Spatial knowledge spillovers in Europe : a meta-analysis

In this paper we focus on one type of externality, namely knowledge spillovers. Empirical studies on effects of knowledge spillovers in Europe have normally focused on localized effects, either on total factor productivity or knowledge production in terms of patent output. The purpose of this paper is to quantitatively review the empirical literature on spatial knowl-edge spillovers in Europe by means of meta-analysis. Our aim is to determine the extent to which such spillovers have been empirically documented as well as the spatial reach of these spillovers. In addition, we will apply meta-regression-analysis to analyze the determinants of observed heterogeneity across and between publications. Our results show that if total local R&D expenditure in a European region increases by 1%, the number of patents in that region increases, on average, by 0.482%. Spatial knowledge spillovers induce a positive effect on local knowledge production, however, this effect proves to be marginally small. Spatial weighting regime seems to matter. If R&D expenditures in other regions are weighted by distance in kilometers or minutes (instead of a binary contiguity matrix) then the spillover effect will on average be larger. Also, public R&D expenditure is found to have a lower impact on local patent production compared to private R&D expenditure.