Safety Evaluation of Traffic Enforcement Improvement Program in Shanghai

Abstract

The “Shanghai Traffic Enforcement Improvement Program” was implemented on March 25, 2016, to address the longstanding issues of traffic crashes and violations in Shanghai. The program aimed to enhance road traffic safety through strict enforcement and targeted all traffic participants across the city, focusing on ten major violations. Despite the importance of such a program, there has been a lack of studies evaluating the effectiveness of these programs. Traditional methods, such as randomized controlled trials (RCTs), often face significant biases owing to the non-random nature of policy interventions, making accurate evaluations challenging. To address these issues, this study employs the interrupted time series method and introduces a Bayesian causal interrupted time series model, based on traffic crash and violation data from 2014 to 2019 in Shanghai. The findings indicate that the program led to a 25.58% reduction in traffic crashes and a 178.55% increase in traffic violations. This study identified the latency in crash numbers, revealing how policy interventions on violations affect the number of crashes. This study also analyzed the effectiveness of rectifying intersection violations, aiding in determining whether such violations should be a focus of enforcement. Finally, a placebo test was conducted by delaying the intervention time. This study primarily explains the estimated causal effect of interventions on the number of crashes and violations by incorporating Bayesian causal effects and a counterfactual framework. The estimated causal effect analysis revealed the latency in crash numbers and explained the principles by which interventions affect road traffic safety levels.