Abstract:
This paper employs firm-level data to analyze the relative importance of firm characteristics and agglomeration externalities in explaining variation in innovation rates across firms. More specifically, we combine micro-data and census data to estimate the probability that a firm will introduce a goods, service or process innovation. We consider internal firm-level characteristics as well as externalities, using information on the regional production structure to test for Marshall-Arrow-Romer, Porter and Jacobs effects. Our results show that most firm-specific variables are highly statistically significant, whereas agglomeration variables are only significant for a few specific sectors, and even then only for some types of innovation.