Title: Report and Recommendations Regarding Automated Traffic Enforcement Systems
Author: Don Hillis Director of System Management
This is a 2011 study by the Missouri Department of Transportation on the effects of red-light cameras in the state of Missouri. This study compares the before and after crash statistics of 55 RLC intersections in that state. The study uses an undefined period of pre-RLC crash data compared to 12 to 36 months of post-RLC data. The statistics used in this analysis are from various municipalities all with different reporting standards and the report did not define what they used as their Distance-From-Intersection inclusion zone for crashes. There are no data or conclusions in this report that suggest a reduction or increase in fatalities alone due to RLCs. There was an overall 14% increase in total crashes and a 12% decrease in severe crashes. The study concludes that there was a: "44.7% reduction in fatality and serious injury right angle collisions!!", and "We believe automated enforcement is a good tool for keeping motorists safe".
Download the crash data below (336k)
Download the draft presentation below (222k)
Click to view the press release (external link)
This is a 3 page report of data together with a press release and 8 page draft presentation, the data and statistical analysis are not standard or thoroughly explained. The selection of 55 RLC intersections is large number of intersections for this kind of study. The lack of a defined Distance-From-Intersection inclusion zone for crashes somewhat diminishes the conclusions and consistency of the data. The exclusion of intersections where they have less than 12 months of real crash data helps with the validity of the data. There are no control intersections used to validate this data which greatly diminishes the validity of the report. There is no explanation of how exactly the data was analyzed in this report and some of the statistical analysis is inconsistent and not explained further reducing the validity of the study. The use of Bayesian analysis on the severe crash data is not a good fit for the data due to the extreme rarity of severe crashes at the studied intersections, for example more than 85% of the intersections studied had no severe right angle crashes at all during the before period, making the statistical change shown for severe right angle crashes insignificant. None of the conclusions of this study are supported by the data supplied.
Even though the press release mentions that the Missouri DOT is an impartial party for Red Light Cameras, they had the assistance of the red light camera vendors(which cannot be considered impartial) in the preparation of this report, something that should have been mentioned in the press release. Something else that should be mentioned is the the RLC crash data from the city of Washington, MO was specifically omitted from this study even though it fit the stated requirements of this study and the data was available to the authors. No reason was given for this omission, and Washington is not mentioned anywhere in this study materials.