Do interdisciplinary research teams deliver higher gains to science?

The present paper takes its place in the stream of studies that analyze the effect of interdisciplinarity on the impact of research output. Unlike previous studies, in this study the interdisciplinarity of the publications is not inferred through their citing or cited references, but rather by identifying the authors' designated fields of research. For this we draw on the scientific classification of Italian academics, and their publications as indexed in the WoS over a 5-year period (2004-2008). We divide the publications in three subsets on the basis the nature of co-authorship: those papers coauthored with academics from different fields, which show high intensity of inter-field collaboration ("specific'' collaboration, occurring in 110 pairings of fields); those papers coauthored with academics who are simply from different "non-specific'' fields; and finally co-authorships within a single field. We then compare the citations of the papers and the impact factor of the publishing journals between the three subsets. The results show significant differences, generally in favor of the interdisciplinary authorships, in only one third (or slightly more) of the cases. The analysis provides the value of the median differences for each pair of publication subsets.