January 06, 2007
Debating red light cameras

Last month, I posted some links from a traffic safety expert named Reed Berry, with whom I'd corresponded after he left a comment on this post about red light cameras. Berry mentioned that he had been involved in a radio debate with Michael Kubosh, and I asked about a recording of it. I have that recording now, which you can get as an MP3 file here. It's at the beginning of the show and lasts about 35 minutes total. A couple of observations:

1. I was not impressed at all by Kubosh, not that I expected to be. He kept insisting that the cameras are nothing but a money grab, but never really cited any evidence to back that up; Berry brought up the studies that show a reduction in fatalities and injuries, but Kubosh just denied them. A question I wish someone would ask Kubosh, since he also repeatedly stated that traffic light laws should be enforced by the police, is whether he'd categorize a concerted effort by HPD to crack down on red light runners (via stings at high-volume intersections) as a revenue enhancement scheme as well. The impression I get from him is that he thinks it should be a matter of bad luck to get nailed for running a light.

The silliest thing Kubosh said was that the selection of the intersections by Houston was essentially random - I don't recall his exact words, but he said something like "there was no study done" in determining where the cameras would go. Well, it seems to me that if the city's motive is raking in the fines, then they'd be strongly incentivized to put the cameras where they'd nab the most offenders - in other words, they'd do a study to figure out which intersections have the worst problems. (Me, I'd say they made some wise choices, but what do I know?) As such, he's saying the city wants to shake us down, but they're too stupid to do it properly.

2. The one study Kubosh did cite was a Texas Transportation Institute study that claimed adding one second to the yellow light time at an intersection cuts down red light running by 40%. I don't trust him enough to accept that without seeing the study myself, but it's not relevant anyway. I support optimizing yellow light times, but doing that doesn't have any bearing on whether red light cameras would cut down on accidents or not. He also mentioned displaying the time till the light changes at intersections, something I've seen in places like Washington, DC, which again I'd support but again doesn't have any bearing on the debate. Maybe these two things together would have enough impact on red light running that the cameras would be superfluous (Kubosh never explicitly made that claim, let alone cited any evidence for it), but I seem to recall that DC has cameras as well. The point I'm making is that this isn't an either-or choice - you could very easily have optimized yellow lights, time-till-change displays, and red light cameras. One might argue that this is the real way to go to have the biggest impact, in fact.

3. I was more impressed by Berry than I was by Kubosh, but that's not saying terribly much. It's a shame that Berry knew nothing about the specifics of Houston's cameras - he might have challenged Kubosh regarding the camera placement, for one, and he might have noted that the fine in Houston is $75, not $400 or $500 as it apparently is in some places in California (Kubosh, who flogged the revenue horse throughout, had no reason to point out the lower fine that he faces for his deliberate infraction). The issue of you getting a ticket for someone else running a light with your car came up, and Berry suggested pushing to change the law so that the cameras record the drivers' faces. Needless to say, that doesn't help assuage my main issue with the cameras, which is privacy and usage of and access to the cameras' data.

4. Similarly, the host brought up a reasonable slippery slope argument, by asking Berry "why don't they just install a device in every car that would automatically generate a ticket for you if it recorded that you were speeding?" Berry dodged this question by saying that he himself would never get such a ticket because he didn't speed (he said the same thing about his never running red lights). The privacy issue simply never came up, which was a disappointment to me.

5. The radio show on which all this took place was the Rusty Humphries show. I'd never heard of the guy, but from the intros and outros it was clear he was your standard aggrieved and put upon right winger; the theme song and show promos were hilariously egotistical. Thankfully, there was a guest host, who was even-keeled and non-confrontational. I'd have liked him to challenge both Kubosh and Berry more, but I'll give him props for letting everyone have their say and not favoring one side over the other, even in selection of callers. The show was done on December 22, so some of the intro/outro clips were in full War On Christmas mode. Amusingly, every time he came back on the air, the guest host said that Rusty was "spending the holidays with his family"; which holiday was never specified. Apparently, the anti-Christmas forces are even more insidious than even Rusty Humphries could imagine.

Posted by Charles Kuffner on January 06, 2007 to Planes, Trains, and Automobiles

I've seen the TTI study, Charles, and the 40% reduction by increasing yellow light times is correct. But I'm afraid I disagree with this comment: "I support optimizing yellow light times, but doing that doesn't have any bearing on whether red light cameras would cut down on accidents or not."

Actually, there is a limit to how much red light running can be reduced - you can try to get that reduction by massive fine schemes or non-coercive means like Kubosh described. But because for the companies this IS all about revenue (whether it is for the city, you know that's all the company cares about), it won't recoup their investment if you added a second to yellow lights and implemented countdown timers. Nobody does both - indeed, some towns have actually been caught REDUCING yellow light lengths to maximize revenue. The vendors are key players in all this - the real beneficiaries of the scheme.

It's also worth mentioning, if Kubosh did not, that red light cameras actually increase injury accidents. The other solutions don't have that drawback.

Posted by: Gritsforbreakfast on January 6, 2007 8:52 AM

Kubosh did make that claim, but as noted in my previous post, there's studies showing that cameras do decrease fatalities and injuries by decreasing T-bone accidents more than rear-enders increase. As I say, Kubosh never cited anything to this point, he just said "no it isn't".

Posted by: Charles Kuffner on January 6, 2007 10:29 AM

Many of the studies claiming reduced injuries were performed or paid for by vendors, so be sure to check sources closely in this debate.

From what I've seen, studies performed by states or foreign governments, as in the link to the Virginia study I cited above, typically find "a 'definite' increase in rear-end accidents and only a 'possible"' decrease in angle accidents. Most importantly, the net effect was that more injuries happened after cameras are installed."

Even if Kubosh isn't the most articulate or credible spokesman, red light cameras are still a bad idea.

Posted by: Gritsforbreakfast on January 7, 2007 9:08 AM

standard aggrieved and put upon right winger

Nice blanket characterization there. Reasoned, not bombastic. :)

Posted by: kevin whited on January 7, 2007 9:50 AM

Oh, Kevin. You're so precious, I just don't know how you make it through the day without someone pinching your cheeks all the time. Thank God we have you here to remind everybody to be nice to one another.

Scott, just so we're clear, my mind is not made up yet on red light cameras. I want to see Houston's data on accident rates before I make a final decision.

Posted by: Charles Kuffner on January 7, 2007 11:49 AM