Red light running at highway intersections has been a major safety concern in the United States because it often results in a disproportionally higher percentage of injuries than other type of... Show moreRed light running at highway intersections has been a major safety concern in the United States because it often results in a disproportionally higher percentage of injuries than other type of crashes. Statistics shows that about 45 percent of red-light running crashes would cause injuries compare to 30 percent associated with other type of crashes. Over the past decade, many city and county governments in the United States have begun to deploy red-light enforcement systems aimed to reduce red-light running related crashes for the benefits of safety improvements. In particular, the city of Chicago has implemented the red-light enforcement system since 2004. As of the end of 2010, more than 200 signalized intersections have been equipped with red light enforcement devices. This research attempts to evaluate the effectiveness of red light enforcement system deployed in Chicago in reducing types of target crashes including head-on, rear-end, angle, turning, sideswipe- same direction, and sideswipe- opposite direction by fatal, injury, and property damage only (PDO) crash severity category. Data on approximately 1,000 signalized intersections are collected for the analysis. These intersections are classified as three categories: untreated intersections without signal modernization and no red light enforcement, Type I treated intersections with both signal modernization and red light enforcement, and Type II treated intersections with signal modernization only. The Empirical Bayesian (EB) approach has been employed to quantify the crash reduction effects of red light enforcement that would avoid the regression-to-mean biases in the estimation. First, data on untreated intersections are used to calibrate safety performance functions for injury, PDO, and total crashes per intersection per year. The SPFs are then used to establish the EB-adjusted crash frequencies for Type I treated intersections for the before modernization period and Type II treated intersections for the before modernization period and red light enforcement period, respectively. The EB-adjusted crashes for Type I and Type II treated intersections for the after treatment period could be established by adjusting the before and after treatment period traffic volumes accordingly. The differences between EB-adjusted crashes and field observed crashes for Type I and Type II treated intersections for the after treatment period are the effect of signal modernization and combined effect of signal modernization and red light enforcement. The difference of the two effects thus represents the pure effect of red light enforcement. The effect of red light enforcement is further assessed by differentiating the Type I and Type II treated intersections according to number of through movement lanes and annual average daily traffic (AADT) range, respectively. For all target crashes including head-on, rear-end, angle, turning, sideswipe in the same direction, and sideswipe in the opposite direction combined, red light enforcement could on average reduce fatal crashes by 4 percent to 48 percent. Red light enforcement may increase injury crashes by 16 percent and decrease injury crashes by 8 percent. Red light enforcement may increase PDO crashes by one to 10 percent. The safety impacts of red light enforcement are separately assessed for different types of target crashes. For fatal crashes, red light enforcement may reduce fatal crashes by -1.4-1.2 percent for head-on crashes, 4-19 percent for rear-end crashes, 0.3-11 percent for angle crashes, -1.2-8 percent for turning crashes, 2-8.6 percent for sideswipe crashes in the same direction, and -0.4- 1.6 percent for sideswipe crashes in the opposite direction, respectively. Red light enforcement is generally ineffective in reducing PDO crashes. For all target crashes combined, the PDO crash increases could vary between 0.6 percent and 10.1 percent. For different types of target crashes, the highest PDO crash increases range from 0.2 percent for head-on crashes to 4.4 percent for rear-end crashes. Mixed results are obtained for red light enforcement on reducing injury crashes. For all target crashes combined, the injury crash could decrease as high as up to 8.2 percent and increase by approximately 16.3 percent. For different types of target crashes, the injury crashes may reduce by 4.7 percent for rear-end crashes to increase by 8 percent for rear-end crashes. Show less
