The relationship between TFP and innovation performance: evidence from EU regions

Research background: Total factor productivity (TFP) determines how efficiently and intensely the available inputs are used and combined in production process. Improvement of TFP performance requires identification of its determinants, thus enabling policy actions to focus on them. Since the ability to create and absorb innovation is considered as a crucial factor of economic development, the investigation of the impact of the level of regional innovative performance on TFP distribution across EU regions is an important research problem.  Purpose of the article: The aim of the paper is twofold. Firstly, we attempt to develop a theoretical framework for the analysis of the impact of innovation on TFP. Secondly, we assess TFP levels for regions in the EU and investigate whether innovations account for the observed regional disparities in TFP. Methods: The research sample consists of 202 European Union (EU) regions at NUTS 1 and NUTS 2 level from 22 countries. The regional data on GDP, employment and gross fixed capital formation come from the Eurostat. The source of data on the level of innovativeness of European regions is the Regional Innovation Scoreboard. To calculate TFP, we use the multiplicatively-complete Färe-Primont index. In turn, to examine the impact of innovation on TFP, we employ the spatially-lagged X model. Findings & Value added: Our findings show a high degree of dispersion in TFP across EU regions. We find a positive impact of regional innovation performance on TFP. Although theoretical papers on economic development and regional economics suggest that improvements in TFP are key to regional economic performance, and that innovations are crucial to gain such productivity effects, there is a dearth of empirical studies on the link between innovation and TFP at the regional level. Therefore, our paper attempts to fill this gap by providing the evidence of positive effect of innovation externalities on TFP in European regions.