Proceed With Caution


The other day I heard a perky public service announcement in which Big Bird urges children to make sure they sit in the back seat of the car. Big Bird does not explain the reason: to avoid being killed by a government-mandated air bag.

Though air bags are credited with saving hundreds of lives each year, they can be a menace to short people, adults as well as children. They've been responsible for more than 100 deaths since the late 1980s. In response, the federal government has begun granting waivers that allow vulnerable individuals to have air bag deactivation switches installed in their cars.

Thus, after being forced to pay a premium for a car that poses a special danger to them, these people must beg for permission to neutralize the danger. Even then, because of liability concerns, they may not be able to persuade their dealer to install the device. As of mid-1998, 30,000 people had waivers, but only 1,000 had switches.

In addition to imposing a potentially deadly risk on a subset of drivers and passengers, mandatory air bags raise the cost of new cars for everyone by several hundred dollars each. In the May/June issue of Contingencies, the magazine of the American Academy of Actuaries, software consultant Gerry Smedinghoff estimates that air bags cost $1.7 million per life saved.

That's comparable to the cost of an $8 million plan to install safety sensors on the escalators in Washington's subway system, an improvement that would have prevented five deaths during the last 13 years. By contrast, Smedinghoff estimates that seat belts cost only $150,000 per life saved.

So even if mandating air bags is questionable, surely it makes sense to require the use of seat belts. Then again, maybe not.

In a recent Cato Institute paper, British risk expert John Adams explains that the evidence concerning seat belts is not as clear-cut as you might think. Wearing a seat belt substantially reduces the chance that you will be killed if you get into an accident, but that's not the end of the story.

"As people perceive themselves as safer or better equipped against a danger," Adams notes, "they are more likely to take more risks." If it encouraged reckless driving, an increase in seat belt use might not reduce the overall fatality rate: Deaths could go down among drivers while going up among pedestrians, cyclists, and rear-seat passengers.

Looking at the experience in Britain, "which has been singled out as the only jurisdiction in the world in which it is possible to measure fatality changes directly attributable to a seat belt law," Adams sees just such a pattern. And after adjusting for a surge in alcohol-related crashes followed by a crackdown on drunk driving that coincided with the enactment of Britain's seat belt law, he finds no evidence of a net reduction in fatalities.

A second report from the Cato Institute suggests that another platitude of traffic safety, "55 Saves Lives," also needs modification. Stephen Moore, Cato's director of fiscal policy studies, notes that after Congress repealed the 55-mile-per-hour national speed limit law in 1995, the traffic fatality rate continued to decline, reaching the lowest level ever recorded in 1997.

Between 1995 and 1997, total fatalities (not taking into account miles traveled) increased by 0.4 percent in the 33 states that raised their speed limits and by 0.2 percent in the other states. Even if the entire difference is attributed to higher speeds, it amounted to only 64 deaths in 1997.

That means the dire predictions of the activists who supported a national speed limit were off by a factor of 100. One explanation for the modest safety cost is that variance–cars traveling at different speeds on the same road–is a much more important factor in accidents than speed itself.

Moore emphasizes that any fatalities have to be weighed against the benefits of higher speeds–otherwise, why not set the limit at 40, or 25? He estimates that "the 55-mph federal speed limit cost America $3.2 billion a year in lost output" because of extra time spent in transit, while fuel savings have been trivial.

Instead of imposing a speed limit on the rest of us, Congress should have proceeded with caution. "Not so fast" is good advice for politicians in a world of unreliable forecasts and unintended consequences.