Yet Another Study

By James J. Baxter. This article originally appeared in the Winter 2011 issue of NMA’s Driving Freedoms magazine.

A recent report in The Wall Street Journal illustrates how public opinion can be misled and manipulated by agenda-driven politics. The Journal is justifiably well-respected for its evenhanded reporting of information regardless of its popularity with the establishment. Nevertheless, it got fooled this time.

The article focuses on a report, originally published in Traffic Injury Prevention, by research professors Michael Sivak and Brandon Schoettle of the University of Michigan’s Transportation Research Institute. The report itself is little more than a tabulation of road deaths from 2005 through 2009, which declined from 43,510 to 33,963.

Up to this point there isn’t anything to quibble about. The federal statistics of highway fatalities should be reasonably accurate. There aren’t many opportunities for interpretation of what constitutes a traffic fatality. However, when it comes to identifying and quantifying the cause of a traffic fatality it can be “Let the games begin!”

Almost invariably, reports by industry experts support the biases of the funding source. The University of Michigan’s Transportation Research Institute is funded, according to its web site, by “federal and state government agencies, motor vehicle manufacturers and suppliers . . . .” So it is not surprising that Messrs. Sivak and Schoettle’s study interprets the unprecedented 22 percent drop in highway fatalities from 2005 to 2009 as evidence that, among other things, people have slowed down, drunk driving laws are working, and, ironically, distractions such as cell phone use have diminished.

A hot-button topic right now is “distracted driving,” currently the surrogate for talking or texting on a cell phone. The study reports that “inattentive” driving is recorded in 2008 as a “primary” factor in only seven percent of fatalities. However, a DOT spokesman quoted by the Journal picked up the distraction-related figure of 16 percent, an increase of “60 percent” from the ten distraction-related percentage in 2005. (There is no indication of how a police officer is able to determine when a driver is distracted. Perhaps more art than science here. But we’ll get to that.) To the credit of the Journal reporter, he juxtaposes DOT Secretary Ray LaHood’s stated intention to eliminate all cell phone use in vehicles, even hands-free, suggesting a possible motive behind the spokesman’s comment.

The unvarnished truth is that no-one has a clue about how many fatalities are actually caused by distracted driving. There is strong reason to believe that distracted driving is grossly under-reported as a cause of traffic fatalities.

How can anyone really know, when most of the time there are no witnesses and scant evidence? A police officer checking a box at the crash scene does not make it so. It’s a collection of those checked boxes from which the federal statistics are created.

Let’s look at how this plays out in the real world. Two people go out for dinner. They have a nice meal and perhaps a glass of wine. On the way home they are moving with traffic, at a safe pace consistent with traffic and conditions, which may not be equal to the posted speed limit. The couple is having an intense conversation about a family conflict and just as the driver briefly looks at the passenger to emphasize a point, the car in front stops abruptly to avoid a pedestrian. Our subject hits the rear of the stopped vehicle.

You can see the problem of identifying a cause of the crash. Perhaps an occupant of one or the other cars or the pedestrian is killed. Speeding, alcohol impairment, following too close, and failure to have car under control are among the boxes that might be checked on the accident report. Of course, the one box not checked, the actual cause of the crash, is “inattentive or distracted driving.” All checked boxes enter the federal data base that the U of M researchers subsequently analyze. They assume the data are accurate.

I can see how honest researchers could be misled by dubious government statistics. But, how can Mr. Sivak say, quoting the Journal’s writer, “alcohol . . . that is the biggie“ when looking at the cause of crashes? Mr. Sivak concludes that alcohol and speed explain why so many people die on the highway alone, without hitting another car.  This is dangerous and inflammatory speculation. What about fatigue, reduced visibility, snow, fog, icy roads? Then there’s a myriad of health issues such as heart attacks, stroke, diabetic shock, and the effects of medication.

And what about the rarely-mentioned, never-documented cause of many single-vehicle, single-occupant fatal crashes, suicide? I fear this number is much larger than most of us dare to think about.

What the Journal missed, though, is–and this is obvious enough to appear intentional–the fatal mistake of not considering vehicle miles traveled (VMT). Raw numbers mean what you want them to mean, as we have seen. That’s exactly what happened when the Journal’s reporter asks the usual suspects.

A published comment by one meritorious reader of the article sums up this fatal flaw:

The number of road fatalities per VMT in the US from the 1920s to today has followed the same log-linear trend downward, decade in and decade out. Short periods of increase in per-VMT road deaths and a large reduction appear as noise, but are uncorrelated with safety features like speed limits and seat belts. The limit to traffic fatalities is cultural, not technological. It’s based on what people find acceptable, and what they adjust to. There’s no way to beat the trend line, not per VMT.

So when you read a research report, look at the funding and follow the money.

Not an NMA Member yet?

Join today and get these great benefits!

Comments are closed.