“Co-Opetition” Effects of Ride-Hailing on Metro Services and its Spatially Varying Determinants: Evidence from Shanghai, China

The rapid expansion of ride-hailing services has sparked intensive debate on their competition and cooperation with public transit such as metro systems. However, empirical evidence on such “co-opetition” effects remains limited, as well as on their spatiotemporal characteristics and influencing factors. To address this gap, we developed a three-level framework for detecting the co-opetition effects of ride-hailing on metro services based on ride-hailing trip data from Shanghai in China. We then performed exploratory analysis to reveal the spatiotemporal patterns of these effects. Lastly, we employed Poisson regression models to assess the impact of trip characteristics, the built environment, and socioeconomic indicators on various co-opetition effects. The results show that: 1) the co-opetition effects of ride-hailing on metro services exhibit distinct temporal patterns, with the complementary effect predominantly occurring during the initial morning peak hours, while the substitution effect is more prominent toward the end; 2) ride-hailing tends to substitute primarily for metro services in the city center, while complementing and supplementing them in the suburban areas; 3) the density of metro facilities is negatively associated with the supplementary effect and potential complementary effect; 4) the potential complementary effect can be categorized as front-end and back-end connections, with front-end connections being more susceptible to the density of facilities at the trip origin, while the back-end connections are influenced by the density at the destination; and 5) travel distance and road congestion demonstrate a negative correlation with the substitution effect, whereas road network density exhibits a positive correlation.