The citation is given in this article, which says there's a 15% decrease, and not the 47% of the nonscientific comparisons: > "But the Tribune study, which acc

The citation is given in this article, which says there's a 15% decrease, and not the 47% of the nonscientific comparisons: > "But the Tribune study, which accounted for declining accident rates in recent years as well as other confounding factors, found cameras reduced right-angle crashes that caused injuries by just 15 percent." It also includes the analysis that revelation calls for (the balance of severity and incidence rate): > "The economic analysis examined the extent to which the increase in rear-end crashes negates the benefits for decreased right-angle crashes," the 2005 study concluded. "There was indeed a modest aggregate crash cost benefit of RLC systems." which is why "The Tribune researchers also suggested that Chicago could reap much better results if it shut down about 40 percent of its cameras, located in about 75 intersections throughout the city where crash rates were lower to begin with."