Regional and firm-specific effects on innovations using multi-level methods

This paper analyses the determinants of different innovation types in Germany using the IAB (Institut fur Arbeitsmarkt- und Berufsforschung-Institute for Employment Research, Nuremberg) Establishment Panel 2006 to 2012 and merged regional data from various sources. The aim of the study is to test the empirical evidence of regional and firm-specific effects on innovations. For the empirical analysis variables of three dimensions are considered: availability and quality of local inputs, knowledge spillover, efficiency of regional innovation system and intra-firm capacity. The effects of these variables are estimated using three-level random effect logit models which account for the clustered and longitudinal structure of the data. In the focus is the empirical analysis of the expected impacts of regional and firm-specific factors on various categories of innovation activities in Germany. The differences between the various kinds of innovations became important when considering the regional and firm-specific variables in detail. While for radical innovations the unemployment rate, the share of MINT graduates, the R&D expenditures and industry's regional concentration have significant, but relatively small impacts, for other types of innovation the regional variables are insignificant. Overall the empirical results show that the regional intra-class correlation is small and the firm-specific variables are crucial for modelling innovation activities in Germany and dominate the regional dimension.