A Spatiotemporal Approach for the Innovative Activity in Europe.

This paper provides exploratory empirical evidence on the innovation characteristics of the European countries using a spatiotemporal approach. Using Bayesian Additive Models for Location, Scale, and Shape (BAMLSS) a heteroscedastic Gaussian geoadditive model is specified. It incorporates different explanatory model terms for innovative activity, spatial autocorrelation, and a time effect and allows a flexible model fit close to the data observed. The timeframe of the dataset used in this paper covers the period from 2000 to 2012 and contains 224 regions. The main results show that spatial proximity and temporal considerations add to the explanatory content and provide insights about the innovative activity in Europe, supporting the importance of regionally differentiated considerations. The findings suggest that action schemes should focus on promoting cooperation among actors, facilitate R&D transfer, and consider spatial spillover effects. [ABSTRACT FROM AUTHOR]