Knowledge creation vs knowledge co-production : knowledge intensive business servises and innovative activity in EU regions : prepared for the ERSA Congress, Bratislava August 21-25 2012

Regional economies are continuously evolving shifting from more traditional manufacturing toward more service-oriented production systems. Despite the increasing relevance of services, however, the analysis of innovation at the regional aggregate level has mainly focused on manufacturing, gathering the attention on the role of R&D expenditure as input in the production process and, in some cases, accounting for research-based knowledge externalities. In this paper the role of Knowledge Intensive Business Services is studied and their contribution to the regional aggregate innovation is evaluated. The aim is twofold. First is to provide insights on the role covered by KIBS as a second knowledge infrastructure. Second is to examine the extent to which KIBS operate as bridges between the general purpose analytical knowledge produced by scientific universities and more specific requirement of innovative firms. A role commonly acknowledged to KIBS is in fact that of knowledge transferors. If on the one side it is however clear to whom they transfer knowledge, their client firms, on the other it is not as clear from whom the knowledge is originally transferred. For this reason a major attention in this work is dedicated to scientific universities considered as a primary source of knowledge. Being this knowledge analytical and highly codified, it probably can be more easily accessed by nearby located firms having higher opportunities of research collaboration and less easily by firms located in different regions. It is argued that KIBS, in transferring knowledge from universities to firms, are therefore specially important in the latter case. To test hese hypothesis a knowledge production function is estimated for a sample of 200 EU NUTS II regions including also information of university research and KIBS concentration. Parameters are estimated using the heteroschedasticity-consistent G2SLS estimator for spatial models and the evidence suggests that the contribution of KIBS to regional innovation is considerable. In fact accounting for the knowledge embedded in business services can considerably contribute to explain the cross-regional variation in innovative activities. Furthermore it is find that the KIBS contribution is more sizeable in regions in which there are not scientific universities. The highlighted results have important policy implications asking to rethink to how much effective an R&D-centered innovation strategy could be, at least in some regions.