The Denver Post has an interesting piece today on a topic I blogged about several days back, indicating that the jury is still out on whether Robocop-style policing actually improves public safety (Colorado Springs' looming experiment with red light cameras gets a brief mention). Although automated ticket-writing is a proven way to fatten municipal coffers, at low- or no-cost to the city, conclusive evidence is hard to come by about whether the cameras prevent accidents and deter scofflaws, according to the Post:

"The Federal Highway Administration reviewed all studies of red-light cameras in 2005 and offered a lukewarm endorsement. While avoiding estimates of injury reductions, the agency said sites with cameras produced an average of $39,000 in annual economic benefits from crashes avoided. But, the agency added, wide variances in methodology and reliability "render useless any conclusions" from most of the studies.

Researchers in Oxnard, Calif., in the late 1990s compared results at that city's red-light cameras with three other cities. They found that overall accidents in Oxnard fell 5.4 percent; yet crashes in the control city of Santa Barbara, without red-light cameras, dropped more than 10 percent.

A Texas study showed a 43 percent reduction in broadside collisions and a 5 percent increase in rear-end accidents — but noted that the cameras' role in those numbers was inconclusive."

Aurora's experience with red light cameras has been equally inconclusive, even though the city is expanding their use, according to the Post:

"Aurora is negotiating a contract to add cameras at 10 intersections to the four cameras operating since 2006. Police say they are convinced the cameras improve safety, but the statistics are less clear.

Intersections with two of the older cameras showed a decline in accidents, from 17 to 11. One intersection had no accidents, before or after the camera was put in, while a fourth intersection actually saw an increase, from zero accidents to five.

"What we're finding is that the numbers are not real, real significant in seeing a decrease in accidents, because there were not a huge number of accidents to start with," said Cloyd, of the Aurora police.

Police officials often point to reduced violations at red-light cameras, separate from accidents, as proof the devices educate drivers and change behavior. Yet Aurora's cameras issued 9,918 tickets in 2007, 8,717 in 2008 and 9,855 in 2009."

Critics of the cameras claim that equivalent intersection safety improvements can be gained, and red light-running can be curtailed, by simply fine-tuning the signal light sequence -- by prolonging the yellow light for just a few seconds, for instance. I've not studied the issue enough to reach my own conclusions. But one thing seems clear enough to me; that something other than hard data is driving the expansion of Robocop technology.

Traffic enforcement should be aimed at improving roadway safety, not generating revenue or creating automated speed traps. One sure way to undermine public support for law enforcement, and erode respect for the law, is to create the impression that police officers are just revenue collectors in blue uniforms. And I don't want to see that happen in Colorado Springs.