Effects of numerical platforms and matching characteristics on individual choices and social welfare: the case of short-distance carpooling

This paper investigates the effects of carpooling organization and individuals' idiosyncrasies on the propensity to carpool. We test whether pre-planning the trip, use of a platform, and socio-demographic similarities or differences between driver and passenger affect choices towards carpooling over solo-driving and public transportation. A stated choice experiment collected answers of 3,600 inhabitants from in the Lyon's urban area, France. Our results indicate platforms raise the individual willingness to carpool, and that this platform's effect is larger for passengers than for drivers. Other things being equal, we identify a clear gender effect: women are more easily accepted as carpoolers than men, and they also are more sensible than men to their carpooler's gender. By contrast, the age of the proposed carpooler does not affect carpooling choices. We illustrate these results with a stylized social welfare analysis which highlights a significant contradiction between what it would take to make carpooling beneficial from a welfare perspective – drivers should pay passengers – and what seems to be individually acceptable.