Identifying Important Roadway Corridor Factors which Affect Crashes Influenced by Unsafe Driving Behaviors to Help Law Enforcement Proactively Reduce Crashes

Abstract

Effective allocation of traffic law enforcement officers can proactively reduce crashes by stopping unsafe drivers before crashes occur, but this can be difficult because of staffing limitations. Therefore, it is important to identify roadway factors that could encourage unsafe driving behaviors so appropriate proactive enforcement methods can be taken. In this paper, signalized arterial roadway corridors in Florida are modeled to identify the important factors that affect the crash rate for seven unsafe driving behaviors. Using a standardized corridor definition, 406 corridors in seven urban Florida counties were identified. These corridors contained 18,518 crashes, including 5053 fatal and injury (FI) crashes, from 2017 to 2021 involving the studied behaviors. Three random forest regression models were developed to identify the important corridor-level factors that affect the crash rate for all crashes (model 1), FI crashes (model 2), and only FI crashes caused by careless or reckless driving (model 3). Corridors in Pasco County, corridors with greater signalized intersection densities, and corridors with six or more lanes had above average predicted crash rates for all three models. In addition, corridors with speed limits greater than 45 mph (model 1), two lanes (models 1 and 3), no school zones (models 1 and 3), bike lanes (models 2 and 3), and no horizontal curvature (models 2 and 3) were predicted to have higher than average crash rates. By considering these factors when making allocation decisions, law enforcement agencies can likely reduce crashes, injuries, and fatalities caused by these unsafe behaviors.