Determinants of crash injury severity for delivery riders: Insights from an error components mixed logit model with heterogeneous means and variances

Abstract

The rapid growth of e-commerce and food delivery services has led to an increase in commercial motorcycle riders, raising concerns about their safety on the road. This study aims to identify and analyze the determinants of crash injury severity for delivery riders in Thailand. Questionairs data was collected from 2000 commercial motorcycle users across five regions of Thailand, incorporating a wide range of demographic, work-related, and environmental factors. The study employs a Heteroscedastic Error Components Mixed Logit with Heterogeneity in Means (HECMLHM) model to capture unobserved heterogeneity and complex interactions between variables. Key findings reveal that rider age, experience, education level, income, work frequency, and rest periods significantly influence crash injury severity, often with varying effects across the population. Counterintuitively, more experienced riders faced a higher risk of severe injuries. Based on these findings, policy recommendations include targeted safety education programs, experience-based training to mitigate overconfidence, work schedule management, and optimized rest period policies. This study contributes to the field by focusing exclusively on delivery riders, employing advanced modeling techniques, and providing a comprehensive analysis of factors influencing crash severity in an emerging market context. The findings offer valuable insights for developing targeted safety interventions and policies to reduce crash injury severity among this growing workforce.