Understanding the complementary effect of bike-sharing on public transit: A case study of subway line expansion in Xiamen, China

The rapid growth of dockless bike-sharing services (DBSs) in recent years has provided a new potential technological solution to the critical last mile problem of urban mobility and is thus considered a complement to public transit. However, there is little causal evidence of such technology-enhanced DBSs complementing public transit. To address this research gap, this paper empirically investigates the complementary effect of DBSs on public transit using a quasi-natural experiment on the opening of Subway Line 3 in Xiamen, China. Our difference-in-differences estimation results show that the number of DBS trips near newly opened stations increased by 84.7 % on average after the subway opening, representing 40.4 % of all bike-sharing trips within the 100-m buffer zone of the subway station entrance. The results of the spatiotemporal heterogeneity analysis revealed that such complementary effects were concentrated within 2 km of rail transit stations, especially during peak commuting hours and in areas with low bus accessibility and high population density. Moreover, the number of street intersections negatively affected the complementary effect of DBS on the subway when the connection distance was short (i.e., less than 800 m). These findings can inform city governments on the effective use of DBSs to solve the critical last mile problem of public transit and to improve urban mobility, while also illuminating how service providers can optimize capacity allocation and vehicle rebalancing decisions.