Regional Innovation Systems and Economic Performance: Between Regions and Nations

Abstract

This article aims at assessing the diversity of regional innovation systems and their economic performance in Europe. We propose to adapt the social systems of innovation and production (SSIP) framework at the regional level by identifying the specific arrangements of each part of the innovation and production system. Three key features of European regions are investigated using this framework: the diversity of regional SSIP, the interplay of regional and national determinants of such systems, and the impact of SSIP on regional performance. We identify a typology of regional configurations resulting from the combination of scientific, technological, educational and industrial indicators, using multivariate data analysis. A variance analysis approach is then developed in order to test the existence of specific regional growth regimes. The results highlight a persistently high level of diversity of regional configurations, notably among knowledge intensive regions, but also show that national institutional settings remain of fundamental importance in shaping a number of regional configurations. A final conclusion relates to the weak correlation observed between the structural characteristics of regions and their performance over the 2003–2007 period: regional performance remains primarily shaped by national trends. Overall, the paper questions the regional dimension of these “systems”.

Funding

This research is part of EURODITE Integrated Project (FP6), funded by the European Union [grant number CIT3-CT-2005-006187].

Notes

1 The four SSIP identified are the market-based SSIP, the social-democratic SSIP, the meso-corporatist SSIP and the public SSIP (Amable, 2000).

2 The analytical knowledge base refers mainly to science-based knowledge creation and absorption, using formal models and highly codified knowledge (publications, patents); Synthetic knowledge is typically related to tacit engineering processes, with a specific importance of problem-solving-related knowledge, notably through interactive learning with suppliers or customers, as well as application of existing knowledge by combination; Symbolic knowledge refers to “the economic use of various forms of cultural artefacts” (Asheim, 2007, p. 226) and includes activities such as branding, design, or advertising.

3 For most variables, except for patents and publications (only available as a 3-year average for the period 2000–2002).

4 See Appendix 2 for the detailed composition of these two subsamples.

5 The initial cluster criteria (high science or technology) isolates 53 regions, the population/density criteria add 5 regions (Lombardia, Zuid-Holland, Noord-Holland, Utrecht and Cataluña) and the complementary criteria on R&D intensity (public or private) add 5 other regions (Tirol, Alsace, Languedoc-Roussillon, Lazio and Groningen). 16 regions initially selected through the cluster criteria are excluded from the “high tech” group: 4 regions with less of 500,000 inhabitants and 12 with too low R&D, patents or publication levels.

6 Significant variables of each cluster are presented in Appendix 4.

7 Regions list is presented for each cluster in Appendix 2.

8 And actually reinforces national clustering of regions.

9 See Appendix 3 for the European weight of these profiles.