What determines the embeddedness of European regions in EU funded R&D networks? Evidence using graph theoretic approaches and spatial panel modeling techniques

This study shifts attention to the embeddedness of European regions in R&D networks, as captured by R&D joint ventures funded by the EU Framework Programmes (FPs) in the time period 1998-2006. By regional embeddedness we refer to a region´s position in terms of different network analytic centrality measures, given aggregated region-by-region collaboration flows in the FP networks. Network embeddedness is measured by a region´s betweenness centrality, a proxy for a region´s ability to control knowledge flows, and eigenvector centrality, measuring a region´s connectedness with central hubs. The objective is to estimate how different region-internal and region-external characteristics affect a region´s embeddedness in the European network of R&D cooperation. We consider independent variables accounting for the knowledge production capacity of a region and the regional economic structure. In modelling regional network embeddedness, we make use of advanced spatial econometric techniques by means of panel spatial error models and panel spatial durbin error models with random effects. The results show that R&D expenditures, human capital and regional technological specialisation are the most important determinants for a region´s network embeddedness. Further, the study provides evidence that spatial spillovers from neighbouring regions influence a region´s FP network embeddedness.