International, national and sectoral determinants of innovation: evolutionary perspective from the Czech, German, Hungarian and Slovak community innovation survey data

Abstract

This paper studies the determinants of innovation activities in Central European economies at the firm, sectoral, country and international levels. It focuses on innovation activities in four countries, the Czech Republic, Germany, Hungary, and Slovakia. Our dataset includes data from the three waves (2010, 2012, 2014) of the Community Innovation Survey (CIS), in a total of 58,880 observations, which we combined with 2-digit NACE rev. 2 sectoral variables (Herfindahl index, average values of age, profits, intangible assets, shares of large and medium-sized enterprises). These variables were calculated from 910,731 observations originating from the Amadeus database. We used Heckman and Bivariate Probit estimation procedures to observe the dynamics of new-to-the-market product and process innovation over time and at different levels of the innovation ecosystem. We focused on sectoral variables and found that determinants of innovation activity differed substantially. Our analysis documented the fluctuation of sectoral variables over time, which aligns with evolutionary theories. At the international level, the results suggest a disequilibrium effect of some sectoral variables over time. National innovation systems differ, and individual coefficients are hardly comparable without an innovation input-output context. We provided several recommendations for future research, such as the issue of non-linearity in some identified determinants.