The Construction of the Academic World-System: Regression and Social Network Approaches to Analysis of International Academic Ties
This paper explores factors responsible for strength of various forms of academic ties between countries. It begins with examining several theoretical models of international academic collaboration: “the republic of letters”, “academic (neo)colonialism”, “the classical world-system”, and “the world-society”. Propositions about factors affecting intensity of ties between countries and configuration of their overall network are then derived from each of the models. These propositions are then tested against empirical data on two kinds of academic ties: volumes of international student flows between pairs of countries (UNESCO statistics) and number of co-authored papers (Web of Science database). Negative binomial regression is used to estimate influence of various independent variables (funding of science, distance, historical experience of dependency) about the significance of which the models make different predictions. We discover that expectations associated with “the classical world-system” fit the data best, with “academic neo-colonialist” factors also important in the case of international student flows. To account for possible differences between disciplines and to capture the directions of evolution of the system, we then explore changes in international collaboration network in two fields: geoscience and economics during a 30-year interval (1980-2010).