Evaluating the role of ride-hailing in multimodal travel to maximize travel utility in urban areas

Nowadays, with the widespread adoption of ride-hailing services, urban residents’ travel mode choice behavior has changed significantly. However, few studies have systematically compared ride-hailing with integrated, door-to-door multimodal public transportation—combining subway, bus, and walking—to better understand travelers’ preferences. In this study, we generated matched travel scenarios using real ride-hailing and public transportation trip records, ensuring consistency in departure times and real-time road traffic conditions. This combined dataset incorporated individual traveler characteristics and built environment factors as both internal and external influences on mode choice. Based on this comprehensive data, we developed an analytical framework and applied a logit model to analyze travel mode selection. Key model variables were identified using SHAP feature importance. Subsequently, we conducted comprehensive numerical experiments and sensitivity analyses to evaluate the impact of differentiated pricing strategies on travel mode choice probabilities and operator revenues. The model was applied to Xi’an, a city with a rapidly expanding subway network. Sensitivity analyses focused on key variables such as departure time, ride-hailing cost, and income. The findings reveal that high-income travelers exhibit lower sensitivity to ride-hailing costs, while low-income groups are more cost-sensitive, particularly during peak periods. Departure time significantly amplifies the effect of public transportation travel time on mode choice. All travelers show a strong preference for ride-hailing when the cost is below 30 RMB, whereas most opt for public transportation when the cost exceeds 50 RMB. Numerical simulations further demonstrate that targeted subsidies for low-income travelers can increase their probability of choosing ride-hailing by more than 30% during peak hours, without compromising operator revenue due to increased ridership and premium pricing for high-income users. These results highlight the need for targeted, data-driven policies to enhance both the efficiency and equity of urban multimodal transit systems.