Measurement and explanation of the intensity of co-publication in scientific research

In order to study networks of collaboration between researchers, we propose a simple measure of the intensity of collaboration, which can be easily interpreted in terms of relative probability and directly aggregated at the laboratory level. We use this measure to characterize the relations of collaboration, as defined in terms of co-publication, between the physicists the French 'Centre National de la Recherche Scientifique' (CNRS), in the field of condensed-matter, between 1992 and 1997, and to investigate how they vary with regards to various factors: mainly the geographical distance between laboratories, but also their specialization and size, their productivity and the quality of their publications, and their international openness. We find that the average intensity of co-publication within laboratories is about 40 times higher than the intensity between laboratories but within towns, and 100 times higher than the intensity between laboratories and between towns. Yet, geographical distance does not have a significant impact, or a very weak one, on the existence and intensity of co-publication of researchers located in different towns. We also find that the productivity labora