Here’s a link to the story in the Houston Chronicle:

http://blog.chron.com/newswatch/accidents-fall-at-houston-red-light-camera-intersections/

The story says that:

In the five months after Houston voters forced city officials to turn off a camera surveillance system that fined motorists for running red lights, traffic accidents at those 50 intersections with 70 cameras have decreased 16 percent, according to recently released data.

There were lot’s of reasons given by officials for this unexpected outcome. Everything from the “weather was good” to “the camera’s had trained people to be safer.”

The interesting statistic that no one mentioned was that it is usual for rear-end collision to increase when red light cameras are installed because, to avoid a ticket, people slam on their brakes when a light turns red and they get rear ended.

There are at least two lessons that I think you can learn from this article.

1. People don’t know how to trend infrequently occurring accident statistics.

In this case, no one on either side of the argument used advanced trending techniques to prove their point. Instead, they chose the statistics that best fit their argument and claimed that those stats proved their point.

2. Sometimes corrective actions can have unintended consequences.

Several times in the past we’ve discussed red light cameras as an enforcement tool and the consequences that the tool could have on accident statistics. Our general opinion is that the cameras would be great for raising revenue but would do little to improve safety. For several reasons, rear end collisions were an unintended consequence of red light cameras that tend to increase accident rates at intersections where the devices were installed. So all people looking to improve performance should learn that your corrective actions may have other consequences than the ones you intend them to have!