attractiveness towards foreign resources based on general technological competences and infrastructural supports (in terms of degree of openness to foreign investors, business climate, corporate and enterprise culture, language, etc.) oVered by the country as a whole. In contrast, the Italian model was closer to the case of a lower order location, attracting foreign-owned research over a narrower range on a more sectorspeci® c basis.

The West Midlands and the North West show, instead, a rather diVerent picture. Here, the pro® le of technological specialization of foreign-owned ® rms is closely related to the equivalent pattern of specialization of indigenous ® rms in both regions. In the case of the North West, the relationship between the two technological pro® les seems to be particularly strong (the coeYcient is signi® cant at the 1% level), as suggested by our earlier discussion of the detailed composition of the RTA indices. This is consistent with our hypothesis that lower order regions, such as applies also to Piemonte in Italy, attract the innovative activities of foreign MNCs more for their speci® c set of specialized expertise which can be accessed by asset-seeking foreign-owned aYliates. By locating research facilities in such regions, aligned with sectoral-speci® c local strengths, foreign MNCs may be able to upgrade their own technological capabilities in some particular technological ® elds.15

C O NC L US I O NS In a rapidly globalizing economy, and particularly in the face of a process of economic integration such as that occurring in the EU, regions forge an increasing number of linkages with other locations within and across national boundaries through the technological activities of integrated parts of MNC networks that they have managed to attract to local development. The ® rst attempt to analyse the UK case has provided further support to the hypothesis that the pattern of M NC networks for innovation conforms to a hierarchy of regional centres, whose composition diVers between industries. Accordingly, the technological specialization of foreign-owned aYliates in diVerent regional locations depends upon the position of the region in the hierarchy. Moreover, without any intervention to in¯ uence the broad forces of globalization, the strong geographical concentration of innovation might bring about a further marginalization of backward regions, excluding them from the possibility of entering the locational hierarchy. The growing interdependence between local and cross-border innovative activities requires a search for a more macro-systemic approach to the organization of technological resources and capabilities. Therefore, a refocusing of regional and industrial policies, to develop and improve local technological competence to attract and retain development eVorts that contribute a locally speci® c source of innovation to international networks, seems to have become central to the agenda of the Èurope of the regions’ . The analysis of the regional position in the hierarchy and the understanding of the strategic behaviours of M NCs ± increasingly guided by the aim of capturing the synergies oVered by combining diVerent local contexts ± may help promote a more even distribution of the creation and diVusion of know-

Multinational Corporations and the Location of Technological Innovation in the UK Regions ledge, providing the basis for a new phase of regional development in Europe. The results obtained above, although themselves suggestive of further lines of research, particularly through the regionalization of equivalent data for other European locations so as to allow for more extensive comparisons, seem to support our belief that such analysis may contribute substantially to a more accurate description of the crucial link between the `global’ and the `local’ in international networks. Acknowledgem ents ± The authors wish to thank the participants at the 45th North American Regional Science Association International, Santa Fe, and the International Economics Study Group meeting at the LSE in March 1999, as well as two anonymous referees for helpful comments and suggestions. Simona Iammarino gratefully acknowledges the support of the European Commission, under the TMR `Marie Curie’ Programme (category 30, contract no. ERBFMBICT961062).

9.

10.

NO T E S 1. Adopting Dicken’s de® nition `networks are relational structures. . . whose nature depends a good deal upon the particular power structure’ (D I CKE N , 1994, p. 105). 2. The term `globalization’ thus refers to a wider dimension than the one evoked by pre® xes to the word national, such as inter-, multi- or trans- (D ICK EN , 1994; A RCH I BU GI and I A M MA RI N O , 1999). In principle, therefore, we could have a higher interrelatedness among geographically dispersed units even with the same level of internationalization of innovative activities of the M NC. 3. For example, in the U K case, the speci® c object of this paper, Taylor and Wren believe that the reorientation of industrial policy towards the regions particularly during the 1990s, re¯ ected mainly national competitiveness exigencies. Ìt became increasingly recognized that the less well-oV areas could play a substantial role in improving the competitiveness of the U K economy as a whole’ (T AY LOR and W REN , 1997, p. 838). 4. The relevance of the regional dimension of innovative systems in the global economy has been discussed more extensively in C A N TWEL L and I A MM A RI NO, 2000. 5. Much more work has been done on the foreign ownership in U K manufacturing at the territorial level (see, among others, D ICK EN and L LOYD , 1980; H I L L and M U N DAY, 1992; S T O N E and P EC K , 1996). 6. As stated above, for the UK study we decided to refer to the NUTS 1 level, instead of the NU TS 2, in order to ensure the comparability among European regions as much as possible. According to Eurostat Basic Principles, in fact, NU TS 1 in the U K is the level corresponding to `standard regions’ (11), therefore comparable to the 20 N UTS 2 Italian regions (see also E U RO STAT , 1995). 7. In the UK, the total number of patents registered in the Reading database over the period 1969± 95 is more than ® ve times that registered for Italy. 8. A number of studies pointed out the strong concentration of (indigenous) innovation in the South East, providing evidence of the fact that signi® cant innovations

11.

12. 13.

14.

15.

329

are more likely to be introduced into the South East regional context than elsewhere in the UK (see, among others, H A RRI S , 1988; T H WA I TE S and W YNARCZ Y K , 1996; C O S H and H U GH ES , 1996; K EEB L E and B RYSO N , 1996; K EEB L E , 1997). Thwaites and Wynarczyk, in particular, show that the innovative performance of the South East is related to elements of macro and corporate culture and of the modus operandi of ® rms located in the region. Just for the sake of clarity, the two output indicators are diVerent not only insofar as the EPO re¯ ects `patent requests’ and the US PTO `patents granted’, but also because in the ® rst case the overall number of requests is reported, whilst in the second the Reading database includes only corporate patents, i.e. patents granted to the largest industrial ® rms. More speci® cally, taking the period as a whole, G DP per capita in most regions show a tendency to fall relative to that of Greater London, moving towards an asymptotic level of 56% of Greater London GD P. Looking at the subperiods, however, their analysis suggests that the UK regions exhibit a stronger tendency to converge in periods of slow national growth than they do in boom conditions (C H AT T ERJ I and D EW H U RST , 1996). Some of the technological sectors were dropped from the table on the grounds of the relatively small number of patents for corporate research in the country as a whole, the outcome being that only 32 sectors are reported in Tables 2 and 3 and thus discussed in the text. However, the total refers to the total number of patents for all 56 technological sectors. The key to the sectoral codes is given in the Appendix. For the UK as a whole, the index obviously considers S i , where i5 1, . . . , 11 (the 11 N UTS 1 regions). Historically, pharmaceuticals was a traditional British strength ± deriving partly from food industries, just as some other chemical areas were linked to textiles or to oil ± but it had fallen away for much of the middle part of the century. It recovered its tradition in the 1970s and 1980s and, as argued by C AN TWE L L , 1987, this was due in part to the favourable interaction between research in US-owned and indigenous British ® rms in this industry, which seems from our current data to have had a regional focus in the North West. The standard assumption of this methodology is that the regression is linear, which is valid if the cross-sector index approximately conforms to a normal distribution. However, as discussed more extensively in C AN TWE L L , 1991, if the sample of patents numbers is not suYciently large, an arti® cially high degree of dispersion in the index is very likely to be found, as in the regional cases considered here. This is what tends to lead to a skewed distribution, rather than any property of the RTA index itself. Thus, for the three regions, the sectoral distribution represented by the RTA conforms more closely to a lognormal than a normal distribution, so taking a logarithmic rather than a linear functional form in such cases improves the values of the t-statistic on b. Furthermore, by analogy with the exercise carried out for Italy, the period was sub-divided into 1969± 82 and 1983± 95 in order to test the validity of our hypothesis at diVerent geographical levels over time, according to

330


the fact that the globalization of innovation, as de® ned above, is likely to have taken oV especially from the 1980s onwards. We thus adopted a Granger notion of sequential causality, and we ran the regression for 47 technological sectors (the reason for having a smaller number of sectors in the lagged cross-section model is that we dropped all the technological sectors with an overall number of patents less than 600 in the world total in both 1969± 82 and 1983± 95, in order to avoid the inclusion of sectors with a relatively low propensity to patent at the world level). As expected, the results were not signi® cant in all cases except the North West, for which the overlap in technological specialization of U K-owned and foreign-owned ® rms holds also in the regression over time (at the 1% level of signi® cance).

A P P E ND I X Table A1. Sectoral groups 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17

Food and tobacco products Distillation processes Inorganic chemicals Agricultural chemicals Chemical processes Photographic chemistry Cleaning agents and other compositions Disinfecting and preserving Synthetic resins and ® bres Bleaching and dyeing Other organic compounds Pharmaceuticals and biotechnology Metallurgical processes Miscellaneous metal products Food, drink and tobacco equipment Chemical and allied equipment Metal working equipment

18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56

Paper making apparatus Building material processing equipment Assembly and material handling equipment Agricultural equipment Other construction and excavating equipment Mining equipment Electrical lamp manufacturing Textile and clothing machinery Printing and publishing machinery Woodworking tools and machinery Other specialized machinery Other general industrial equipment Mechanical calculators and typewriters Power plants Nuclear reactors Telecommunications Other electrical communication systems Special radio systems Image and sound equipment Illumination devices Electrical devices and systems Other general electrical equipment Semiconductors OYce equipment and data processing systems Internal combustion engines Motor vehicles Aircraft Ships and marine propulsion Railways and railway equipment Other transport equipment Textiles, clothing and leather Rubber and plastic products Non-metallic mineral products Coal and petroleum products Photographic equipment Other instruments and controls Wood products Explosive compositions and charges Other manufacturing and non-industrial

RE F E RE N C E S A M IN A. (1992) Big ® rms versus the regions in the Single European Market, in D U NF O RD M. and K A FKALAS G. (Eds) Cities and Regions in the New Europe: The Global± Local Interplay and Spatial Development Strategies. Belhaven Press, London. A RCH I BUG I D. (1992) Patenting as an indicator of technological innovation: a review, Sci. & Public Policy 19, 357± 68. A RCH I BUG I D. and I A M MA RI N O S. (1999) The policy implication of the globalisation of technology, Res. Policy 28(2± 3), 317± 36. A U D R ET SCH D. B. and F EL D M AN M. (1996) R&D spillovers and the geography of innovation and production, Am. Econ. Rev. 86(3), 630± 40. B R ESCH I S. (1995) Spatial patterns of innovation: evidence from patent data, Department of Economics, University of Urbino (mimeo). B R ESCH I S. (1997) The geography of innovation: a cross-sector analysis, CESPRI Working Paper No. 95. C A N IEÈ L S M. C. J. (1997) The geographic distribution of patents and value added across European regions, paper presented at the 37th Regional Science Association European Congress, August, Rome. C A N TWEL L J. A. (1987) The reorganization of European industries after integration: selected evidence on the role of multinational enterprise activities, J. Common Mkt. Studies 26(2), 127± 51. C A N TWEL L J. A. (1989) Technological Innovation and Multinational Corporations. Blackwell, Oxford. C A N TWEL L J. A. (1991) Historical trends in international patterns of technological innovation, in F O RE MA N - P EC K J. (Ed) New Perspectives on the Late Victorian Economy. Cambridge University Press, Cambridge. C A N TWEL L J. A. (1992a) The eVects of integration on the structure of multinational corporation activity in the EC, in K L E IN M. W. and W EL F EN S P. J. J. (Eds) Multinationals in the New Europe and Global Trade. Springer-Verlag, Berlin. C A N TWEL L J. A. (1992b) The internationalisation of technological activity and its implications for competitiveness, in G RA N STRA N D O., H AÊ KAN SO N L. and S J OÈ LAN D ER S. (Eds) Technology Management and International Business: Internationalization of R&D and Technology. Wiley, Chichester.


331

C A N TWEL L J. A. (Ed) (1992c) Multinational Investment in Modern Europe: Strategic Interaction in the Integrated Community. Edward Elgar, Aldershot. C A N TWEL L J. A. (1993) Corporate technological specialisation in international industries, in C A SSO N M. C. and C RE EDY J.(Eds) Industrial Concentration and Economic Inequality. Edward Elgar, Aldershot. C A N TWEL L J. A. (Ed) (1994) Transnational Corporations and Innovatory Activities. Routledge, London. C A N TWEL L J. A. (1995) The globalisation of technology: what remains of the product cycle model?, Camb. J. Econ. 19(1), 155± 74. C A N TWEL L J. A. and H A RD I N G R. (1998) The internationalisation of German companies’ R&D, Nat. Inst. Econ. Rev. 163, 99± 115. C A N TWEL L J. A. and H O D SO N C. (1991) Global R&D and U K competitiveness, in C A SSO N M. C. (Ed) Global Research Strategy and International Competitiveness. Blackwell, Oxford/Cambridge, MA. C A N TWEL L J. A. and J AN N E O. E. M. (1999) Technological globalisation and innovative centres: the role of corporate technological leadership and locational hierarchy, Research Policy 28(2± 3), 119± 44. C A N TWEL L J. A. and I AM M AR IN O S. (1998) MNCs, technological innovation and regional systems in the EU: some evidence in the Italian case, Int. J. Econ. Bus. 5(3), 383± 408. C A N TWEL L J. A. and I A M MA RI N O S. (2000) Regional systems of innovation in Europe and the globalisation of technology, in B A RTZ OKAS A. (Ed) Technology Policy and Regional Integration. Routledge, London. C H AT TE RJ I M. and D EW H U R ST J. H. L. (1996) Convergence clubs and relative economic performance in Great Britain: 1977± 1991, Reg. Studies 27(8), 727± 43. C O SH A. and H U GH ES A. (1996) The Changing State of British Enterprise: Growth, Innovation and Competitive Advantage in Small and Medium Sized Firms 1986± 95. ESRC Centre for Business Research, University of Cambridge. D I CK EN P. and L LOYD P. E. (1980) Patterns and processes of change in the spatial distribution of foreign-controlled manufacturing employment in the United Kingdom, 1963 to 1975, Environ. Plann. A 12(1), 404± 26. D I CK EN P. (1990) Seducing foreign investors ± the competitive bidding strategies of local and regional agencies in the United Kingdom, in H E BB ERT M. and H A NS EN J. C. (Eds) Unfamiliar Territory: The Reshaping of European Geography. Avebury, Aldershot. D I CK EN P. (1994) The Roepke lecture in economic geography: global± local tensions: ® rms and states in the global space economy, Econ. Geogr. 70(2), 101± 28. D U N F OR D M. (1996) Regional disparities in the European Community: evidence from the REGIO databank, Reg. Studies 30(1), 31± 40. D U N N IN G J. H. (1995) Reappraising the eclectic paradigm in an age of alliance capitalism, J. Int. Bus. Studies 26(3), 461± 91. D U N N IN G J. H. (1996) Globalisation, technological change and the spatial organisation of economic activity, Discussion Papers in International and Business Studies No. 211, University of Reading E U RO STAT (1995) NU TS: Nomenclature of Territorial Units For Statistics. Eurostat, Luxembourg. F AGE RBE RG J. and V ERS PAGEN B. (1996) Heading for divergence? Regional growth in Europe reconsidered, J. Common Mkt Studies 34(3), 431± 48. F AGE RBE RG J., V E RSPAGEN B. and C A NI EÈ L S M. C. J. (1997) Technology, growth and unemployment across European regions, Reg. Studies 31(5), 457± 66. F E L D MA N M. P. (1994) The Geography of Innovation. Kluwer Academic, Boston. F O R S G. (1998) Locating R&D abroad: the role of adaptation and knowledge-seeking, in B RAU N ERH J EL M P. and E KH O L M K.(Eds) The Geography of Multinational Firms. Kluwer Academic, Boston/Dordrecht/London. G RI L ICH ES Z. (1990) Patent statistics as economic indicators: a survey, J. Econ. Lit. 28, 1,661± 707. H AG ED OOR N J. and N ARU LA R. (1996) Choosing organisational modes of strategic technology partnering: international and sectoral diVerences, J. Int. Bus. Studies 27(2), 265± 84. H AR RI S R. I. D. (1988) Technological change and regional development in the U K: evidence from the SPRU database on innovations, Reg. Studies 22, 361± 74. H ILL S. and M U N DAY M. (1992) The UK regional distribution of foreign direct investment: analysis and determinants, Reg. Studies 26, 535± 44. H IR ST P. and T H O M PSO N G. (1996) Globalization in Question. Polity Press, Cambridge. J A FF E A.B., T RA J T EN BERG M. and H EN D ER SO N R. (1993) Geographic localization of knowledge spillovers as evidenced by patent citations, Quart. J. Econ. 63(3), 577± 98. K EEB L E D. (1997) Small ® rms, innovation and regional development in Britain in the 1990s, Reg. Studies 31(2), 281± 93. K EEB L E D. and B RYSO N J. (1996) Small ® rms creation and growth, regional development and the north-south divide in Britain, Environ. Plann. A 28, 909± 34. K U EM M ERL E W. (1999) The drivers of foreign direct investment into research and development: an empirical investigation, J. Int. Bus. Studies 30(1), 1± 24. M ALE CKI E. J. (1980) Dimensions of R&D location in the United States, Res. Policy 9, 2± 22. M ALM BE RG A., S OÈ LVE L L O. and Z A ND ER I. (1996) Spatial clustering, local accumulation of knowledge and ® rm competitiveness, Geogra® ska Annaler 78B(2), 85± 97. P AT EL P. and P AV IT T K. L. R. (1991) Europe’s technological performance, in F REEM A N C., S H A R P M. and W ALK ER W. (Eds) Technology and the Future of Europe. Frances Pinter, London. P AV IT T K. L. R. (1988) Uses and abuses of patent statistics, in V A N R A AN A. (Ed) Handbook of Quantitative Studies of Science Policy. North Holland, Amsterdam.

332


Q UA H D. T. (1996) Regional convergence clusters across Europe, LSE Centre for Economic Policy Discussion Paper 274, London School of Economics. S AC H WA L D F. (1998) Co-operative agreements and the theory of the ® rm: focusing on barriers to change, J. Econ. Behav. & Organiz. 35(2), 203± 25. S O ETE L. L. G. (1987) The impact of technological innovation on international trade patterns: the evidence reconsidered, Research Policy 16, 101± 30. S TO N E I. and P EC K F. (1996) The foreign-owned manufacturing sector in UK peripheral regions, 1979± 1993: restructuring and comparative performance, Reg. Studies 30(1), 55± 68. T AYLO R J. and W RE N C. (1997) UK regional policy: an evaluation, Reg. Studies 31(9), 835± 48. T H WA I TE S A. and W Y NA RCZ YK P. (1996) The economic performance of innovative small ® rms in the South East region and elsewhere in the U K, Reg. Studies 30(2), 135± 49. Z A ND ER I. (1999) How do you mean global? A taxonomy of innovation networks in the multinational corporation, Research Policy 28(2± 3), 195± 213.