Modelling the interdependent relationship of motorcyclist injury severity and fault status: A recursive bivariate random parameters probit approach

Abstract

Motorcycles constitute the primary mode of transportation in Thailand. However, their prevalence has created an acute safety crisis, with motorcyclists representing over 70% of road fatalities. While previous studies have examined either injury severity or fault status in isolation, the potential interdependency between these outcomes remains poorly understood. This study addresses this gap by investigating motorcycle crash injury severity through both direct effects and indirect effects mediated through riders' at-fault status. This study examines motorcycle crash injury severity by considering both direct effects on outcomes and indirect effects mediated through riders' at-fault status. Analyses were conducted for three timeframes—2017–2018, 2019, and 2020—to capture potential temporal instability. A recursive bivariate modeling framework with random parameters was adopted to address unobserved heterogeneity and endogeneity between fault likelihood and crash severity outcomes, revealing key demographic, behavioral, and infrastructural predictors of fatal injuries. Findings indicate that younger riders, nighttime conditions, and high-risk behaviors (e.g., speeding, and alcohol use) increase the probability of being at fault. Meanwhile, factors such as being male, older, riding without a helmet, traveling against traffic, and riding at night significantly raise fatal injury risks. Although at-fault data provide valuable insights, they must be interpreted alongside roadway conditions, as Thailand's infrastructure offers limited protection for motorcyclists. Accordingly, this study recommends an integrated approach that combines improved infrastructure (e.g., motorcycle-friendly barriers and segregated lanes), robust education initiatives (targeting speed, helmet use, and alcohol awareness), and enhanced enforcement of traffic regulations. Implementing or enhancing licensing standards—such as a Graduated Licensing Scheme—can help curb risky behaviors and foster safer riding practices among young riders. These evidence-based recommendations can help policymakers develop more effective mitigation strategies to reduce the number of severe and fatal motorcycle-related crashes.