Exploring the spatially heterogeneous effect of the built environment on ride-hailing travel demand: A geographically weighted quantile regression model

Investigating the relationships between ride-hailing travel demand and the built environment is of utmost importance to sustainable development. Previous studies have explored the heterogeneity in the relationships across different spatial regions or quantiles of travel demand distribution, but rarely both. To fill the research gap, this study develops a geographically weighted quantile regression (GWQR) model using ride-hailing order data collected in Shenzhen, China. The results show that: (1) the GWQR model outperforms the quantile regression model in predicting the expected travel demand and capturing the unobserved heterogeneity; (2) the continued increase in the number of ride-hailing services will reduce the attractiveness of bus, taxis, bicycle, and subway, which is more significant on weekdays; (3) Concerns about privacy and safety decrease the preference of pregnant women and female for this service, and the negative attitude do not eliminate as the service become increasingly popular; (4) men and youth prefer online-hailing services, and their interests will increase with the number of services provided; (5) the construction of commercial buildings is more important in peripheral areas than in downtown areas, although the measure may not initially enhance the generation of ride-hailing trips. These findings enable us to provide tailored strategies and suggestions for decision-makers.