Heterogeneity in Women's Nighttime Ride-Hailing Intention: Evidence from an LC-ICLV Model Analysis

TL;DR

This study investigates the heterogeneity in women's nighttime ride-hailing intentions using a novel LC-ICLV model, revealing two distinct female subgroups with different influencing factors, informing tailored safety and policy strategies.

Contribution

Introduces a mixed LC-ICLV model to capture unobserved heterogeneity in women's ride-hailing decisions, identifying two distinct female subgroups with different decision drivers.

Findings

Two female subgroups identified: Attribute-Sensitive and Perception-Sensitive.

Service attributes influence young women more, safety perceptions affect older women.

Tailored safety features and policies are recommended for different subgroups.

Abstract

While ride-hailing services offer increased travel flexibility and convenience, persistent nighttime safety concerns significantly reduce women's willingness to use them. Existing research often treats women as a homogeneous group, neglecting the heterogeneity in their decision-making processes. To address this gap, this study develops the Latent Class Integrated Choice and Latent Variable (LC-ICLV) model with a mixed Logit kernel, combined with an ordered Probit model for attitudinal indicators, to capture unobserved heterogeneity in women's nighttime ride-hailing decisions. Based on panel data from 543 respondents across 29 provinces in China, the analysis identifies two distinct female subgroups. The first, labeled the "Attribute-Sensitive Group", consists mainly of young women and students from first- and second-tier cities. Their choices are primarily influenced by observable service attributes such as price and waiting time, but they exhibit reduced usage intention when matched with female drivers, possibly reflecting deeper safety heuristics. The second, the "Perception-Sensitive Group", includes older working women and residents of less urbanized areas. Their decisions are shaped by perceived risk and safety concerns; notably, high-frequency use or essential nighttime commuting needs may reinforce rather than alleviate avoidance behaviors. The findings underscore the need for differentiated strategies: platforms should tailor safety features and user interfaces by subgroup, policymakers must develop targeted interventions, and female users can benefit from more personalized risk mitigation strategies. This study offers empirical evidence to advance gender-responsive mobility policy and improve the inclusivity of ride-hailing services in urban nighttime contexts.