Turning off the cameras: Red light running characteristics and rates after photo enforcement legislation expired

► Red light running rates at camera-enforced locations increased after the program's enabling legislation expired. ► Recidivism at previously camera-enforced locations could not be explained by driver characteristics alone. ► This is the first known peer-reviewed study documenting impacts of turning off automated enforcement. In 2005 the Virginia legislature allowed the law permitting automated enforcement for red light running violations to expire. An opportunity presented itself to evaluate what would happen to red light running behavior at formerly enforced locations. Using intersections previously studied to document one city's deployment and use of photo enforcement (see Martinez and Porter, 2006), we mobilized multiple pre-expiration, immediate post-expiration, and one year post-expiration observations at camera-enforced intersections as well as two control groups consisting of same-city and a different city's non-camera locations. More than 2700 direct observations were made in these time periods, documenting the near-immediate increase in red light running at previously camera-enforced intersections. These intersections had a rate that nearly tripled immediately after the law expired, and more than quadrupled one year later. Further, within a year of the law's expiration, the low red light running rates at the previous-camera locations had recidivated to red light running rates of the control locations. Driver characteristics were not significant predictors of these rates once intersection group and traffic volume (and their interaction) were controlled, meaning red light running in this study was not linked to a particular driver type. Our results are important for scholars of intersection safety, as this is the first known peer-reviewed study documenting estimates of what could happen when automated enforcement is removed.