Modeling the impact of e-hailing services on regional public transit considering transit-dependent people

Abstract

We propose an analytical model to theoretically and numerically examine the impact of the introduction of e-hailing services into a region with traditional public transit, considering transit-dependent people who cannot drive themselves. For simplicity, this study considers many-to-one trip demand in a monocentric linear city. Our theoretical investigations confirm that in the short term when the service level of public transit is fixed, automobile users shift to public transit due to road congestion in the city center brought on by e-hailing services. For non-auto users, there exists an optimal connecting point where users transfer from feeder e-hailing services to public transit. Our theoretical analysis reveals that in the long term when public transit provider changes the service frequency according to the ridership, (1) the optimal connecting point remains at the same location, and (2) all public transit users in the suburb side of the optimal connecting point shifts to feeder e-hailing services. Our numerical investigations further reveal that the impact of e-hailing services on regional public transit depends on the regional dependency on automobiles. Shortly after introducing e-hailing services, the costs of users would significantly increase in a transit-dependent region, while the revenue of public transit significantly decreases in an automobile-dependent region. In the long term, the total cost for users is higher in all scenario settings compared to the base scenario without e-hailing services, which is analogous to the well-known Downs–Thomson paradox. Based on the above results, we discuss possible integrated regional transportation designs to realize desirable long-term outcomes.