Who and where are the co-authors? The relationship between institutional and geographical distance in scientific publications

This paper analyses the phenomenon of collaborations between university and other research institutions with industry, seen through the lens of co-autorship of scientific publications in the Italian biotech sector. We analyse a database including the publications done by the Italian biotech firms from 2003 to 2005; the institutions the authors of the publications belong to are registered and classified in four categories (firms, universities, hospitals and research centres); their localization is registered too. The main goal of our research is to analyse the relationship between geographical and institutional distance. Given that there is institutional distance if different kinds of institutions collaborate, we want to verify if such distance changes, and in what direction, when the physical distance increases. This analysis is conducted at an aggregated level, than at a more disaggregated one, taking two factors into consideration: the quality and relevance of the papers; the different nature and aims, therefore the different behaviour, of the different institutions. Regarding the empirical tools, the social network analysis is joined with the regression analysis. The more relevant results may be synthesized in this way: at a more aggregate level the direction of the relationship between spatial and institutional distance does not emerge with full statistical evidence; at a more disaggregate level the results emerge more clearly: taking into consideration the papers that receive few citations (that may be considered as results of project of limited scientific relevance or quality), the relationship between geographical and institutional distance is inverse (the variety of institutions is more limited among papers written as results of international collaborations than among papers deriving from national co-autorships); among more cited papers, the relationship is direct. On another side, taking the behaviour of different institutions into consideration, we observe an inverse relationship between spatial and institutional distance for firms, universities and research centres, a direct relationship for hospitals. This kind of analysis may shed more light on the way knowledge flows among innovative agents and should be taken into consideration by the policy maker that aims to promote research collaboration between different institutions.