Turning on the brakes

Safety is a necessary but not sufficient condition for buying a car.

If a given vehicle has no air bags, seat belts or NCAP crash test rating, you’d have to be an idiot to buy it. Contrarily, there is no hard and fast correlation between a car’s safety rating and its market value (in fact, without looking at the numbers, I’d guess there’s an inverse correlation primarily due to luxury sports cars). Selling a car, then, rests on achieving some minimum standard of safety and maximizing the quality of other features—comfort, design, handle—that drivers care about.

The latter qualification is something consumers can pretty easily gauge through experience and the Internet. The former, unfortunately, defies any supposedly fair method of assessment. We have no idea how safe our cars actually are until we crash them.

When we do crash them, it’s usually due to a confluence of factors, only some of which may directly relate to the vehicles involved. A car accident is so complex an event it defies simple classifications like “stickiness of an acceleration pedal” or “improperly placed floor mats.” Obviously, an automaker alone can significantly bump up the conditional probability of an accident, but the real probability of a lethal crash from driving, a Prius, for instance, is determined by myriad disparate forces of the universe acting in combination over a significant time horizon. When Toyota issued a recall of the Prius in early February because of a software glitch resulting in a small but potentially deadly time lag in the line’s braking system, blame could reasonably have been placed on engineers, regulators and environmentalists. Cars are a much more complex version of pencils, and, in the words of Milton Friedman personifying a pencil, “Not a single person on the face of this earth knows how to make me.”

To Congress, however, Akio Toyoda makes Toyotas, Mark McGwire destroys baseball and CEOs deface the economy. Toyota, specifically company President Toyoda, is strictly to blame for any and all injuries sustained from failures of their product line. The House Oversight and Government Reform Committee made this abundantly clear on Wednesday, chastising Toyoda in the process for not showing enough remorse and waffling on questions. Once again, it was a prototypically unproductive session of lawmakers.

When Congress spends several hours scape-goating a big-name executive, it necessarily precludes itself from doing any thorough analysis on the problems that landed said scape goat in Washington. Toyoda himself, recipient of a law degree and MBA, probably had very little to do with the safety defects in Toyota vehicles; he does not design braking software or floor mats or acceleration pedals. He has no formal schooling in engineering or software programming. He cannot tell you why exactly so many Toyotas’ brakes failed.

This question should be the focal point of any extended Toyota investigation. To Congress’s credit, a number of House and Senate members did broach this question with both Toyoda on Wednesday and U.S. President of Motor Sales James E. Lentz III, a marketing executive with an MBA, on Tuesday. Was the problem mechanical or electrical in nature? Toyota does not know for sure at this point, although Lentz leans mechanical. For recalling purposes, a mechanical mistake is far less costly than an electrical one; consequently, without presentation of empirically gathered data, Lentz’s testimony about anything even remotely technical should be disregarded.

Even if Lentz could fluidly discuss the physics of anti-lock brakes or the coding errors behind software malfunctions, how could Congress or, more importantly, consumers understand him? It is likely the case that understanding why Toyotas are unsafe requires a level of technological literacy uncommon among the people who drive them. Short of enlisting the help of our engineering brethren, there’s little we can do in the way of properly evaluating a vehicle’s safety. Brakes and car software are merely the tip of the iceberg: To fully get a sense of how safe a car is, you’d have to have detailed knowledge of crash test methodology, the deployment of air bags, how other drives react to the prospective car on the road, etc. Even if you did all that, it’s probable most of your assumptions from the research would be disproved in five years.