Clustering of Italian Higher Education Institutions Based on a Destination-Specific Approach

Abstract Student mobility flows are typically examined by using gravity models. These models generally assume the same relationship for each origin-destination pair. This assumption may not hold for the Italian higher education system, where, as a consequence of the decentralization process, some universities have a national vocation, while others primarily target local populations. For this reason, we adopt a destination-specific approach to gather detailed insights for each university. Indeed, our empirical analysis reveals distinct interaction dynamics among Italian universities, corroborating our hypothesis of spatial heterogeneity. Furthermore, we propose a fuzzy clustering technique based on the estimated distance parameters that measure the deterrence effect. Through this clustering procedure, we identify two distinct groups: one comprising universities with a national vocation and the other comprising those with a local focus.