Potentials for reducing spatial inequalities in innovation: A spatial econometric perspective

Localised capabilities and endowments to produce new knowledge are widely considered as the fundamental basis of a region´s innovative capacity, and accordingly viewed as a major source of spatial inequalities in innovation and, thus, economic competitiveness in the long-run. In this article we identify drivers for regional knowledge production, with the aim to illustrate potentials for reducing such inequalities in regional innovative capacities. We shift attention to technological knowledge production, measured in terms of a region´s patenting activity, specifically focusing on three main drivers for technological knowledge production outputs t the regional level. The drivers we focus on came into intensive discussion recently and are of particular interest in an inequality context since they capture very specific capabilities of the region's knowledge base, among them technological relatedness density, knowledge complexity and the average complementarity of technological knowledge bases of spatially close region. The analysis is conducted for 430 European NUTS-adapted regions, spanning the period of 2005-2019. Methodologically, a spatial Durbin panel model (SDM) is employed, allowing us to address a question largely neglected so far in the literature, namely, the identification of spatial effects of the analysed drivers of inequality in terms of knowledge creation, such as spillover effects of economic complexity or technological relatedness. The results of the estimated model are then implemented in different scenarios, where the future evolution of inequalities is simulated. Among them, a scenario that resembles the current path, and three other scenarios where the drivers under consideration, i.e., average relatedness density, average complementarity, and knowledge complexity, are synthetically altered for certain regions. The preliminary results point to a statistically significant positive effect of both spatial complementarity and average relatedness density on technological knowledge production, while the effect of knowledge complexity on the overall patent output appears to be negative for the observed European regions. These initial results are promising in terms of deriving conclusions for regional innovation strategies and policies, e.g pointing to different pathways for lagging regions to catch-up in terms of their innovation capacities. However, the results have to be consolidated looking deeper into the econometric estimations, in particular the spatial effects, but also in the scenario analysis still to be conducted.