Does Coauthorship matter for Scientific Productivity? Evidence from Geography's Top Journals

The aim of this paper is to investigate the relationship between scientific productivity and \r\ncollaborative behaviours (formal and informal). Despite the different approaches available in the \r\nliterature, we will focus on what we call the “relational trend”: our goal is to detect some of the \r\nfactors which might affect researchers’ productivity, considering “relational variables”. In \r\nparticular, the tendency to write papers in co-authorship will be used as a proxy of formal scientific\r\ncollaboration amongst scientists, while the number of acknowledgements will be assumed as a \r\nproxy of the scientist’s ability to build informal collaboration networks. Both co-authorships and \r\nacknowledgments indices are interpreted as two of the main forces which could affect and drive \r\nscientific production, apart from individual talent. Using the dataset developed by Togni (2009) \r\nwhich collects data about geographers’ publications on the Top Journals in the years 2000-2007, an \r\neconometric analysis using two-stage least squares has been performed, in order to regress \r\nproductivity on a series of other indices, including (amongst the others) a typical SNA index of \r\ncentrality (betweenness centrality). Three results clearly emerged from the analysis: co-authorships \r\nnetworks affect productivity in a negative way, but a variety of co-authors may increase \r\ngeographers’ productivity; on the contrary, informal influence (acknowledgments network) on \r\nproductivity seems not to have any effect on productivity. Finally, a trade-off between the \r\ntransaction costs from the collaboration and the need to mutual exchange of skills and knowledge \r\ncomplementarities which boost scientists to vary their co-authors.