Cleveland is one of the many cities that are deploying red light cameras. These are the cameras that catch cars running red lights and mail you a ticket afterward.
I just saw
an article [autoblog.com] claiming that red light cameras increase the rate of accidents at intersections, while bringing in a lot of revenue.
So, now there are people on one hand arguing that red light cameras make the roads safer by changing the way people drive, and people on the other hand arguing that these devices are actually increasing the rate of accidents, thereby making the roads less safe.
First off, it's difficult to tell how accurate those studies are. Red light cameras are fairly new, so it's difficult to determine whether the accident rates will return to normal.
I am opposed to red light cameras, though. The problem is that they provide an incentive for the city to not provide an adequate yellow light duration. If the city were to shorten the duration of the yellow lights by 10%, their increase in revenue would probably be huge, even if the actual change were almost imperceptible.
Furthermore, if there are lights that already had insufficient yellow light durations, it would be against the city's interest to fix the timing of those lights.
I'm sure it would be no surprise to most people if studies were done that concluded that most parking meters did not keep the correct time. In other words, if you put in enough coins to park legally for 30 minutes, and the studies found that the parking meters expired after 29 minutes, that wouldn't be shocking. If you got a parking ticket at a faulty meter, would you be willing to stand there for a while with a stopwatch in order to generate the evidence you would need to fight that in court? And it's in the best interest of the city to have faulty parking meters, because that increases revenue.
I wouldn't be surprised if the same thing were happening with traffic lights.
There are two main differences between parking meters and traffic lights. One is that it is much easier to prove that a parking meter is faulty than to prove that the duration of a yellow light at an intersection is insufficient. The other difference is that if a parking meter is faulty, no one really gets hurt. Whereas if the duration of a yellow light at an intersection is too short, it can cause accidents. Furthermore, the city probably doesn't care if there is a slight increase in accidents, as long as their revenue sharply increases.