Continued from part one of the article. View full article on a single page

Regressing Behavior
It turns out the financial success of red light cameras depends on short yellow light durations. According to a report by the California State Auditor, more than three out of every four red-light camera tickets were given for violations that happened when the light had been red for less than one second. What's more, the vast majority of the tickets were the results of infractions that happened within the blink of an eye, literally less than 0.4 seconds.

Union City's camera vendor, Redflex, despite knowing the importance of yellow light durations to their bottom line, assured the city police that the reduction in photographed violations was only temporary. Bizieff reported that Redflex told him that the city, "should expect a rebound in following months with citation numbers gradually approaching, but falling short of, the original numbers."

Did this happen? Nope. A year later, Union City red light violations were still down -- by 73 percent in the period since the yellow light timings were properly set. Indeed, longer yellow timings make our roads safer. This is backed up by the Texas Transportation Institute, which concluded that an additional second of yellow light time can cut not just violations, but also accidents by 40 percent.

Such an encouraging conclusion runs counter to what you'll read in studies (usually funded by the insurance industry) that celebrate red-light cameras -- and not longer yellows -- as boons to public safety. What these biased studies don't explain is that a statistical phenomenon known as "regression to the mean" often explains decreasing accident rates.

Think about it this way. When a jurisdiction decides to install traffic safety cameras, the first practical question is, Where should they be installed? The answer is obvious: in locations that have had a lot of accidents. Let's take the hypothetical example of a certain road that normally has about two fatal accidents a year. One day, a van full of people loses control and rolls down a cliff, killing six people instantly. That road's fatality tally jumps to eight for that year, making it a prime candidate for camera installations. So speed cameras are installed.

The next year, without any fluke van accidents, fatalities settle back down to two per annum. This return to an average number is known as the regression to the mean. You could have installed potted plants to monitor the road, and produced the same result -- but this doesn't stop public officials from holding a photo op to celebrate the cameras as the cause of success. Make no mistake, regressions to the mean are recognized by traffic engineers; they're just not publicized. We give you the case of an official U.K. Department for Transport report, which attributed a 54.5 percent reduction in accidents to speed cameras. Sadly, the fact that regression to the mean accounted for 44.1 percent of this "success" wasn't mentioned until the appendix way at the end of the document.

Studying the U.K's accident rates a bit further is useful, because one can barely drive a few miles without seeing a speed camera on the island. Indeed, the U.K. is now home to more than 6,000 cameras, and about two million drivers are ticketed every year for driving just a few miles per hour over the limit.

In 1991, the year before speed camera citations were first issued, there was a total of 311,368 injury accidents in the U.K. One decade and 3,561,817 automated speeding citations later, injury accidents had not declined at all -- the 2001 total stood at 313,309, despite significant advances in trauma care, road engineering, and vehicle safety equipment, such as airbags and ABS.

After 2001, U.K. government statistics did begin showing a decline in the number of serious accidents -- but were the stats valid? After all, the British Medical Journal found that the road injury rate increased slightly from 90.0 per 100,000 population in 1996 to 91.1 in 2004, and this conclusion was drawn from the review of actual hospital records. In explaining this discrepancy, an official government watchdog group, the Statistics Commission, said last year, "There is known undercounting of road accidents in police statistics which are used to... inform policies on traffic safety."

In other words, in the profit-driven world of camera enforcement, reports are being manipulated to make speed cameras appear more effective than they really are.

It gets worse when we turn our attention to red light cameras, since these devices can actually introduce hazards of their own. When a motorist knows a camera is watching, he's wont to slam on his brakes at a yellow light rather than risk a ticket. In these cases, the camera does indeed keep the driver from running the red. But if the motorist behind doesn't know a camera is watching and is caught off-guard by the lead car's sudden braking, he'll rear-end the car that stopped early. Indeed, studies prove that red-light cameras lead to more rear-end collisions -- in some cities up to 139 percent more, according to a 2007 study sponsored by the Virginia Department of Transportation.

Though this effect is well-documented, it's often dismissed by camera supporters who suggest that rear-enders are merely "fender benders," and are acceptable alternatives to the "deadly" T-bones that cameras prevent. Sounds plausible enough, right? After all, when you're struck from behind, you have a trunk and large crumple zone to protect you. Conversely, just thin sheet metal and a pane of glass offer protection during a side impact.

Right?

Well, it turns out that rear-end collisions can be plenty injurious. We'll give you the case of Winnipeg, Canada, whose city auditor obtained data on every intersection accident that occurred in town between 2002 and 2004. The data was collected by Manitoba Public Insurance, which, as the province's only insurance company, had the sole burden of paying the hospital and repair bills. Now, keep in mind, this insurance carrier was providing numbers on actual costs, not statistical estimates, which are often quoted in studies of this type.

The Winnipeg auditor found that the number of injuries at intersections with red light cameras jumped 64 percent. The cost of property damage claims increased up to 113 percent -- all despite a significant drop in the number of T-bone collisions. With devastating results like this, it's difficult to see how rear-enders can be considered "low-impact" accidents.

Several U.S. studies have reached similar conclusions. The Washington Post discovered in October 2005 that injury accidents increased 81 percent (and total accidents doubled) at D.C. intersections equipped with red light cameras. In the same Virginia DOT study that reports red-light cameras cause more rear-end collisions, we learn that while T-bone accidents decreased slightly, the overall number of injury accidents increased -- by as much as 24 percent.

Follow the Money

So if safety isn't the primary motivation behind camera enforcement, what is?

To find the answer, in 2006 a reporter for the U.K.'s Daily Mail posed as an Eastern European looking to buy a speed system. A few cocktails at a posh restaurant brought out the truth from Jon Bond, the former Warwickshire Police Chief Superintendent and current CEO of Tele-Traffic UK Ltd., the number one U.K. vendor of laser speed cameras.
A hidden camera captured a speed camera executive and former police chief admitting that their photo enforcement tools were part of a "scam."

The systems make "buckets of cash," Bond told the reporter. "There will be so much money coming in, you won't know what to do with it." Ironically, these damning words were captured on a hidden television camera.

Profit motives also drive the issue in the United States. In 2001, I spoke with Ron Arnone, a former employee of Lockheed Martin IMS -- then the red light camera contractor for San Diego, California. Arnone served as a leading witness in a criminal traffic court trial in which hundreds of defendants joined together to fight the machines that had accused them of running red lights.

Arnone told me (as well as testified) that money was the only topic photo enforcement employees ever talked about behind the scenes. He said safety was the last thing on their minds, a fact confirmed by confidential memos obtained during the trial.

Judge Ronald L. Styn absorbed this information and ruled, "The evidence obtained from the red light camera system as presently operated appears so untrustworthy and unreliable that it lacks foundation and should not be admitted."

The fact is, despite all the legal, ethical, fairness, and safety problems they introduce, traffic cameras are just too attractive a revenue source for local and state governments to turn down. Even San Diego couldn't resist re-installing cameras a few years after Styn's rebuke
.
There's still hope, as the camera issue is blowing up in court rooms across the nation. The Minneapolis camera program that bedeviled Don Zimmerman shut down in September 2006 after a state court of appeals found the program violated state law, a decision upheld this year by the Minnesota Supreme Court.

The Ohio Supreme Court is currently hearing a challenge to speed cameras, and early this year, court rulings forced San Jose, CA to drop the speed camera program it had operated since 1996. San Jose was using speed cameras despite a specific law the California legislature wrote in 2000 stating that it, "does not authorize the use of photo radar for speed enforcement purposes by any jurisdiction."

You would think this wording is clear enough, but it's not. In fact, just this March, the California Mountains Recreation and Conservation Authority -- a governmental group with jurisdiction that stretches from Santa Monica to Simi Valley -- signed a contract to set up speed cameras just outside the famous Hollywood Bowl.

With local governments blatantly flouting legislation, it's clear that drivers must get directly involved to protect their rights. Indeed, in the four times that photo ticketing has been the subject of a referendum, the people loudly and clearly said "no" each time. Last year, for example, 76 percent of voters in Steubenville, OH rejected speed cameras. Cameras have also been rejected in Alaska, Arkansas, Hawaii, Nebraska, Nevada, New Jersey, Utah, West Virginia and Wisconsin. If traffic cameras are on the horizon in your area, contact your representatives and demand the public have a say in the matter.

If it's too late and you've received a photo ticket, fight it. Californians, in particular, should take advantage of a free website, www.highwayrobbery.net, to learn more about how to beat automated tickets. And all drivers would do well to heed Don Zimmerman's advice: "Don't ever pay a ticket until you go online, and are sure you're guilty."

First published in the Summer 2007 edition of High Performance Driving magazine.