Red Light Cameras Raise Crash Risk, Cost

concealment writes with news of dissatisfaction with a pilot program for stoplight-monitoring cameras. The program ran for several years in New Jersey, and according to a new report, the number of car crashes actually increased while the cameras were present. "[The program] appears to be changing drivers’ behavior, state officials said Monday, noting an overall decline in traffic citations and right-angle crashes. The Department of Transportation also said, however, that rear-end crashes have risen by 20 percent and total crashes are up by 0.9 percent at intersections where cameras have operated for at least a year. The agency recommended the program stay in place, calling for 'continued data collection and monitoring' of camera-monitored intersections. The department’s report drew immediate criticism from Assemblyman Declan O’Scanlon, R-Monmouth, who wants the cameras removed. He called the program 'a dismal failure,' saying DOT statistics show the net costs of accidents had climbed by more than $1 million at intersections with cameras." Other cities are considering dumping the monitoring tech as well, citing similar cost and efficacy issues.

Comments

Cost vs injury

[EdZ]

Hard to tell without access to the raw figures, but if the number of T-bone crashes has reduced, replaced by more rear-end incidents, is it possible that the injury rate, or at least number of serous injuries or fatalities, has decreased? Even if the net cost in car damage increases, that would still be a win in my books.

Re:Cost vs injury

[Jah-Wren Ryel]

is it possible that the injury rate, or at least number of serous injuries or fatalities, has decreased?

I'm as anti-traffic-camera as they come, but I have to agree that this is prime territory for "lies, damned lies and statistics." How do we know that the increase in accidents wasn't due to some other factor and the cameras actually mitigated it? What happened in the bordering towns that didn't deploy cameras? Where there any other changes, like increased number of drivers on the road, etc?

Re:Cost vs injury

[IndustrialComplex]

T-bone crashes aren't the type of crashes prevented by red-light cameras to any significant degree.

A T-bone crash can only occur a few seconds into the light when the other lane of traffic has already begun to pass through the intersection. No driver WANTs to run a true red-light. Those times you see someone blow through a red-light that wasn't just someone squeaking through or missing a yellow? Those mid-red light runners completely missed that there was a redlight there at all! They didn't run the red-light because they wanted to, they ran the red-light because they weren't paying attention.

So what effect does a red-light camera have on people who aren't paying enough attention to see that there is a red-light there in the first place? Well, as we can see by these numbers, not much of an effect at all.

I'd be willing to be that there are fewer T-bone style crashes, because of an increased number of people stopped at the intersection, providing something else for the inattentive driver to see (or as the numbers suggest, rear-end)

Re:Cost vs injury

[artor3]

According to this article, which references the same study, the cameras actually seem to make things much better.

On the two intersections where the cameras have been in place the longest, t-bone accidents are down 86%!! And this isn't a revenue stream either, as the number of tickets issued, while spiking initially, drops off substantially as drivers get used to the cameras.

Of course, the article submitter wanted a good ole fashioned anti-government hate-fest, and was happy to massage the numbers accordingly. Lies, damned lies, and statistics, indeed.

Re:Cost vs injury

[dgatwood]

No, what is needed are not cameras. What is needed are some combination of:

• Stripes on the pavement to indicate that at the speed limit, you should not stop if you are past this point. This reduces the guesswork that currently makes it difficult to assess whether to enter a light or not.
• Countdown timers on lights to indicate how long the driver has before the light changes to yellow. Again, this reduces the guesswork.
• Longer yellow cycles. Studies consistently show that above a few seconds, drivers do not ever adjust to longer yellow cycles. Thus, lengthening the yellow cycle by only a couple of seconds reliably and reproducibly reduces red light violations to near zero.

Any one of these would result in a far, far greater reduction in traffic light violations and the resulting collisions than any camera system possibly could. The combination of all three would almost (if not completely) eliminate the problem entirely.

The whole rear-end collisions thing is bunk with respect to the cameras. The at-fault is always the driver who is following too closely to the vehicle in front of them so they can't stop in time for the braking vehicle--not the stationary red light camera.

Technically, yes, but the fact of the matter is that increasing the probability of a driver slamming on his or her brakes increases the probability of a rear-end collision, which is a simply inexcusable thing for law enforcement to be doing, given that there are so many better ways of solving the problem in question that do not result in such a negative side effect.

Therefore, given that red light cameras are significantly less effective than alternative techniques, the only real reason to consider them is revenue generation. And if that's the government's only purpose for enforcing traffic laws in a given community, its leaders should resign.

Re:Cost vs injury

[dgatwood]

For those folks who don't want to take the time to read the article, briefly put, the number of T-bone crashes decreased, but the average severity of those crashes increased, presumably because of people racing to try to get through the light so that they wouldn't get ticketed, failing to do so, and thus being at an unsafe speed at the time of collision.

Thus, even the costs associated with T-bone crashes increased, and although this study did not break down the cost into medical and non-medical costs, one can reasonably assume that although the injury rate decreased because the total number of accidents decreased, the severity of injuries was probably greater, resulting in increased medical costs. That's just not a very good trade.

Red light cameras do not merely encourage positive behavior. They equally encourage both positive behavior and far more reckless behavior, and on average, make things worse according to nearly every metric. Increasing yellow light times (by starting the yellow cycle a second earlier) make things better by nearly every metric. Why don't we do this? Greed. Write your legislators and demand a ban on red light cameras and a return to more reasonable yellow light times.

Re:Cost vs injury

[hduff]

So what effect does a red-light camera have ...

It increases government revenue by ticketing mostly people who make rolling-right turns on red, a practice that does not result in traffic fatalities for the most part.

Re:Cost vs injury

[icebike]

Is it still a win in your books when the cities shorten the yellow to generate more tickets?

You've probably hit on the principal reason for the mixed results. The systems were designed to serve one purpose, but the money was just too good to pass up. The system was perverted to serve a different purpose.

People are so worried about entering an intersection on red that they are causing rear-end accidents by sudden stopping when the safest thing to do is just to roll across the intersection. Most intersections have an all-ways-red interval to handle the guy who enters on what he thought would be a late yellow, but actually was red due to a shortened yellow.

If cameras were not allowed to trigger until the crossing lane's lights were GREEN, and there were statutory yellow durations and statutory all-ways-red durations, it would eliminate all this yellow shortening nonsense, and maybe the cameras would catch the scoff-laws they were intended to catch.

Re:Cost vs injury

[TheGratefulNet]

There are a lot of folks out there who think stopping at lights is optional.

my friend had this strange habit; when we approached an intersection he would speed up and plow thru the red lights. I would ask him why and he would say 'relax, my brother does this all the time'. every time, the same thing: running past a red light and telling me not to worry. it all worked up until we came to a green light and, yes, he slams on his brakes. when I ask why he does *that*, his reply was 'look, my brother could be on that road!'

Re:Cost vs injury

[Dahamma]

Actually, it's damn lies in the slashdot summary, too. From TFA:

"The DOT’s report noted that two intersections in Newark have been part of the camera program for two years, and that 24 others in six communities have been recording violations for at least one year. At the Newark sites, the report said, crashes in the latest year were down by 57 percent from the “pre-camera year,” with decreases of 86 percent for right-angle collisions and 42 percent for same-direction crashes.

It said the number of citations issued at the Newark intersections fell by 85 percent over the two-year period. “

So, crashes are up a total of 0.9% in a year, but the intersections that have had cameras for 2 years are *way* down, as are citations. So, it's also very possible (though still a hypothesis) that it takes a year or so for locals to get used to the concept of cameras, at which point they really start proving their worth.

And even in the "new" cameras, an increase of only 0.9% means while there was a slight increase overall, it's practically a given that a collision from behind because someone couldn't stop fast enough in reaction to hard braking is going to be less dangerous that T-boning someone trying to beat the light...

Re:Cost vs injury

[dasunt]

No, what is needed are not cameras. What is needed are some combination of:

One thing you missed: Better traffic lights.

Traffic lights do contribute to delays of traffic flow. They slow down traffic, and the contribute to useless delays, as anyone who has sat at a long red light, waiting for cross traffic that doesn't exist.

I'd argue that any large country (I'm looking at you, US) has an incentive to improve traffic light. Right now, the most common intelligent traffic light has ground loops to detect when cars are at the intersection. In this day and age, couldn't we do better?

Imagine a traffic light that could "see" traffic approaching from a distance. It could also see traffic backed up to the light. Instead of turning red when a car or two is approaching and the cross traffic is non-existent, it could remain green. Instead of turning green when traffic is backed up to the light, it could let other traffic, which is clear to continue, go.

Just imagine the time and fuel savings!

And finally, I'd have to say, in this day and age, sometimes intelligence is overrated (even though this negates the above). As a young adult, the traffic lights in my town sensibly shifted to blinking yellow and red at late hours when the traffic was light. The most common through traffic would get blinking yellows. The underused cross streets would get a blinking red. Then, years later, they replaced the traffic lights with a more modern type that would detect when a car was waiting and change the light. This was useless in a small town - approaching from a cross street would result in a longer delay (as the other light would go through a yellow/red change), and if you were on the main street when a car had to cross, you'd have to wait for the red/green change for just one vehicle.

Really, considering how much time Americans spend on the road, and how much time is lost, small improvements result in a big gain in time not lost and gas not burned.

Re:Cost vs injury

[NIK282000]

That sounds like logic! I wont have any logic when it comes to traffic control and regulation, its obscene!

Re:Cost vs injury

[Concerned Onlooker]

"People are so worried about entering an intersection on red that they are causing rear-end accidents by sudden stopping..."

That's a basic misunderstanding of safety. The people tailgating are the ones causing the accidents.

Re:Cost vs injury

[anagama]

That's the issue right there: revenue.

I live in a border county with Canada. They've started installing signs about a hundred yards ... well, meters ... in front of stoplights on major streets. The signs have two yellow flashing lights that flash and a message that says "be prepared to stop when flashing" -- in other words, the warning sign starts flashing and by the time you reach the intersection, the stop light has gone red even though way back when you noticed the warning sign, the stop light still showed green. So you know that even if the stop light is green, if the warning sign is flashing, it'll be red by the time you get there so slowing down is a good idea.

I suppose it will be some time before we know if the warning signs cause people to gun it for the intersection, or if it works as intended. I hope it is the latter because it is 1) unintrusive (no data collection), 2) directly aimed at the issue rather than a backdoor means for private profit at public expense, and 3) pretty darn simple (pole, board, paint, light bulb, timer).

Re:Cost vs injury

[thegarbz]

And yet the person behind didn't leave enough space. I suppose it's still the person in front's fault if a kid runs out on to the street?

I'm not sure where you live but here the person behind is ALWAYS at fault if he rear-ends a car. Driving down the road, nothing to be seen, step on the break as hard as possible putting faith of god in my ABS system and the guy behind me hits me? He's at fault. Guy behind doesn't have ABS and I do? He's at fault. Snow on the road? He's at fault.

The fundamental reason behind this is that the guy behind was not leaving enough room to react to a very sudden change in the conditions ahead, aka tailgating. Also tailgating doesn't mean you're riding the guy's bumper, no it means different things at different speeds. In Germany at 100km/h if you're within 50km of the car in front you're tailgating. (Speed in km/h)/2 = (distance in m) in space you need between the car in front of you.

Not that it matters because if you weren't tailgating and you still manage to hit the car in front you're likely asleep at the wheel and guess what, still your fault.

People should be following for the worst case, not thinking about which intersection is safe.

Re:Cost vs injury

[Xest]

Yes, it's sounds like the real problem here is that these areas of the US are suffering from a major tailgating problem above all else.

All a red light camera does is catches people who run a red light. I was always taught that when approaching traffic lights, approach them as if they're going to change and that's exactly what they do. I've never had a problem stopping for a red light, not once and neither have I had anyone go into the back of me at traffic lights.

From the comments on this thread it sounds like people in the US race up to lights and then slam their brakes on and everyone behind them tailgates as that's the only reason red light cameras would increase crash risk and cost. If that's the case then it sounds like there's a more fundamental issue here - that driving test standards in the US simply aren't high enough, and policing of bad driving isn't done well enough.

Maybe the US does need more raised roundabouts, you have little choice but to slow down coming up to them, otherwise you crash right into them and have to explain to the cops how the fuck you managed to land your car in the middle of a roundabout. That way people who drive like idiots automatically get caught because with a bit of shrubbery on the roundabout too it makes it hard for them to get away from the scene, assuming their car is even in good enough state to do so still. They soon learn.

related to yellow-light shortening?

[Trepidity]

There have been a number of scandals, including in New Jersey, where installation of cameras was found to coincide with, or be followed shortly thereafter by, shortening the yellow-light duration, presumably to make more money from the resulting tickets.

This article implies that the cameras themselves are at fault, but I wonder if the shortened yellow-light duration is actually the primary culprit.

Re:Honestly...

[qeveren]

What honest excuse do you have for running a red light? It isn't like you don't get plenty of warning that the thing is going to change.

You want lawsuits ...

[Taco Cowboy]

...regarding ...

* 1. exploitation of homeless/unemployed people and putting them in harm's way ?

* 2. drivers getting into accident due to the distraction by the appearance of homeless/unemployed/unkempt people at busy traffic intersections?

* 3. little girls in cars got spooked by homeless/unemployed/unkempt people taking pictures of them?

 
If your answer is "YES" to all of the above questions, then, sure, go ahead, start distributing digital cameras to homeless/unemployed people and putting them in the middle of busy traffic intersections.
 

Re:You want lawsuits ...

[Spy Handler]

1. It's not exploitation if the homeless/unemployed person does this voluntarily for a wage (aka a job).

They wouldn't be in harm's way if they're positioned somewhere sensible where cars don't travel, such as a sidewalk.

2. Drivers get into accidents for all kinds of stupid reasons. Cops sometimes appear at busy intersections, as well as hookers, protesters, news cameramen, fruit vendors, etc. I don't see any outrage to ban these.

3. Little girls in cars, oh my... won't someone please think of the CHILDREN!

Re:You want lawsuits ...

[homey of my owney]

The City of Albuquerque recently cancelled its Red Light camera program after a vote resoundingly said to get rid of them.

Re:Why not

[Anonymous Coward]

This actually did happen in Korea several years ago:

Float back several years in time with me for a moment. There was once a time in Korea when the government got really serious about curbing traffic violations. This was probably due to ranking highly on all sorts of international statistic lists for traffic fatalities.

Anyhow, the Korean government’s solution was to implement a reward system whereby normal citizens would receive a monetary reward for submitting photos of other drivers violating the law. Brilliant idea right? Yeah, and it failed brilliantly too. Wanna guess why? The Korean government failed to take the following into account:

1. The number of false reports and staged photos was absolutely through the roof.

2. People began CAUSING traffic violations in order to profit. For example, they’d block a street momentarily so that an intersection would get filled up with cars during a red light, and then they’d have a friend photograph all the cars stuck in the intersection.

3. People began blackmailing each other. Instead sending the photos into the police, they started trying to sell the photos to the drivers of the cars being photographed while breaking the law, and it turned out to be even MORE profitable.

4. Korean people began quitting their jobs, buying expensive camera gear, and setting up elaborate photograph traps in areas where they knew they could make money. That’s right, people actually quit their day jobs because blackmailing or turning in their fellow citizens all of the sudden became more profitable than working in an office.

5. The government didn’t consider that they would receive hundreds of thousands of photographs, and without some type of standard or rules set in place, would be obligated to pay out insane amounts of money to the thousands of amateur photographers who suddenly materialized across the peninsula. The profits generated by traffic fines went to pay off the photographers, which means no profit for the government.

6. Traffic violators would see another person photographing them, and then they’d get out of the car and beat the shit out of the cameraman.

7. Men would take pictures of women violating traffic laws, and then demand sexual favors in exchange for not submitting the photos to the police.

Thus the “turn in your poorly driving neighbor” policy was scrapped almost as quickly as it started. And no, this isn’t fiction. Ask a Korean about it.

Simple way to improve intersection safety

[fiannaFailMan]

Roundabouts.

They also improve traffic flow and eliminate the need for 4-lanes in each direction to store stopped idling cars.

Re:Simple way to improve intersection safety

[ionix5891]

Oh god no! In ireland all we have are fucking roundabouts, and are now being replaced by traffic junctions and traffic flows much better at rush hour

Re:Simple way to improve intersection safety

[alcourt]

They are also correlated (according to the state transportation official I talked to who was pushing one in my area) with higher car on pedestrian injuries, and are more likely for new drivers to have loss of control accidents compared to more traditional intersections. These loss of control accidents often end up with the vehicle striking the very areas pedestrians are expected to stand, waiting for minutes for a break in traffic to safely cross.

Re:Simple way to improve intersection safety

[Anonymous Coward]

Roundabouts are a bit like transistors. When a heavy flow starts in one direction, the side directions are blocked indefinitely until the flow lightens, but both directions of flow stop or slow down before entering because they are unsure as to the behavior of the car in the circle (is it going to exit? continue around?). Thus they act as bottlenecks.

Red light cameras increase both crashes and safety

[Ichijo]

The Department of Transportation also said, however, that rear-end crashes have risen by 20 percent and total crashes are up by 0.9 percent at intersections where cameras have operated for at least a year.

That isn't surprising. The Federal Highway Administration found that red-light cameras increase rear-end collisions but reduce more severe right-angle collisions, saving $50,000 in collisions per intersection per year in medical and repair costs.

Assemblyman Declan O'Scanlon, R-Monmouth..wants the cameras removed.

That would be an extremely bad idea. "In the immediately aftermath of the [red light camera] law's expiration, the risk of someone running a red light at an intersection was three times higher than it had been when cameras were on."

If safety is the goal, they should keep the red light cameras and lengthen the yellow light duration. "An Institute study conducted in Philadelphia, Pennsylvania, evaluated effects on red light running of first lengthening yellow signal timing by about a second and then introducing red light cameras. While the longer yellow reduced red light violations by 36 percent, adding camera enforcement further cut red light running by another 96 percent."

Re:So...

[norpy]

If you rear end someone you are by definition following too close and the accident is your fault. That means that if you get rear ended the person you should be suing is the one that dented your ride.
At least that is the law where I live (not in the US)

Scrap signs altogether!

[mcrbids]

There's a movement growing rapidly in Europe to reduce traffic signs and lights, and they are finding that removing signs and lights can cause a rather dramtic reduction in accidents. A number of cities have done away with traffic lights and signs entirely with surprisingly good results. (EG: average trip times drop dramatically, accident rate plummets, people report greater satisfaction, etc)

I'm not saying that we should do away with all signs everywhere, but there is sufficient evidence available that the "common sense" utility of the traffic sign or a traffic light is clearly unproven.

Re:Red light cameras increase both crashes and saf

[alen]

Maybe people need to learn to gently stop and not on a dime like a race car driver?

Re:Why not

[D'Sphitz]

You must have skipped over #1 and #2.

Re:Light Counter?

[Anubis IV]

This. I've been wondering the same thing since they introduced LEDs.

One change I'd suggest is to make it some sort of a progress indicator, rather than an actual countdown using numbers, that way it's more understandable to everyone and more discernible from a distance. For instance, light the LEDs in a filled circle at the center of the yellow light, that way people can always tell it's a yellow, but then have the unlit outer ring of LEDs get filled up in a sweeping motion that goes around clockwise, that way we can tell both how quickly it is going and how much time we have left, without having to making out the number or parse it. Doing that would make it easier to see at a distance, and it'd also mean that even if you missed the first half of the yellow, you could still tell how fast it's going, giving you an idea of whether it's a fast or a long yellow light for future reference.

I'm not sure that I'd do it with greens, however, since I know plenty of intersections that stay green for several minutes at a time (making both of our ideas a poor choice), and I'm pretty sure I've even seen some at night that default to staying green until they detect a car at the cross street, meaning that a countdown or progress indicator would be irrelevant.

Re:Honestly...

[dgatwood]

The problem you've got in New Jersey is not red light cameras. It's poorly trained, inconsiderate and selfish drivers.

Don't forget pissed off drivers. When you've just hit the last eight lights in a row red, most people find themselves a lot more motivated to try to make it through one, because doing so means that they'll probably make it through the following one or two lights on green. There's only so much that most people can handle, and once people cross that threshold, their driving becomes markedly worse. Unfortunately, our nation's traffic light management sucks harder than a Hoover, so this is a much more common problem than it should rightfully be in any sane universe.

Re:T-bone vs read-end

[IndustrialComplex]

So the real problem is people approaching the intersection too quickly. Before cameras, they just went right through, causing right-angle collisions. Now they scared of the fine and breaking quickly, causing read-end collisions.

I'm not comfortable with having 1/10th of a second to decide if I need to decelerate my vehicle from 45MPH to zero or if the 0.25s it takes me to approach the intersection is too long for me to proceed through the arbitrary length of the intersection before the unknown duration yellow light changes to red and nails me with a $200-$400 ticket.

Now, my only options when I see a yellow light are:
1. Too close to stop safely: ACCELERATE to avoid taking too long
2. Unsure of yellow duration: STOP NOW ASAP don't want to risk going slightly past that white line.

Before I had the option:
1. Too close to stop safely: Decelerate and proceed cautiously. If it turns red, that's ok because I have enough room, and I'm taking my time to make sure no pedestrians are crossing against the light. I'll also be out of the intersection before opposing traffic starts so I'm not inconveniencing them.

2. Unsure of yellow duration: Stop the car, using care not to decelerate too quickly. I may stop just slightly beyond the white line, but I'm not in the plane of traffic, so no one is impeded. The car behind me could also decelerate safely.

If I made a mistake on 1 or 2 in the non-traffic camera situation, there is always the potential for a police officer to pull me over and let me know it was too far out of bounds. Strangely enough though, that has NEVER happened. Why, it's almost as if I'm a cautious driver, without a single citation for running a redlight or a speeding camera despite living in DC, yet the cameras have modified my behavior to be much less safe (for you), but much safer for my wallet.

Well-established

[Curunir_wolf]

This seems like pretty old news at this point. Red-light cameras are put in place by private companies promising revenue. It was never about safety, and study after study has shown increased hazards at intersections where they are installed.

As usual in these cases, people need to remember to follow the money. One person you've never heard of, but should be thanked for exposing this issue, is Shawn Dow of Arizona. He has been all over the country teaching activists how to fight these things and make local legislators afraid of the people, instead of kow-towing to the rich lobbyists. He's been beating up on politicians (figuratively) for years, and winning.

Re:Why not

[MickLinux]

Actually, I prefer web cams placed all along the roads, each reachable from a city website. line the roads with stripes, so speed can be digitally measured, and keep the last day of footage for public review.

Then, when someone gets cut off, or sees reckless driving, they can post a pointer to the police. Police either confirm or deny claim -- as well as any OTHER infractions they see in the incident. Appropriate citations get mailed out [fine to license only, no points]. Citizen who pointed out the incident first gets 10%. Citizens whose claim is denied cannot register another claim for a week, by automatic lockout.

Now, the unemployed can freelance in traffic patrols, and the wealthy who flaunt the law can fund them. The poor who flaunt the law can stop driving. The reckless can learn a cheaper way to drive.

I think it'd be a win-win situation .