Sunday, July 3, 2011

Red-Light Cam "Analysis," Lynnwood-Style

The Herald has been doing some first-rate reporting lately, shining a spotlight on Lynnwood's continued use of flagrantly extortionist photo-enforcement devices. Scott North's July 1st article should be required reading for anybody concerned about this issue.

North sums up the reality in a perfect one-line zinger: "Bottom line: It's premature for anyone to say red-light cameras have caused an outbreak of traffic safety in Lynnwood." (I would prefer to have substituted the word "impossible" for "premature," but let's not quibble.)

Now, on the face of it that doesn't really say anything that any intelligent and honest person didn't already know. So then why should this article be required reading? Because it highlights and underscores the extraordinary lengths to which this city will go in order to try to justify these devices and present them in a favorable light, despite basic common sense and overwhelming evidence to the contrary.

Way back in the Paleolithic Era when I was a young lad in college, I majored in a fairly math-heavy discipline and statistics was a basic part of those studies. Consequently, I nearly laughed out loud when I saw Lynnwood's "analysis" document, which you can read for yourself at the above-linked Herald article (click on it and look for the Adobe Acrobat link near the top).

This is one of the most blatant and pathetic exercises in desperate spin-doctoring I've seen in a very, very long time.

Where to begin... Well, first off, you'll note that the city did not simply provide the base data and allow the reporter to draw his own conclusions. Rather, the city "helpfully" drew some conclusions for him in the form of percentage decreases in total collisions and total injuries.

This is a gloriously hamfisted attempt to influence the reporting! And to a statistician, it's nothing short of hilarious. I'm reminded of a child trying to make the case to his parents that pizza is actually a health food because it features all the major food groups, so of course we should have it every night for dinner, etc.

Specifically, the report concludes with two figures that show a 9% decrease in total collisions and a very dramatic-seeming reduction of 32% in total injuries, implying that these significant-sounding reductions are all thanks to the cameras. Who can argue against cameras that cause such significant-sounding reductions in collisions and injuries, right?

I can! First off, the cameras may well have contributed to the reduction in collisions and attendant injuries, but not for the right reasons. To the extent that they create an environment overtly hostile and predatory to motorists, they encourage potentially tens of thousands of people to avoid Lynnwood entirely. I guess I have to concede this point to the camera cheerleaders -- ensuring that there are fewer people on our roads is indeed a sure-fire way to reduce accidents. Of course, it's also a perfect way to sabotage the many businesses and employers that reside in this town, which is particularly counterproductive considering how catastrophically dependent the city is on sales tax and employee head taxes, but nevermind all that.

Secondly, there were no benchmarks presented at all -- just a few selectively chosen figures cherry-picked by the city. Without knowing how non-camera intersections have fared during these same time periods, or how other municipalities have fared, these figures don't say much of anything.

Thirdly, the percentage delta decreases mean absolutely nothing considering how small are the numbers of collisions and injuries. A figure like 32% sounds so significant! So huge! My God, it's a full third and that's just enormous!

Except that it's none of those things. It's 30 as opposed to 44, and distributed as it is among the numerous intersections in question, it only amounts to a half handful of reductions per intersection. Suffice it to say this reduction doesn't rule out mere chance -- and that's the understatement of the year.

And finally... how does the city explain those intersections where collisions or injuries increased after the installation of red-light cameras? It's right there in their very own figures.

I could go on poking more holes in this "analysis," but at the end of the day, the real bottom line is that all of this presumes that the figures presented are actually correct -- and considering Lynnwood's established reputation of misinforming people regarding the validity of photo-enforcement evidence against them, I have every reason to doubt the the veracity of these figures, right out the gate. (See Robert Mak's report from last December if you doubt that -- visit the link, click on the video and scroll ahead to 9:50 and see what David Andree's experience was -- talk about an eye-opener.)


Anonymous said...

Hey, I have that same book, as well as the terrific "Statistical Tricks and Traps" by Ennis Almer. Good blog and please keep up the expose.

BTW, my husband and I are on Whidbey Island and we have made it a point to avoid Lynnwood these days.

Anonymous said...

I haven't shopped/eaten/watched a movie in Lynnwood in years, and I have no plans on going back. Their Gestapo police force have made it an unfriendly environment to motorists. I just don't need the hassle when I can shop elsewhere.

Anonymous said...

Nice work!

Anonymous said...

And let's not forget that 50% of this increase in ticket camera revenue goes to a for-profit private corporation.

I'm surprised there hasn't been a huge backlash from retailers. There has to be a large number of people who avoid Lynnwood entirely now just like my circle of friend do.

Anonymous said...

Another huge factor is the major road realignments in the past six years on 196th and Highway 99. It takes time to get used to them.

ted trepanier
not the WDOT one

Anonymous said...

For nearly every argument "con" with respect to statistics there is also an argument to the "pro", and this case is no different. There are also numerous national and international studies suggesting there are safety implications, although I will agree they are all still quite short in term so correlations are difficult. I guess my only issue is that you only argue from one perspective and don't remain objective. You say the city isn't doing this, so what's your excuse?

People's Republic of Lynnwood said...

The simplest and most straightforward answer is that it doesn't matter that there are just as many pro-camera as there are anti-camera positions. A cursory analysis reveals that pro-camera arguments are based on fallacious reasoning, distortions, or a similarly promiscuous relationship with objective truth.

As for the "numerous national and international studies suggesting there are safety implications," it's interesting to note that they tend to be written by bureaucrats in the pay of red-light camera profiteering cities and therefore know on what side their bread is buttered... as opposed to being written by individuals who are actually qualified to make such assessments, such as engineers. Likewise for organizations whose existence is based on their monopsonistic relationship with government bodies on the lookout for fresh revenue sources.

In closing, one would think it relatively obvious that I would only argue from one perspective, as I am an exponent of that perspective. Regarding my supposedly "not remaining objective," I can only enjoin you to reread the article and brush up on your analytical skills, as you will be hard-pressed to find an example of a subjective argument presented from my part.

The fact that my objective argument was compelling enough to make you think I was being partisan tells me plainly that it was compelling.

Thanks for the comments.