One or many knowledge production functions? : Mapping innovative activity using microdata

This paper discusses the determinants of three alternative measures of innovative output by looking at firm's own formal R&D activities and at the acquisition of external technology (TA) in its embodied and disembodied components. These input-output relationships are also discussed by distinguishing between small and large firms and those belonging to low-tech and high-tech sectors. The empirical analysis focuses on the Italian industrial sector over the period 1998-2000, using a subsample of 2,949 firms from the third European Community Innovation Survey (CIS 3). A bivariate probit analysis framework is used to investigate the determinants of product and process innovations, while truncated regressions are used to discuss innovation intensity. This paper also discusses an alternative test procedure that permits an extension of Cragg's test in the analysis of survey data with weighted observations. Results show that R&D is strictly linked to product innovation, while TA is crucial in fostering process innovation; however, both inputs increase a firm's innovative intensity. Significant evidence is also found that small firms and firms belonging to low-tech sectors rely more on the acquisition of external technologies and on cooperation agreements, while larger firms in high-tech sectors rely more on their own formal R&D. -- R&D ; product and process innovation ; embodied technical change ; CIS 3 ; bivariate probit ; Cragg's test