Does the information provider choose to promote public transport use or protect public health? A case study in China

Background: Health risks from air pollution require cutting motorized private vehicle use. Providing new public travel patterns, along with appropriate smartphone-based multimodal travel information services (SMTIS) is considered an effective way to alleviate this stress. However, air pollution is not likely to be solved overnight. There is a growing demand for more detailed air pollution information accessible through mobile apps to avoid exposure to pollution, which may increase the use of private cars. So far, little is known about this contradictory behavior. Methods: Based on 1926 observations obtained from a stated preference experiment conducted in Shanghai, a mixed logit model was developed to investigate the effect of SMTIS on travelers’ mode choice behavior. Specifically, this study aims to reveal the impact of progressively providing typical multimodal information, social influence information, and air pollution information expressed in cigarette equivalents on commuters’ choice of travel modes, including private car, park-and-ride (P&R), and customized bus. Results: Information offered by SMTIS, including travel time, travel cost, available parking spaces, comfort level, and air pollution information, exerts significant influences on travel mode choice. Also, there is observed taste heterogeneity for such information across the population. Respondents placed less emphasis on air pollution information during commute trips than on travel time. Providing information on modal pollution exposure prevents people from using public transport. Even when air quality is good, we still observe an increase in the proportion of people choosing cars. With regard to social influence information, people were more likely to choose P&R when online word-of-mouth about P&R displayed a positive propensity, while the percentage of other travelers who had selected P&R did not exhibit a significant effect. Implications: The findings provide valuable implications for the design of low-carbon traffic management policies and the deployment of multimodal information services.