Remember when Albemarle first started planning red light cameras to reduce accidents at intersections and lots of people pointed out that VDOT’s own studies found that they increase the rate of accidents? The Board of Supervisors didn’t believe that, and one year ago they put up red light cameras at the intersection of Rio and 29N. The numbers are in and—lo and behold—the number of accidents at 29 and Rio increased with red light cameras, Rachel Ryan reports for CBS-19. There were 23 crashes from December through July, up from 22 during the same period one year earlier.
OK, so the goal of reducing accidents didn’t work out. But this intersection was chosen because of the high rate of people running the light there—they’ve at least caught a lot of people, right? Nope. The county pays the first $10,000/month in traffic fines to a private firm (RedFlex, an Australian company) to operate the system, and not once have the cameras even brought in enough money to cover those costs. Still, Albemarle County Sergeant Darrell Byers claims that they’ve been effective. By what metric, though, he’s not saying. The county has a three-year contract with Redflex, and getting out of it requires paying unspecified termination fees. There are no plans to reconsider the use of the cameras.
Minor quibble: I don’t see anything supporting the allegation that the county planned the red light cameras in order to reduce accidents at intersections in the links provided above.
Also, an increase to 23 from 22 is hardly statistically significant. Sure, the cameras didn’t decrease the number of accidents but, again, I don’t see where anyone said that they would. I’m not saying that no one said they would decrease the number, I just don’t see it.
The hook’s site is down right now, so maybe the offending quote is there….
The solution to this failed govt program is obvious to me. We need more cameras and larger fines. The idea isn’t bad, it is simply underfunded.
My statistics show there would have been 50 more crashes without the cameras.
Oh, I have no idea if it’s linked anywhere there—I just remember that was the whole point of the cameras, as the county talked about widely at the time. (There was a long, drawn-out debate, in which folks like John Whitehead—and me, though I’m no John Whitehead—pointed to VDOT’s study that they make intersections more dangerous, and the county responded with other studies that demonstrated otherwise.) For instance, from the Progress in December, Albemarle’s senior assistant attorney Annie Kim said: “The county’s goal has been, and remains, the reduction of accidents and dangerous driving at this intersection…”
I started to write an explanation about that in this blog entry, but realized that it didn’t matter, since the goal was to reduce accidents, and increasing them (or even maintaining them) is clearly not, in fact, reducing them. :)
“increasing them (or even maintaining them) is clearly not, in fact, reducing them”
Sad really that things like that need explaining. Makes me nostalgic for the days of poll taxes and literacy tests.
1. Not at all statistically significant. 22 to 23 should be reported as “no change.” You even italicized “increasing.”
2. One year is not enough time for to say anything. Too small of a sample and there may be a transition period that does not reflect a new norm.
3. Red light cameras reduce severity of accidents. I’d take a few rear-end collisions caused by sudden stops over high-speed herringbone collisions any day.
4. I thought the main criticism against these things was that they were just being implemented as a cash cow for the government. Now that this is obviously false, the problem is that they don’t generate enough revenue?
5. Compare the cost of running the camera with the cost of citing red-light offenders the traditional way. That’s how to determine cost-effectiveness.
“Sad really that things like that need explaining. Makes me nostalgic for the days of poll taxes and literacy tests.”
If I may point out that MY point was that in the absence of documentation of the assertion that they WOULD reduce accidents (which Waldo has since provided), the fact that there was a statistically insignificant increase in the accidents was relevant to nothing at all.
I’m sorry you missed that, ‘the boss of me’. I guess I assumed that people could infer that one statement made (I didn’t see where people were claiming there would be a reduction in accidents) was related to another (an increase from 22 to 23 is not significant).
Sad, really, that I have to explain this.
Oh yeah, one more point:
6. Total daily traffic is increasing on the US 29 corridor, which means we would expect to see increased collisions as well. The proper metric to evaluate the cameras is collisions per AADT or comparing the change in an intersection with the camera to a similar one without.
I have not calculated the p-value—I have no idea if it’s statistically significant. If you own a copy of SPSS and want to plug the numbers in and report the results, be my guest, but until somebody does so, I don’t think we’re actually qualified to conclude that.
That’s a criticism that some people were making. You write that it’s “obviously false,” but you also wrote that “[o]ne year is not enough time for to say anything”—I think you’ve got to pick something and go with it. :)
Compare the amount that we ticket people for running red lights when police catch them running red lights with the cost of how much we fine them when they’re caught with the cameras, and I think you’ll find the numbers are pretty lousy.
What I’m really concerned with is (and what I wish the Newsplex article would have presented is): are there any changes in crash types? Are we still seeing the same number of right-angle/head-on crashes (the most dangerous kind)? How about a statistic on injury/fatal crashes in those time periods? And again, as has been asserted already, 8 months is an incredibly short time period to make any conclusions. In traffic engineering, 3 years is the norm.
Bottom line: if you’re opposed to red light cameras on the basis of personal privacy or unjustified spending of taxpayer dollars, that’s fine. But saying that crashes are “increasing” (whether or not you attribute that to the red-light cameras) is not a valid statement, scientifically.
I’m sorry, I hate to belabor the point, but this is simply wrong. There was one more crash than in the same period in the prior year. Adding one to any number is an increase. Period.
It may be accurate to say that the change is not statistically significant—we don’t have enough data to know that. It’s definitely accurate to say that “it barely increased” or “there wasn’t much of a change” or “rates stayed about the same.” But it’s completely wrong to argue that the number of crashes hasn’t increased. It absolutely, unquestionably has.
None of this, BTW, is to argue with your point, Anthony, that we need more data. You’re right, we do. This is very little to go on, but it’s all we’ve got.
I’m sorry, but that’s a bunch of hooey. By saying “CRASHES INCREASED” you, as a journalist, are implying that there was some sort of meaningful change. Yes 23 is a bigger number than 22, but it is without the context of the number of vehicles passing through the intersection, and is it is disingenous to present this in a news headline as having any sort of meaning.
“I think you’ve got to pick something and go with it.”
Ok, I’ll stick with probably all around. We’ll wait and see if the revenue does come in (I bet there is a delay due to processing time anyway). I doubt we’ll see a windfall though.
Against my better judgment, I wandered over to the Newsplex site to find commenters STILL enraged that the government is just using this to pick their pockets. Not only does the county extort hard-earned money but they do a lousy job of it!
What’s really going on here is our confusion between private and public space. Lots of Americans instinctively believe that when they are in their cars they are on private property and they should be allowed to do whatever they want. In reality, they are in a very public realm where their decisions have real impacts on other people.
I, for one, can’t wait until there are plenty cameras on me at every moment of my life. Constant scrutiny makes me tingle.
The comforts you’ve demanded are now mandatory.
If you want absolute privacy, don’t use public roads.
“If you want absolute privacy, don’t use public roads.”
Well. There’s some sound practical advice, and not at all hyperbolic. But I think you misunderstand – I enjoy the scrutiny. That’s why I bought a car with OnStar…
“One could also imagine an eager police chief acquiring the data to issue speeding tickets en masse.”
from this article:
http://www.wired.com/threatlevel/2011/09/onstar-tracks-you/
Mmmm. I love the smell of data streams in the morning. Smells like… victory.
Let me get this straight,
Albemarle county pays $10,000 a month for monitoring one intersection?
For $120,000 we could have 2 more cops on the street and they could sit at that intersection all day long.
Why?
@ Waldo and failing
The numbers don’t have to be statistically significant. The point is they’ve spent a ton of money and nothing has changed. Continuing
No, you’re inferring it. I don’t really imply things, as a general rule—I just come out and say things.
In this case, I thought through the problem of whether it had actually increased, looked at all available data, determined that a) it had increased b) that there was no further data and c) that whether presented as an increase or as a stable number, the important part was that the numbers had failed to go down. I’m perfectly happy with that decision.
Listen, I don’t know how dumb you think that this site’s readers are, but you should know that they’re pretty bright. I provided all available numbers (if you know the number of vehicles that passed through the intersection, I hope you’ll share), giving everybody exactly as much information as I have. And then, because I’m such a swell guy, I provided this whole commenting system so that we could have this very discussion that we’re having right now. Not a lot of people come here just to read my little summaries—basically everybody clicks through to read the discussion. That’s kind of the point of the site. I’ve got faith that when I write “increasing,” and then provide the numbers “22” and “23,” that people are capable of doing the math and verifying that 23-22=1, that a positive number means “increase,” recognize that a difference of 4.5% is not enormous, and contextualize this information.
cvillenews.com has far and away the most intelligent, informed discussion on any area website. That’s because the audience consists of some of the most intelligent, informed people in the area. So I feel pretty good that nobody’s getting the wool pulled over their eyes.
I go through that intersection regularly. I’ve seen people cram on their brakes in the intersection and an increase in rear-ending “close calls” since the camera was installed. My observation is that the camera causes poor decisions.
Now I learn that the camera results in the county’s citizens making a substantial expenditure to a foreign firm on a monthly basis.
The camera does not reduce accidents. It does not generate revenue. It costs us taxpayers a bunch of money with no benefit. Why have it?
In addition to the practicalities of an upside-down cost benefit analysis, these cameras are a detriment to society in that they advance acceptance of a dehumanizing “Big Brother” environment where a person can have a machine accuse them of a crime. As I see it, that is antithetical to the societal structure set forth in the Constitutions of both the USA and Virginia generally referred to as “the land of the free”. We have the right to be faced by our accuser and a machine just won’t do.
No benefit. Costs lots of money. Dehumanizing. Antithetical to the way of life we stand for. Why do we have it?
Is is a different take with more detailed stats: http://www.readthehook.com/101104/seeing-green-rio29-cameras-nab-7638-drivers
[…] decreased after the installation of red light cameras last year, Dave McNair reports for The Hook. Initial data demonstrated that crashes had increased at the intersection—and they have—but police say that those numbers include lanes that are not monitored by red […]