Hu, Wen; Cicchino, Jessica B.
Insurance Institute for Highway Safety
July 2016

Introduction: Although numerous studies have demonstrated that automated enforcement reduces red light running, a growing number of communities have deactivated their red light camera programs in recent years. This study updates estimates of the effects of turning on cameras and offers a first look at the effects of turning them off.

Method: Among the 117 large U.S. cities with more than 200,000 residents in 2014, trends in citywide per capita rates of fatal red light running crashes and of all fatal crashes at intersections were compared among 57 cities that initiated camera programs during 1992-2014 and 33 cities without cameras during this period to examine the effects of activating camera programs. Trends also were compared between 19 cities that turned off cameras and 31 regionally matched cities with continuous camera programs to evaluate the effects of terminating camera programs. Because several cities turned cameras off during 2005-08, the estimated effects might have been confounded by the U.S. economic downturn immediately afterward. The primary analyses were limited to the 14 cities that turned off cameras during 2010-14 and compared trends in the 14 cities with those in 29 regionally matched cities with continuous camera programs. Poisson regression was used to examine the relationship of activating and deactivating camera programs with fatal crash rates.

Results: After controlling for temporal trends in annual fatal crash rates, population density, and unemployment rates, rates of fatal red light running crashes and of all fatal crashes at signalized intersections in cities with cameras programs were 21 and 14 percent lower, respectively, after cameras were turned on than what would have been expected without cameras. Rates of fatal red light running crashes and of all fatal crashes at signalized intersections in 14 cities that terminated cameras programs during 2010-14 were 30 and 16 percent higher, respectively, after cameras were turned off than would have been expected had cameras remained. Increases in rates of fatal red light running crashes (18%) and of all fatal crashes at signalized intersections (8%) in all 19 cities that turned cameras off were not significant.

Conclusions: The current study adds to the body of existing research indicating that red light cameras can reduce the most serious crashes, and is the first to demonstrate that terminating camera programs increases fatal crashes.

Practical applications: Communities interested in improving intersection safety should consider this evidence. Legislators and communities thinking about terminating camera programs should consider the impact to safety if programs end.