Cognitive Dissonance on Vehicle Safety

"Sure, you would expect that with the recession on, right? Well, actually, deaths per 100 million miles fell to 1.16 in 2009, down from 1.25 the previous year, an an all-time low."

Wouldn't you expect this anyway? Fewer cars on the road could easily lead to a lower crash likelihood for each car, not just a lower # of total crashes.

What is the Deaths/100 million miles for the recalled Toyota vehicles?

@David Henderson:

Doesn't this turn on the definition of "safe"? Sure, today's vehicles and roads are less hazardous than those of the past. That doesn't mean they couldn't be made a lot safer, if we were willing to abandon other desirable characteristics.

It seems to me that Strickland's quote could be interpreted as meaning that for regulations to improve safety to the point that the regulators would be satisfied, those same regulations would result in vehicles which were undesirable overall. Not an unreasonable position as far as I can see.

David Henderson makes an excellent point. Auto manufacturers could make cars that are built like tanks and have incredible safety ratings; they don’t, however, because no one would buy them. Safety is not an absolute—you have to sacrifice economic resources (e.g. fuel efficiency, cost to purchase), and features/style (e.g. engine horsepower, weight of car, styling of car) for safety. In other words, as desirability due to safety-rating goes up, desirability in other aspects often goes down. There’s no golden safety-standard; it’s a normative decision based on each consumer’s wants.

NHTSA certainly has some responsibility to ensure auto safety, but the ultimate decision on what the acceptable safety risk is lies with the purchaser and not a regulatory agency that has no vested economic or personal interests in the purchase. It is the individual who pays for the car and takes the safety risk when driving it. Strickland’s comment is partially correct, but not for the reason he intended. The problem was a functional issue—the cars did not operate as intended. The danger did not come from an intentional trade-off that made the cars more desirable but reduced safety in the event of an actual crash. Roll-over safety or air-bag deployment are testable, an unintentional design flaw is not. It is not that the NHTSA was at odds over the safety trade-offs that manufacturers knowingly made, it is that the problem was not something that anyone could have foreseen, thus it is incredibly hard to regulated.

The data suggests that driving has been getting safer in recent years (though perhaps it would be wise to examine the study closer). Safety technology is improving as a whole and the NHTSA has been fulfilling its mission. People like Strickland, however, choose to ignore the data as a whole and instead focus on specific cases and claim they are the norm when in fact they appear to be deviations from the norm.