Analyzing the Effects of Police Enforcement on Freeway Crashes Across Spatiotemporal Characteristics

Abstract: This study seeks to investigate the impacts of police enforcement on freeway crashes by analyzing data across various spatial and temporal dimensions. Previous research has often aggregated data at broader levels, missing critical segment-specific details that could refine understanding and intervention effectiveness. This study fills that gap by categorizing freeway segments with high spatial resolution. In addition, leveraging Vector Autoregression (VAR) and Structural Vector Autoregression (SVAR) models, this research will address the complex, interdependent relationships between enforcement (measured by traffic citations) and crash occurrences over time. The VAR model will capture lag effects, indicating how past citations influence future crash patterns, thus revealing patterns that contribute to the anticipation of risk. Additionally, SVAR models will assess the contemporaneous, or immediate, impact of citations on crashes. The SVAR findings are expected to show a significant, negative impact of enforcement in the short term, with larger effects in urban areas compared to rural segments. The study will further employ Impulse Response Functions (IRFs) to visualize how enforcement effects diminish over time. These IRFs are crucial for understanding the persistence of the enforcement impact. For example, while enforcement initially reduces crash risks, its impact typically diminishes within a few days. Understanding the effective duration across different segments can provide valuable guidance for strategic police allocation. Additionally, the study aims to incorporate contextual factors like weather and time-of-day effects to gain a more comprehensive understanding of the diverse influences on freeway safety. Ultimately, this research could lead to refined enforcement practices that prioritize high-risk segments with the greatest potential for crash reduction. By focusing on high-resolution data and advanced modeling, this study will inform data-driven, proactive approaches to traffic safety, supporting law enforcement efforts to effectively mitigate freeway crashes.