The relationship between network position and regional innovation capacity taking spatiality into account

Abstract

According to the literature, geographical location and position within knowledge networks are two of the most important aspects of interregional knowledge flows. The latter is often measured by certain types of centrality measures that indicate how good a position is within a network. This paper applies these types of indicators as well, namely degree, closeness and betweenness centrality measures. It analyzes interregional knowledge flow between NUTS2 regions of the European Union relying on both relational data from the Framework Programmes and patenting data. At first, a network dataset is built based on the Framework Programme research and development collaboration projects. Secondly, using this relational data, OLS regressions and different spatial models are applied.

The goal of the paper is twofold. First, the authors attempt to verify the relationship between knowledge network position and regional patenting capabilities using the above network and centrality measures as main explanatory variables. The authors conclude that their hypothesis of a positive network position effect on patenting activity can be proven with the use of any of the three different types of centrality measures (degree, closeness and betweenness centrality). This effect can be shown in case of unweighted and weighted networks as well, even though the explanatory powers of the models are different in each case. These results are consistent with the findings of previous studies that show a connection between better network position of a region and increased regional innovation activity. The main contribution of the current study is that it verifies these effects on a European level for patenting activities as well, with the use of the Framework Programme network. The study also proves that there is a difference between the explanatory powers that different centrality measures hold. The second goal of the paper is to measure whether there are any spatial effects in interregional knowledge production. In other words, the paper seeks to shed light on whether regions can benefit from better patenting activity achieved by their neighbors who hold a better position in the presented knowledge network. Based on a spatial econometric analysis the authors conclude that with the use of certain centrality indicators in a spatial lag model, an interregional spatial knowledge spillover effect is present. However, this effect can only be shown in the cases of weighted degree and betweenness centrality, which are calculated by using the number of shared projects between regions as weights. These results show that spatial location also has an impact on the innovative performance of regions. Consequently, it can be stated that a region’s elevated patenting activity generated by a better network position positively affects the innovation capacity of neighboring regions.