Crowding on public transport (PT) is a major issue for commuters around the world. Nevertheless, economists have rarely investigated the causes of crowding discomfort. Furthermore, most evidence on the costs of PT crowding is based on contingent valuation studies. First, this paper assesses discomfort with PT crowding over different density levels, trip durations and across different individuals using a different methodology. Based on a survey of 1,000 Paris PT users, the negative, linear relationship of in-vehicle density on reported travel satisfaction is remarkably similar to previous studies investigating PT crowding costs and stable across most individual characteristics. Contrary to the identifying assumption of most contingent valuation studies, we find little increase in crowding costs over travel time, in line with an additive specification of the generalized PT cost function. Second, we investigate the causes of this discomfort effect. We identify three key drivers: (a) dissatisfaction with standing and not being seated; (b) less opportunities to make use of the time during the journey; (c) the physical closeness of other travellers per se.
public transport crowding stated satisfaction travel cost survey data