Research orientation and agglomeration: Can every region become a Silicon Valley?

Not only success stories, such as Silicon Valley, but also non-success stories can inform regional innovation policy. In order to provide a benchmark for regional innovation systems we compare both success and non-success stories. Regional innovation systems differ in structural and functional requirements, because development processes are path dependent. We suggest that regions' development paths emerge from agglomeration patterns and research orientation. Accordingly, we have developed a typology of regions including (1) their agglomeration patterns (either MAR or Jacobs' type) and (2) the degree to which their research is predominantly oriented towards obtaining fundamental understanding or addressing considerations of use. We combine qualitative and quantitative data on thirty-six European regions to categorize them according to research orientation and agglomeration, thereby developing a typology. We use this typology and some basic quantitative economic data to see how success and non-success regions are distributed. Our results indicate that a better understanding of how to combine agglomeration patterns with research orientation can guide context-sensitive policy.