Innovation decision and firm productivity: An empirical analysis for the services sector

The primary aim of the paper is to explore the relationship between productivity and innovation based on a large firm-level data set for service and manufacturing sector in the period 2010-2011. The novelty of the approach lies in make an analysis from innovation surveys of the Colombian Statistical comparing Knowledge intensity business sector (KIBS) and manufacturing. Innovation activities of firms are modelled as a four stage model (CDM) which allows studying several interrelated questions while controlling for public support for innovations. In the first two stages determinants of being supported by the government and the decision to innovate and consequent innovation investment are separated. In the third stage innovation input (innovation investment) is linked to innovation output, and finally, in the fourth stage it is determined how the productivity of firm is related to its innovation activities. In this final stage we explore the heterogeneous effects across the productivity distribution function using quantile regressions. Our analysis proved that innovation input significantly increases innovation output, but it does not true for services companies. This result may throw a shadow on the efficiency of supported firms and have some implications for competition policy.