The world of scientific research has undergone dramatic changes in the last decades. Increased specialization, collaborations are becoming more and more diffused in the community of social scientists, while the practice of evaluating scientific research and personnel through bibliometric indicators is increasingly used by departments, universities, government bodies and funding agencies.
These two phenomena are hardly independent; on the contrary it has been shown that they are strongly interdependent since the increased pressure to publish on academics has caused a rinsing propensity to co-authoring papers due to a series of demand-side and supply-side factors.
Aim of this paper is thus to analyse how the collaborative behaviour (i.e. co-authorships) of Italian economists influences their scientific output. In particular the empirical analysis is based on an original database based on ECONLIT and MIUR-Cineca databases.
In the paper, the structure of co-authorship patterns is firstly investigated through Social Network Analysis techniques, to obtain relational metrics (density, clustering coefficient, average path length) and analyse its topology (searching for small world properties). We then rely on econometric estimations, to explain the individual productivity of Italian economists, in terms of both “attributional” variables (age, gender, academic position, tenure, geography, etc.) and “relational” variables (propensity to cooperate and stability in cooperation). We also estimate the role of “positional” variables (such as betweenness and closeness centrality indexes and clustering coefficient) in explaining the scientific productivity of Italian economists.