The causal effects of removing Hook-Turn regulation on road safety

Traffic safety and related policy interventions have garnered increasing attention from both scholars and policymakers. While prior research has primarily focused on the role of infrastructure improvements in mitigating traffic accident risks, relatively little attention has been given to the impact of hook-turn (HT) traffic regulations. HT is a specialized traffic regulation implemented at intersections, requiring vehicles to proceed to the far side of the cross street and wait for the green light in the intersecting direction before completing their turn. This regulation has been adopted in countries or regions such as Japan, Australia, and Taiwan. This paper empirically examines the causal effects of HT regulation on traffic accidents by exploiting a policy reform in Tainan City, Taiwan, where HT was removed in some townships. Using this policy change as a quasi-natural experiment, we apply a difference-in-discontinuity design to estimate its impact on traffic accident outcomes, including the number of accidents, the number of victims, and the number of vehicles involved. Our findings indicate that the policy reform led to a 21% reduction in accident cases and a 19% decrease in the number of victims. The reduction is primarily driven by injury-related rather than fatal incidents. Additionally, the total number of vehicles involved in accidents declined by 28%, with larger reductions observed for motorcycles (–26%) than for automobiles (–18%). A back-of-the-envelope calculation suggests that the policy resulted in a 4.7% decrease in total medical expenditures among residents in the treated areas during the study period.