Increasing emphasis has recently been placed on the regional dimension in economic analysis. The re-discovered importance of the regional dimension lies in the role played by spatial factors in economic performance. Spatial agglomeration of economic activity seems to generate a competitive advantage stemming from localised higher value added embedded in the local socio-economic environment. In spite of the development of information and communications technologies (ICT), MNCs seem be highly sensitive to untraded externalities and localised knowledge spillovers in their location strategy. Therefore, in an age in which space and time constraints have been reduced (or even eliminated) by the fast pace of technological change, proximity still matters.
If this is true in production, empirical evidence has shown that it is all the more true in the development of technology. Local agglomeration economies are top targets in the corporate location of R&D laboratories. This seems to be particularly the case of cutting-edge industries (such as ICT) highly relying in localised knowledge spillovers. Although a large number of studies have focused on the regional dimension in the US economy, only recently has economic analysis started exploring regional locational aspects across Europe. The initial findings show clearly the existence of higher and lower order regions as a result of vicious and virtuous cycles reinforced over time.
In the global-local nexus debate, promoted by the re-discovery of the regional dimension in economic theory, it is interesting to analyse the role played by regions from European small countries in the development of technology within the European ICT industry. Using regionalised patent data granted in the US to large electrical firms for their European-located research, this paper aims to investigate the interaction between local and foreign corporate technological development in small country regions in order to evaluate the role of these locations in the geographical division of labour in European-owned ICT multinationals.
Due to the role played by spatial factors in economic performance, the importance of the regional dimension has been re-discovered in the knowledge-driven age. Although the fast pace of technological change has reduced and to some extent eliminated space and time barriers, proximity and face-to-face interaction still matter in corporate technological development. Therefore, increasing emphasis has recently been placed on the regional dimension in economic analysis.
This seems to apply particularly to science-based industries (such as information and communication technology – ICT) highly reliant on knowledge spillovers, which can be better captured locally. Therefore, in spite of the development of ICT, in the compilation of corporate location strategies multinationals are highly sensitive to untraded externalities, which may go well beyond intra-industry knowledge spillovers as in the Marshall-Arrow-Romer (MAR) model. The analysis of corporate spatial strategies seems to suggest that inter-industry and university knowledge spillovers are the main factors attracting foreign direct investment (FDI).
Although in the innovation literature there is a well-established tradition of spatial analysis in US regions (Jaffe et al., 1993, Audretsch and Feldman, 1996, Audretsch and Stephan, 1996, Feldman and Audretsch, 1999), the research interest in the European context is rather recent. Moreover, the few studies focusing on Europe have so far neglected regions from small countries. In attempting to bridge the gap, the aim of this paper is twofold. First of all, the paper argues the significance of a region- over a nation-State-based analysis particularly in a heterogeneous socio-economic environment as Europe. Second, the focus is on the interaction between local and foreign technological development in European small country regions in order to evaluate the role of these locations in the geographical division of labour of European-owned ICT multinationals.
By using regionalised patent data granted in the US to large electrical firms for their European-located research, paper is organised in 5 main sections. The following section discusses the reasons for a region-based analysis. Section 3 describes the regionalised data and the sample. In section 4, the argument for a region- over a nation-State-based analysis is examined statistically, and corporate location strategies investigated by analysing the interplay between European foreign and indigenous electrical companies. A few conclusions are drawn in section 5.
The last two decades have withnessed the rise of ICT as a core pervasive technology. This phenomenon has impacted on the development of corporate technological profile, intra- and inter-firm division of labour and corporate location strategies. The pervasive character of ICT calls for corporate expertise in the development and management of this key technology even when it lies outside core corporate competencies. Similarly, the complex character of technology together with the increasing technological interrelatedness has promoted a more heterarchical division of labour within the multinational coporation (MNC) as well as an exploration in the adoption of collaborative agreements as a new organisational form of economic activity. As far as the spatial dimension is concerned, the ICT revolution has impacted on corporate location strategy by reducing and to some extent eliminating space and time constrains. Although geography might have expected to be neglected by the MNC as a result of these new techno-economic conditions, spatial factors continue to play a great role in corporate economic performance. Spatial agglomeration of economic activity seems to generate competitive advantage steaming from localised higher value added embedded in the local socio-economic environment. Therefore, in spite of the development of ICT, MNCs seem to be highly sensitive to untraded externalities and localised knowledge spillovers in their location strategies.
Following the stream of literature generated by "new economic geography" (e.g. Krugman, 1991), increasing returns are essentially a regional and local phenomenon arising from regional economic agglomeration and specialisation. This applies particularly to cutting-edge industries (such as ICT) highly reliant on localised knowledge spillovers. In these industries, university-industry links, user-producer relationships and proximity to corporate competitors are key factors in generating corporate success. Conversely, nation-State-based analyses seem to neglect intra-border dynamics at both a micro and a macro level. On the micro level, indigenous processes of knowledge accumulation are determining corporate location strategies aimed at building sustainable competitive advantage by through interacting with complementary external expertise. On the macro level, within national borders over time indigenous economic performance appears to be reinforced by vicious and virtuous cycles, further fed by the interplay of foreign and indigenous activity (see Cantwell, 1987, 1989). The immediate results of these phenomena is the establishment of geographical hierarchies of local centres distinguished between higher and lower order locations.
In evaluating the re-discovery of geographical proximity, it should be borne in mind that the phenomenon of spatial agglomeration of economic activity is parallel and complementary to corporate technological globalisation. The term globalisation is here adopted to indicate the increasing geographical dispersion of intra-firm networks across a large number of locations and the consequent interdependency between the units which constitute the MNC. The establishment of geographically dispersed networks allows the transfer of technology, skills and assets across national borders between the parent company and the subsidiaries as well as between the subsidiaries and the parent company. In this context, the subsidiary plays a great role in interacting with the indigenous techno-socio-economic environment in order to tap into local expertise of technological development. The competencies of the indigenous location may be transferred and absorbed in the technological path of the whole MNC through the geographically dispersed intra-firm network. In order for this corporate strategy to be successful, a key point is to identify local centres of knowledge accumulation complementary to the corporate technological path and/or indigenous dynamic environments generating knowledge spillovers. In the context of the global-local nexus debate, the regional dimension has been re-discovered in economic theory as a significant spatial unit of analysis. Corporate spatial organisation is driven by the different types of indigenous agglomeration forces whether shaped as general external economies or sector-specific economies (see e.g. McCann, 1995).
In this theoretical and factual context, the region-based dimension in economic analysis is particularly meaningful in the European context when considering the furthering of the integration process. The pervasive character of ICT is provoking a transformation in urban and regional systems by generating a new network model involving core regions and neglecting the peripheral (Hepworth, 1986, Malecki, 1994, Antonelli, 1995, 1996, Castells, 1985). As technological change affects not only the long-term growth of economies, but also the spatial distribution of economic activity, a concentration phenomenon has emerged (Feller, 1975). In order to account for this new techno-socio-economic conditions, economic and social cohesion across the Union is increasingly pursued through a more balanced inter-European regional development. The recognised domestic socio-economic inequalities between the member states have pushed the European Union (EU) to implement its own regional policy. Similarly, MNCs are extremely sensitive to indigenous specialisation and economic performance when looking at the map of their investments across Europe. However, the interest in a regional analysis in the European context is rather recent by comparison with the established tradition of this kind of studies in the US. Moreover, current investigations concerning European regions have neglected small country realities.
The regional data, upon which the empirical analysis of this study is based upon, concern US patents granted to the largest European electrical firms between 1969 and 1995. As this paper does not aim to discuss the uses and abuses of patent statistics as a measures of innovative activity, reference is made to the large literature on the topic (see e.g. Pavitt, 1988).
The data are drawn from the University of Reading database. Using a classification scheme derived from the US Patent and Trademark Office (USPTO) patent class system, each patent is classified into one of the 399 original patent classes. In the Reading database, those patent classes have been allocated to one of 56 technological groups according to related technological fields/sectors; 6 of the 56 technological sectors comprise the main fields of ICT. Our study is restricted to all European firms classified in the broad electrical corporate industrial group in the database held at the University of Reading. The broad electrical industrial group includes the ‘electrical equipment’ (communications) and ‘office equipment’ (computing) industries.[]
The regional dimension is captured through patent document information concerning the geographical location of the research activity (i.e. the address of the inventor). The sub-national spatial entities identified correspond to the units classified by the European Nomenclature of Territorial Units for Statistics (NUTS) established by Eurostat. By providing a single uniform breakdown of territorial units, this classification is based on the institutional division currently in force in the member states.[] Therefore, in the Reading database, a NUTS code is attributed to each patent on the grounds of the address of the inventor (research establishment) ultimately responsible (see Cantwell and Iammarino, 1998, 1999 for a further discussion).
The spatial analysis of the corporate research activity in the European ICT industry is carried out at both national and sub-national level. At the national level, the European host locations considered are small European countries (i.e. Belgium, the Netherlands, Sweden and Switzerland). While the first three countries are European Union member states, the inclusion of Switzerland is due to the historical tradition of the country as a major European centre of technological development. For each of these countries, at a sub-national level the geographical distribution of European electrical corporate patenting activity in the ICT technological sectors is investigated across territorial units as classified by the NUTS nomenclature. In order to ensure as much comparability as possible, the NUTS 1 level is used to identify Belgian and Dutch regions, while, as far as Swedish regions are concerned, the NUTS 2 level is adopted. As pointed out by Eurostat (1995), despite the aim of ensuring that comparable regions appear at the same NUTS level, each level still contains great differences among the territorial units identified in terms of area, population, economic weight or administrative power. Therefore, the 3 Belgian regions, the 4 Dutch provinces and the 8 Swedish counties seem to guarantee grounds of comparability in the analysis of innovative activity. As far as Switzerland is concerned, no NUTS subdivision is available for the Swiss territory as it is a non-EU member. In the Reading database, Switzerland is geographically subdivided in 12 regions according to proximity to big cities.
To account for the most significant locations in terms of development of innovative activity, a selection criterion is adopted to carry out the econometric exercise. European small country regions are included in the econometric exercise if hosting more than a sectoral total of 250 patents granted to European-owned electrical corporations. This implies that the econometric analysis focuses on 18 regions.
As shown in Figure 1, the European small country regions considered as a whole account for 27.36% of the total research and development (R&D) in ICT carried out by European-owned electrical corporations. As discussed below, this percentage is almost fully located in the 18 selected regions.
THE EMPIRICAL EVIDENCE
The neglected regional dimension
The significance of a region-based analysis is confirmed by the analysis of Tables 1 and 2, where the distribution of corporate research facilities in ICT sectors is investigated at national and regional level respectively.
Host country | |
Netherlands | 19.83 |
Sweden | 4.05 |
Switzerland | 2.33 |
Belgium | 1.23 |
Total | 27.36 |
Total all other European countries | 72.64 |
European total | 100 |
Region | |
Zuid-Nederland (NL) | 16.55 |
Stockholm (SE) | 2.40 |
Oost-Nederland (NL) | 1.77 |
West-Nederland (NL) | 1.43 |
Basel (SZ) | 0.96 |
Vlaams Gewest (BL) | 0.86 |
Zurich (SZ) | 0.67 |
Östra Mellansverige (SE) | 0.52 |
Västsverige (SE) | 0.38 |
Luzern (SZ) | 0.37 |
Sydsverige (SE) | 0.37 |
Bruxelles (BL) | 0.24 |
St. Gallen (SZ) | 0.24 |
Norra Mellansverige (SE) | 0.24 |
Smäland med öarna (SE) | 0.10 |
Noord-Nederland (NL) | 0.06 |
Region Wallonne (BL) | 0.05 |
Fribourg (SZ) | 0.02 |
Region | |
#### | |
Zuid-Nederland (NL) | |
Stockholm (SE) | |
Oost-Nederland (NL) | |
Source: As for Table 1. | |
Table 1 ranks the European small countries on the grounds of the distribution of research in all ICT sectors carried out by European-owned electrical corporations. The Netherlands appears to be a higher order location in the European corporate development of ICT by hosting nearly all the research activity located in the countries under analysis. The presence of Philips’ headquarters explains this large percentage (19.83%) of research and development in a cutting-edge technology such as ICT. Sweden is revealed as the second preferred choice in the R&D location decision of European-owned corporations, followed by Switzerland and Belgium respectively. In the Swedish case, the host success in the development of telecommunications and mobile communications especially is likely to be a major factor attracting European research in the country. As far as Switzerland and Belgium are concerned, the two countries host relatively low percentages of ICT research from European-electrical corporations. Although this analysis allows us to build an overall picture of the geography of ICT in these countries, intra-border inequalities are neglected as shown when considering Table 2.
Table 2 reports the distribution of ICT research carried out by European-owned electrical companies across regional units as identified in section 3. At a first glance at the figures, if the Dutch regions are confirmed as higher order locations on average, the ranking of Swedish, Swiss and Belgian regions follows a less clear-cut pattern by comparison with the national picture drawn above. The relatively high percentage of ICT research activity located in Zuid-Nederland seems to be fully due to the presence of Philips’ headquarters in the region (in Eindhoven), which is the most industrialised of the Netherlands. The strong university commitment in the forefront of applying new technologies together with high skilled labour may explain the high ranking of Oost-Nederland and West-Nederland. Conversely, the relatively low attraction of Noord-Nederland for European-owned R&D in ICT (neglected in the national level analysis) is fully revealed when investigating the intra-border dimension of the phenomenon in question. Similarly, the evaluation of Sweden as host location in the development of ICT changes when considering the regional breakdown of the national territory. Although the position of Sweden as the second best choice in the R&D location decisions of European-owned electrical companies in the countries in question is confirmed by the high ranking of Stockholm in Table 2, the other Swedish regions attract research in ICT to differing extents. As far as the Swiss regions are concerned, the disaggregation of the data at the level of regional units appears to provide a clearer picture of the intra-border dynamics characterising the local development of ICT. In the case of Belgium, the significance of a regional analysis is confirmed when considering the percentage of ICT research hosted in Vlaams Gewest by comparison with other Dutch, Swedish and Swiss regions that are relatively less attractive.
The interplay between European foreign and host corporate research activity in ICT
The interplay between foreign and host R&D activity in ICT in European small country regions is investigated with a probit analysis across the 18 selected regional locations and the 6 ICT technological sectors for a total number of observations of 108 (= 18*6). The likelihood of a foreign presence in a region in a particular sector is investigated as a function of the specialisation of the indigenous firms in the region and sector in question. The dependent variable for foreign presence is proxied through a dummy variable (FORPRE) set equal to 1 if at least one patent is granted to foreign European-owned electrical companies in the sector and region in question, and to 0 otherwise. The revealed technological advantage (RTA) index is adopted as a proxy for the indigenous technological specialisation. For each regional group of European indigenous electrical firms (i = 1, with i = 2 denoting foreign-owned), the RTA is defined as the share of US patents in an ICT sector (j) in a region (r) relative to the total number of US patents granted (to both European indigenous and foreign electrical firms) in the same sector across all regions divided by the share of patents granted to the same regional group of European indigenous electrical firms (i) in the same region (r) in all ICT sectors relative to all patents granted (to both European indigenous and foreign electrical firms) in all ICT sectors (j) across all regions. Thus, the index can be mathematically formalised as:
RTAirj = (Pirj/å irPijr)/ (å jPijr/å ijrPijr) (1)
where Pirj is the total number of patents granted to European indigenous electrical firm (i = 1) in region (r) in an ICT sector (j). It is worth emphasising that (i) is defined in terms of groups of European indigenous electrical firms only in the numerators and in terms of both European indigenous and foreign electrical firms in the denominators. As the index is a comparative measure, high (low) values of RTAirj indicate corporate technological specialisation (despecialisation) in a specific ICT sector in the region in question. Therefore, within the European electrical industry, the index enables one to evaluate for each group of European indigenous electrical firms the technological specialisation in a specific European small country region and ICT sectors relatively to all other ICT sectors in the same region.
The results of the probit analysis do not show any statistically positive significant relationship between the presence of European foreign electrical firms in a particular ICT sector in small country regions and the corporate indigenous technological specialisation in the same ICT sector, as illustrated in Table 3.
Dependent Variables |
FORPRE
|
RTAIF |
-0.11442
|
(-0.91103)
|
|
|
|
No. obs. |
108
|
|
|
Log-likelihood |
-74.0569
|
This implies that on average European electrical corporate research in ICT sectors does not seem to be attracted by indigenous corporate technological specialisation in the sectors in question. In order to verify whether it is possible to identify particular patterns beyond the average trends, these results have been investigated further.
For this purpose, Figure 2 reports the distribution under analysis in the probit model discussed above by categorising each region-sector combination as belonging to one of the four boxes describing corporate location strategies.
|
|
|
|
|
Indigenous Specialisation |
|||||||||
|
|
|
|
|
|
|||||||||
|
|
|
|
|
|
|||||||||
|
|
|
RTA > 1 |
|
|
RTA < 1 |
||||||||
|
|
|
|
|
|
|||||||||
|
|
Telecommunications |
|
Computing |
|
Other Electrical Communication Systems |
||||||||
West-Nederland |
St. Gallen |
at least 5 ICT sectors |
Office Equipment and Data Processing Systems |
|||||||||||
Stockholm |
Basel |
Bruxelles/Brussel |
Sydsverige |
|||||||||||
FP |
Sydsverige |
Luzern |
Vlaams Gewest |
Norra Mellansverige |
||||||||||
Fribourg |
Zurich |
|||||||||||||
Basel |
Computer sectors mainly |
|||||||||||||
Stockholm |
Östra Mellansverige |
|||||||||||||
Communications sectors only |
Region Wallon |
Computer sectors only |
||||||||||||
Zuid-Nederland |
Luzern |
West-Nederland |
||||||||||||
Noord-Nederland |
||||||||||||||
St. Gallen |
||||||||||||||
Foreign Presence |
ASSET SEEKING ATTRACTION |
COMPENSATING INDIGENOUS WEAKNESS |
||||||||||||
Image and Sound Equipment |
||||||||||||||
|
|
|
Computing |
all ICT sectors |
at least 2 communications sectors |
Semiconductors |
||||||||
|
Communications |
|
Fribourg |
Noord-Nederland |
Oost-Nederland |
Sydsverige |
||||||||
|
Östra Mellansverige |
|
|
West-Nederland |
Zuid-Nederland |
West-Nederland |
Västsverige |
|||||||
|
Sydsverige |
Oost-Nederland |
|
|
Region Wallon |
|
||||||||
NFP |
Västsverige |
Zuid-Nederland |
|
Semiconductors |
Fribourg |
Luzern |
||||||||
|
Norra Mellansverige |
|
|
Stockholm |
Östra Mellansverige |
St. Gallen |
Office Equipment and | |||||||
|
Luzern |
|
|
Vlaams Gewest |
data proceesing systems | |||||||||
Bruxelles/Brussel |
||||||||||||||
Special Radio Systems |
||||||||||||||
. |
DETERRED BY INDIGENOUS STRENGTH |
|
|
|
LOCATION DISADVANTAGE |
|
|
|
|
|
|
|
The 108 observations are allocated to one of the four boxes in Figure 2 on the basis of whether they are characterised by indigenous specialisation (RTA > 1) – or despecialisation (RTA < 1) - in ICT sectors, and foreign presence (FP) – or no foreign presence (NFP). In each box (corresponding to one of the four possible combinations of the two variables considered), this exercise allows us to identify groups of European small country regions hosting corporate research activity in specific ICT sector(s). In turn, each box describes four sectoral location strategies in the development of ICT, as described in the labels at the bottom of each of them. A clear pattern appears to emerge in the top and bottom left hand boxes as discussed below.
The top left hand box seems to describe an asset seeking strategy by categorising all European small country regions hosting European foreign electrical research in ICT sectors and showing an indigenous specialisation in the technological sectors in question. Two regional groups are clearly identified as clustered around specific ICT sectors. In both cases, the presence of foreign research activity appears to be driven by indigenous assets, which may go well beyond the sectoral indigenous specialisation. The regional locations clustered in these groups can be considered as higher order centres characterised by excellent infrastructures and dynamic indigenous socio-economic environments. As far as the group of regions clustered around the "telecommunications" sector is concerned, the well-developed communications systems together with the local economic dynamics of Stockholm and Sydsverige may explain the asset seeking strategy driving the R&D location decisions of foreign European-owned electrical firms. The inclusion of West-Nederland in this group seems to confirm this pattern when considering the excellent performance of the very internationally oriented Randstad economy together with Philips’ technological expertise. Similarly, the other group of regions (concerning Swiss regions only) may be targeted by foreign European electrical companies as a result of the presence in the regions of ABB’s (previously Brown Boveri) research in computing. However, the indigenous development of the service sector and consequent economic performance should be also taken into account.
The combination of indigenous specialisation and absence of foreign research in ICT considered in the bottom left hand box suggests that European-owned electrical research is deterred by indigenous strength in the sectors and regions listed in the box. The three groups of regions identified are clustered around ICT sectors in which indigenous firms are European leaders. This argument is clearly described by the Swedish regions clustered around the communications sector. Östra Mellansverige, Sydsverige, Västsverige and Norra Mellansverige host large percentages of Ericsson’s and Electrolux’s R&D activity. The agglomeration of indigenous corporate research facilities is encouraged in these regions by local dynamics reinforced by university-industry links and a well-developed educational system over time. The same argument seems to apply to Fribourg, where the indigenous strength in certain computing-related sectors (mainly semiconductors) discourages the location of European-owned foreign research. Similarly, the strong indigenous commitment in R&D in both computing and communications sectors acts as a deterrent for the location of European foreign laboratories in Oost-Nederland and Zuid-Nederland - identifying a regional group lying at the intersection of the other two. The latter is the most typical example of the corporate location strategy described in this box as Eindhoven is Philips’ headquarters. Two points should be made when considering this group. First of all, the absence of European-owned R&D in both communications and computing confirms Philips as a "technological generalist" active in a wide spectrum of technological sectors (Santangelo, 1998). Second, it is worth noting that the absence of foreign R&D facilities refers to European-owned companies, which are technologically too weak to compete with indigenous specialisation expertise (Ibid.). Conversely the two Dutch regions host numerous US-owned R&D laboratories (European Communities, 1993). This emphasises the European technological leadership of Philips able to compete with US giants.
As far as the top and bottom right hand boxes are concerned, the combinations of the two variables considered in each of them seem to draw two different corporate location strategies: compensating indigenous weakness and location disadvantage respectively. At a first glance at the regional groups clustered around specific ICT sectors, the patterns illustrated in the other two boxes in the left hand side of Figure 2 are confirmed. The technological weakness of Swedish firms in computing emerges clearly as shown by the occasional presence of foreign R&D (despite an indigenous disadvantage in the sectors in question) in Sydsverige and Norra Mellansverige, and Östra Mellansverige. Similarly, a simple location disadvantage in others of these sectors is confirmed in the case of Stockholm, Sydsverige and Västsverige. The orientation towards agriculture and mature-industry of the Noord-Nederland economy is also captured by the categorisation of the Dutch region in the boxes under discussion. A specific trends is shown by the Belgian regions, which appear in the top and bottom right hand boxes only. This reveals a poor local performance in ICT sectors overall, which allows for different strategies of European-owned foreign firms. European foreign R&D seems to compensate for indigenous weakness in almost all ICT sectors – as shown in the top right hand box – and neglect completely the regions in question in specific computing sectors because of location disadvantage – as in the case of the "special radio systems" group in the bottom right hand box.
CONCLUSIONS
Paradoxically, the knowledge-driven economy age is characterised by ICT as a pervasive core technology and, at the same time, by a re-discovery of geographical proximity as a key factor in building corporate advantage. As pointed out above, although the importance of the regional dimension in a heterogeneous socio-economic environment as Europe, a tradition of a spatial analysis of corporate economic activity has been slow to emerge by contrast to the well-established stream of literature focusing on locational factors in the US economy. Moreover, the few recent studies focusing on European regions have so far neglected small country realities. Within the new socio-economic conditions, this paper attempts to fill the gap.
First of all, the significance of the regional versus the national dimension in ICT corporate development of European small country regions has been empirically evaluated. Second, the geographical division of labour of European ICT corporations has been analysed. The results confirm the need for region-based analysis in order to identified intra-border inequalities and possible policy solutions for a more even long-run growth. By neglecting domestic dynamics, nation-State-based studies hardly provide a clear picture of local economic performance and growth. As far as the econometric results are concerned, the lack of a relationship between FDI and high indigenous specialisation in the technological sectors in question has been further explored. The interplay between foreign and indigenous research activity allows us to describe corporate sectoral location strategies shaped by the specificities of local socio-economic structures. The identification of the four corporate location strategies discussed above confirms the significance of local geography and the complementary of global and local phenomena. Globalisation and the spatial agglomeration of economic activity mutually reinforce (rather than excluding) one another. In this sense, MNCs’ operations can be considered as a connecting element between the two.
References
Antonelli, Cristiano. 1995. Localized technological change in the network of networks: the interaction between regulation and evolution of technology in telecommunications. Industrial and Corporate Change, Vol. 4, No. 4, pp. 737-754.
Antonelli, Cristiano. 1996. Localized knowledge percolation processes and information networks. Journal of Evolutionary Economics, Vol. 6, pp. 281-295.
Audretsch, David B. & Feldman, Maryann P. 1996. R&D spillovers and the geography of innovation and production. American Economic Review, Vol. 86, No. 3, pp. 630-640.
Audretsch, David B. & Stephan, Paula E. 1996. Company-scientist locational links: the case of biotechnology. The American Economic Review, Vol. 86, No. 3, pp. 641-652.
Cantwell, John A. 1987. The reorganisation of European industries after integration: selected evidence on the role of multinational enterprise activities. Journal of Common Market Studies, Vol. 26, No. 2, pp. 127-151.
Cantwell, John A. 1989. Technological Innovation and Multinational Corporations. Oxford: Basil Blackwell.
Cantwell, John A. and Iammarino, Simona, 1998. MNCs, technological innovation and regional systems in the EU; some evidence in the Italian case. International Journal of the Economics of Business, Vol. 5, No. 3, pp. 383-407.
Cantwell, John A. and Iammarino, Simona, 1999. Multinational corporations and the location of technological innovation in the UK regions. Regional Science, Vol. 33, forthcoming.
Castells, Manuel, editor. 1985 High Technology, Space and Society. Beverly Hills: Sage Publications.
European Communities. 1993. Portrait of the Regions, Vol. 1. – Germany, Benelux, Denmark. Luxembourg.
Eurostat. 1995. Nomenclature of Territorial Units Statistics. Luxembourg.
Feldman, Maryann P. & Audretsch, David B. 1999. Innovation in cities: science-based diversity, specialization and localized competition. European Economic Review, Vol. 43, pp. 409-429.
Feller, Irwin. 1975. Invention, diffusion and industrial location. Collins, L. and R. Walker, editors. Location Dynamics of Manufacturing Activity. London; John Wiley & Sons.
Hepworth, Mark. 1986. The geography of technological change in the information economy. Regional Studies, Vol. 20, No. 5, pp. 407-424.
Jaffe, Adam B., Trajtenberg, Manuel & Henderson, Rebecca 1993. Geographical localisation of knowledge spillovers, as evidenced by patent citations. Quarterly Journal of Economics, Vol. 58, No. 3, pp. 577-598.
Krugman, Paul 1991. Geography and Trade. Cambridge Mass.: The MIT Press.
Malecki, Edward J. 1994. High technology and local economic development. Journal of American Planning Association, Winter, Vol. 50, No. 1, pp. 262-269.
McCann, Philip. 1995. Rethinking the economics of location and agglomeration. Urban Studies, Vol. 32, No. 3, pp. 563-577.
Pavitt, Keith L.R. 1988. Uses and abuses of patent statistics. In van Raan; A. H., editor Handbook of Quantitative Studies of Science Policy. Amsterdam: North-Holland.
Santangelo, Grazia
D. 1998. Corporate technological specialisation in the European information
and communications technology industry. International Journal of Innovation
Management, Vol. 2, No. 3, pp. 339-366.
Telecommunications |
Other Electrical Communication Systems |
Special Radio Systems |
Image and Sound Equipment |
Semiconductors |
Office Equipment and Data Processing Systems |
AEG-Telefunken (DE) | ||
ASEA AB (SE) | ||
BICC (DE) | ||
Bosch-Siemens-Hausgerate (DE) | ||
Brown Boveri (SZ) | ||
CII-Honeywell Bull (FR) | ||
Compagnie General d’Electricité (CGE) (FR) | ||
Electrolux (SE) | ||
General Electric Co. (UK) | ||
ICL (UK) | ||
LM Ericsson (SE) | ||
Nixdorf Computer (DE) | ||
Nokia (FN) | ||
Olivetti (IT) | ||
Philips (NL) | ||
Plessey (UK) | ||
Racal Electronics (UK) | ||
Siemens (DE) | ||
Standard Telephones and Cables (STC) (UK) | ||
Thomson-Brandt (FR) | ||
Thorn EMI (UK) | ||
Zanussi Group (IT) | ||
Legend | DE: | Germany |
FN: | Finland | |
FR: | France | |
IT: | Italy | |
SE: | Sweden | |
SZ: | Switzerland | |
UK: | United Kingdom |
Belgian regions (NUTS 1) |
Dutch provinces
(NUTS 1)
|
Swedish counties
(NUTS 2)
|
Swiss regions
|
|
|
|
|
Bruxelles/Brussel |
Noord-Nederland
|
Stockholm
|
Zurich
|
Vlaams Gewest |
Oost-Nederland
|
Östra Mellansverige
|
St. Gallen
|
Region Wallon |
West-Nederland
|
Smäland med öarna
|
Basel
|
Zuid-Nederland
|
Sydsverige
|
Bern
|
|
|
Västsverige
|
Thun
|
|
|
Norra Mellansverige
|
Luzern
|
|
|
Mellersta Norrland
|
Davos
|
|
|
Övre Norrland
|
Lugano
|
|
|
|
Sion
|
|
|
|
Lausane
|
|
|
|
Fribourg
|
|
|
|
Geneve
|
|
* Italics denotes
the selected regions included in the econometric anlaysis.
|
|
|
|