University–Industry collaboration in the biopharmaceuticals: the Italian case

We investigate the determinants of University–Industry (U–I) interactions in the biopharmaceuticals in Italy over the period 2004–2010, choosing co-publishing as a proxy of U–I partnerships. We construct a novel dataset of co-published articles, that contains measures of proximities, agglomeration, firms’ and universities’ characteristics. Following a consolidated methodology, we integrate our dataset of effective interactions with the set of all potential interactions, to estimate probabilistic models for the occurrence and the intensity of U–I interactions. Our main findings confirm and extend the predictions of the previous literature: (1) geographical proximity and prior partnership increase the probability and the intensity of co-publication; (2) the proximity of a firm to other biopharmaceutical firms and universities attenuates the relevance of geographical proximity; (3) there exists complementarity between prior partnerships and geographical proximity. A novel result is that firms’ and Universities’ size, firms’ R&D and patents expenditure and the composition of the academic staff as well as quality of academic research exert a significant impact on the intensity of co-publishing.