Valuing crowding in public transport systems using mixed SP/RP data: the case of Santiago.

Abstract

This paper presents the valuation of comfort in public transportation using mixed Stated Preferences (SP) and Revealed Preferences (RP) data. In this case, comfort is measured mainly as the level of crowding in the vehicles (bus or train) using the in-vehicle passenger
density (pax/m2). To value comfort we use a stated preference survey where crowding levels are presented by means of figures, and revealed preference data on route choice inside the subway network of Santiago, Chile. The data collected in the surveys are used to estimate discrete choice models and obtain subjective valuation of passenger density through the parameters of the utility function. We assume that disutility for traveling in crowding conditions is proportional to the travel time; therefore, the longer the trip, the higher the utility loss. The results indicate that passenger density has a significant effect on utility of public transportation modes. In fact, marginal disutility of travel time in a crowded vehicle (6
passenger/m2) is two times the marginal disutility in a vehicle with low level of crowding (1 passenger/m2 ).