Modelling commuting tours versus non-commuting tours for university students. A panel data analysis from different contexts

University students’ mobility represents a significant part of the mobility demand, since the right to mobility becomes yet more significant, as it directly translates into the right to education. At the same time, lifestyle evolution and changes has yield to a boost in the number of non-commuting tours, which are now recognized as a key component of any travel demand system. However, their analysis is often overlooked due to their randomness and difficult detectability. Motivated by this shortfall, the current study sought to explore the university students’ mobility by focusing on i) a comparison among commuting and non-commuting tour, ii) analysing non-commuting patterns and iii) identifying factors affecting the tour generation. A joint mixed logit model was specified and estimated using panel data collected in two Italian Universities (Cagliari and Rome). This study represents a pilot test conducted for the purpose of providing scientific justification for implementing Voluntary Travel Behaviour Change programmes and Travel Demand Management policies in Italian Universities. Our results indicate that the number of non-commuting tours, when compared with commuting tours, is not negligible (around 28% of tours are non-commuting tours) and we detected no-significant differences between Cagliari and Rome with respect to the tour characteristics. In Cagliari women, individuals who have a high number of household members, people living in areas characterized by high building densities and a small number of shops, and in Rome students living in small families and those who own a car, are more apt to travel for discretionary purposes