Investigating the determinants of a Peer-to-peer carsharing. The case of Milan

Carsharing systems are seen as a promising alternative to private mobility. Despite several systems are flourishing worldwide, their effective impact on the private mobility still remains negligible. One of the possible evolutions of carsharing system is to move towards a “peer to peer carsharing” (hereafter P2P): a carsharing in which cars are not centrally provided, but are owned by individuals who rent them when unused, in exchange of a monetary reward. Starting from a survey carried out in the year 2012, the paper aims at investigating the main determinants to adopt a P2P carsharing among the inhabitants of the municipality of Milan. Specifically, the probability to undertake a P2P carsharing is investigated by means of a descriptive statistics and two discrete choice models: binomial logit model and multinomial logit model, where the multinomial logit model allows to investigate the probability to share the personal car among different groups of people (nobody; friends, colleagues, and neighbours; all people). To our knowledge the paper represents the first attempt to investigate the propensity to join a P2P scheme by sharing the personal vehicle.