Bikeshare impacts on bus ridership: Unraveling the rail proximity effect

This research investigates the causal effects of bikeshare on bus ridership, with a special focus on quantifying how rail transit influences this relationship. Drawing on data before and after the introduction of station-based bikeshare service in the City of Chicago, we estimated two sets of fixed-effects models. Initially, a general fixed-effects model was employed to study the influence of bikeshare service provision on ridership at both rail stations and bus stops. Subsequently, we used a fixed effects model with interaction terms to delve into the specific effects of bikeshare on bus stop ridership while effectively controlling for the proximity to rail stations. Besides, the heterogeneous effects of bikeshare on bus stop ridership under varied household income levels was studied. Our findings reveal that the introduction of bikeshare service is associated with a significant increase in rail ridership at the rail station level and a notable decrease in bus ridership at the stop level on weekdays and weekends/holidays. Moreover, we observe a higher substitution of bikeshare for bus trips at bus stops located near rail stations, indicating a manifest overlap between feeder bus services and bikeshare in these areas. The heterogeneity analysis further suggests the more pronounced negative impact of bikeshare provision on bus stop ridership in neighborhoods with high household income. These findings emphasize the importance of resource rebalance considering the interconnected relationships among rail, bus, and bikeshare services. Rather than evaluating bus and rail separately, agencies should assess the combined impact of bikeshare, as this approach will aid them in allocating limited transportation resources (e.g., rail, bus, and bikeshare) more effectively to communities and individuals with the greatest needs.