Cost-based Road Safety Dynamics Under an Exogenous Mobility Shock: Evidence from Interrupted Time Series and Difference-in-differences on Rural Roads

Abstract

This study investigates cost-based road safety dynamics on rural roads under an exogenous mobility shock using an exposure-adjusted Accident Cost Rate (ACR). The analysis focuses on curve-related injury crashes on Category B rural roads in two Austrian states during 2012–2023. Geocoded crash records were integrated with official traffic counts to construct exposure-adjusted safety indicators at monthly, road, and road-segment levels. The mobility disruption beginning in March 2020 was treated as an exogenous intervention and analysed using interrupted time series models, complemented by difference-in-differences, count-data, and two-part models to evaluate regional heterogeneity and the robustness of the findings. The results show a statistically significant immediate reduction in the Accident Cost Rate following the onset of the mobility shock, mainly driven by reduced traffic exposure. However, the post-intervention period reveals a gradual recovery of cost-based crash risk despite persistently lower traffic volumes, indicating increased conditional crash severity when crashes occurred. These findings demonstrate that reductions in traffic volume do not necessarily result in sustained improvements in road safety when crash severity is considered alongside exposure. To support the interpretation of crash mechanisms, the study introduces a physically interpretable risk factor based on the ratio between operating speed and the critical curve speed derived from vehicle dynamics. Higher values of this indicator are consistently associated with increased Accident Cost Rates, linking roadway operating conditions with economic crash consequences.
The findings highlight the value of exposure-adjusted, cost-based safety indicators for evaluating rural road safety and supporting evidence-based infrastructure management and transportation safety decision-making.