Money & Company

Tracking the market and economic trends
that shape your finances.

« Previous Post | Money & Company Home | Next Post »

Does pricey gas make cars safer?

February 6, 2009 |  6:22 pm

Crash Test Dummy Does costly gas save lives?

That's a question touched on in an interesting new study published this week by the Governors Highway Safety Assn., which looked at fatalities in 44 states plus the District of Columbia. Overall, the study found that between 2007 and 2008 the total number of fatalities declined in all but four states, led by a 28.5% reduction in mortality in Massachusetts.

But by comparing the percentage change in number of fatalities to percentage change in the number of miles driven in the state — a statistic called Vehicle Miles Driven, or VMT, reported by 22 of the states in question — the study makes it possible to zero in on whether the safety gains are entirely linked to the cost of gas.

How? Because time and again, surveys (and common sense) have shown that when gas is expensive, people drive less; conversely, when gas is cheap, they eat up the miles. Since cars that aren't on the road can't kill their drivers, passengers, other drivers or pedestrians, the theory goes, pricier gas would logically spell fewer fatalities. So: Cheap gas kills. And since high gas prices took a big bite out of VMT last year, it would appear a perfect year to establish that relationship.

At least in theory. Reality would appear a bit more complicated....

Here's what the study found: in 17 of the 22 states reporting fatality and VMT statistics, fatalities decreased substantially more than the decline in miles driven. In Illinois, for example, the number of miles driven decreased by 4%, year over year, but the fatality rate decreased by 16% — four times the rate expected under the pricey petrol postulate. On average, fatalities decreased at more than three times the rate that miles traveled declined.

The big outliers were Wyoming and New Hampshire, where VMT decreased, but fatalities increased. Also of interest were the roads of Vermont, which was the only state with a VMT increase (a 1% uptick), yet 10.6% more people died there than in 2007. 

What is the takeaway from all this? We're not statisticians here at Up To Speed — in fact we barely passed trigonometry and got a C- in calculus — but it would seem to indicate one of three things:

  • That there is either a nonlinear, but real, relationship between miles driven and fatality rates. That seems complicated by the fact that states with the largest delta in VMT had fatality declines that were relatively smaller than those with smaller changes in VMT (for example, South Carolina had a 5% dip in VMT and a 15% reduction in fatalities, a 1:3 ratio; in Minnesota, VMT went down only 2%, but fatalities decreased 12%, a 1:6 ratio).
  • That fatalities and miles driven are independent of one another. On the face of it, this seems unlikely because there is clearly a relationship to total amount of driving and vehicle-inflicted death. If nobody drives, nobody dies in car wrecks.
  • That there is some relationship between the amount of driving done and the total carnage inflicted, but there are other factors out there — independent variables, so to speak — that can raise or lower the death rate regardless of changes in miles driven, such as better safety equipment.

This last one, according to our way of thinking, and the Governors Highway Safety Assn., would seem the most likely. The association's executive director, Barbara Harsha, points a finger at increased seat-belt use and increased enforcement of seat-belt laws as relevant variables.

Another spice in the soup: slower driving speeds. High gas prices may not just cause people to drive less, but more slowly. In Oregon, the average speed was down more than 1 mph last year (who knows how they calculated that). And there's little doubt that slower speeds lead to fewer deaths.

Nowhere was mentioned improved safety systems in automobiles. Can't side-curtain airbags get a little bit of glory?

—Ken Bensinger

Photo of a crash-test dummy by athena1970 via Flickr