Europeanisation of science

Data from the Science Citation Index on scientific collaborations within and between European countries are used to address the question of whether the European science system is integrating over time. It is argued that a simple comparison of the number of national collaborations and European collaborations is misleading as a means of analysing European integration, as this procedure does not control for differences in countries’ sizes. The larger a country, the more collaboration is expected to be oriented nationally because there are more opportunities to interact within the national borders. An alternative statistical analysis is proposed that compares the observed propensities to collaborate with the propensities that would occur when partner selection is random. The results show that, typically, larger countries are better integrated in the European system when size is controlled for, which suggests that scale advantages render larger countries more attractive partners than smaller countries.