Study Says Red Light Cameras Cause Death, Mayhem, Acne

Thursday, March 13th, 2008

Okay, they didn’t go that far. But in a study published this month in the Florida Public Health Review, University of South Florida researchers did find that red light cameras are little more than revenue generators, and actually make intersections less safe than doing nothing at all.

"The rigorous studies clearly show red-light cameras don’t work," said lead author Barbara Langland-Orban, professor and chair of health policy and management at the USF College of Public Health.

"Instead, they increase crashes and injuries as drivers attempt to abruptly stop at camera intersections. If used in Florida, cameras could potentially create even worse outcomes due to the state’s high percent of elderly who are more likely to be injured or killed when a crash occurs."

What else they found:

• The injury rate from red-light running crashes has dropped by a third in less than a decade, indicating red-light running crashes have been continually declining in Florida without the use of cameras.

• Comprehensive studies from North Carolina, Virginia, and Ontario have all reported cameras are significantly associated with increases in crashes, as well as crashes involving injuries. The study by the Virginia Transportation Research Council also found that cameras were linked to increased crash costs.

So what about those studies frequently trotted out by legislators eager to install intersection cameras?

Some studies that conclude cameras reduced crashes or injuries contained major “research design flaws,” such as incomplete data or inadequate analyses, and were conducted by researchers with links to the Insurance Institute for Highway Safety. The IIHS, funded by automobile insurance companies, is the leading advocate for red-light cameras. Insurers can profit from red-light cameras, since their revenues will increase when higher premiums are charged due to the crash and citation increase, the researchers say.

One of those flawed studies credited red light cameras credit for downward trends in intersection injuries that began long before red light cameras were actually installed. Others lumped continuing decreases in injuries at intersections without red light cameras with actual increases in injuries at the considerably fewer intersections with cameras. They’d then come up with conclusions such as, "our intersections are safer since we installed red light cameras," taking care to use words like "since" intsead of "because."

One particularly perverse problem the study didn’t address is the temptation among some city governments to actually shorten yellow lights at camera-monitored intersections to increase revenue, despite well-documented research showing that shortening yellows is pretty much guaranteed to cause more accidents.

Digg it |  reddit | |  Fark

8 Responses to “Study Says Red Light Cameras Cause Death, Mayhem, Acne”

  1. #1 |  Not That David | 

    So, I wonder, did the study control for that? Are camera-fied intersections less safe because of the cameras, or because intersections fitted with cameras are typically also altered in other ways?

    Granted, it probably comes to the same thing in the end, given the temptation to maximize revenue. Still, it seems worth asking.

  2. #2 |  Dave Krueger | 

    This reminds me of city council meetings where they blankly listen to arguments against some proposal on the table only to pass the proposal without discussion immediate after the conclusion of those objections. You get the distinct impression that the decision was made in a back room before the meeting and that actual citizen involvement is nothing more than an annoying formality.

    In other words, red light cameras are here to stay no matter how much evidence city governments are forced to ignore to keep them.

  3. #3 |  Against Stupidity | 

    How many cities would be inclined to install red light cameras, if the fines collected went to the state general fund and any municipal funding received from the state could not be based on the amount of fines collected?

  4. #4 |  central texas | 

    I suppose that the increased number and effectiveness of passive passenger restraint could not possibly have anything to do with that decrease in INJURIES (as opposed to collisions…)?

    One would also suppose that if the intent was to improve safety, the cameras would be installed in demonstrably dangerous intersections. Since I have not had the time to read the study (assuming that the actual text is available rather than the collection of sound bites) I wonder whether the intersections with cameras became more dangerous.

    It would also be interesting to understand what effect concealing the camera might have. If one could not predict which red light was “safe” to run, then perhaps folks would consider the silly notion of stopping for all red lights. A grave imposition on freedom to be sure, but perhaps otherwise desirable.

    And by the by, did anyone else notice that the author who appears most eager to provide sound bites just had a lawsuit dismissed after she rear ended someone? They never get around to actually asking the $64 question: Did she rear end someone who stopped for a red light? At a camera equiped intersection?

    Lots of interesting oxen to be gored here.

  5. #5 |  supercat | 

    A motorist who enters an intersection 0.5 seconds after his light turns red (that’s not even close to making the light) will generally cause far less danger to other motorists than one who tries to stop at a yellow light but either is unable to do so, or is in front of someone else who is unable to do so. Indeed, the only circumstance in which missing a light be 0.5 seconds would cause an accident in the absence of gross negligence by another motorist would be if someone making a late left turn is expecting him to stop. Even there, such assumption would generally be unreasonable on the part of the person making the turn unless the late-entering motorist had given some indication of stopping.

    To be sure, someone who enters an intersection 0.5 seconds late, thereby delaying other motorists, does commit a significant courtesy violation and should probably receive some sort of citation for that. I am unaware, however, of any effort to impose a fine scale that distinguishes safety violations from courtesy violations.

    BTW, if someone enters an intersection 0.5 seconds late, do any of the camera systems delay the opposing green? If not, such failure would demonstrate a lack of real interest in safety.

  6. #6 |  Max D. | 

    “Women, Minorities Hit Hardest.”

    Isn’t there an argument against red light cameras on Constitutional grounds?

    Related: The ongoing war against Gatsos in the UK is civil disobedience at its most entertaining.

  7. #7 |  SayUncle » Revenue err Red light cameras | 

    […] Study says the cause accidents. […]

  8. #8 |  Traffic Cameras are no Boon | Bili Buzz | 

    […] The Agitator Share and […]