The wheel deal

Which is more dangerous – the steering wheel or the driver?


In 2006, Kazakhstan banned the use of cars with the steering wheel on the right-hand side. It argued that such vehicles, designed for countries in which people drive on the left-hand side of the road, lead to accidents here, where we drive on the right-hand side. Banning the cars would lower the rate of accidents per mile driven.

In 1975, a Chicago economist, Sam Peltzman, identified the flaw in this argument: Drivers decide for themselves how much risk to take. They buy cars with a right-hand steering wheel because they cost a few thousand dollars less. They understand that the cars are dangerous on Kazakhstani roads, and so they take precautions. If they are forced to buy cars with a left-hand steering wheel, then they will no longer have reason to drive like the little old lady from Pasadena. They may press the pedal to the metal in order to reach their destination more quickly. This is risky driving, but risk has benefits as well as costs. In short, a ban on right-hand steering wheels might not avert accidents.

Here’s the intuition. A driver will speed a bit more if the benefit to him (getting to the concert on time) exceeds the cost (a possible crash). Requiring him to drive a safer car reduces the cost for every speed. So he will drive faster.

Peltzman was writing about the decision of the United States government in the 1960s to mandate seat belts and other safety features that reduce the chances of injury in a given accident. He pointed out that wearing the seat belt may tempt the driver to speed, making an accident more likely. The overall chance of injury to the driver may not fall. In fact, if speeding endangers pedestrians, then the total number of injuries – to drivers, passengers and pedestrians – may rise.

From his statistical work, Peltzman concluded that requiring seat belts, and steering columns that absorbed energy, had not saved lives.  “The one result of this study that can be put forward most confidently is that auto safety regulation has not affected the highway death rate.”


Building a safer driver


Peltzman’s analysis was sophisticated, but a few technical glitches may be more apparent now than in 1975. The dataset contained observations that varied with time: An accident rate for 1961, another for 1962, and so forth. Some of his estimated models assumed that the relationships between accidents and variables representing potential causes were stable over time. In reality, the relationships might not have been stable, because the underlying variables – e.g., the accident rate, vehicle speed and income – might have been changing over time independently of one another. In addition, the dataset included observations for each state in the U.S. The unique characteristics of a state may affect auto safety on its roads in ways that a statistical model cannot capture explicitly.

These small potential flaws do not blunt Peltzman’s basic point: Regulation too often takes human behavior for granted. Auto regulators viewed safety as an engineering problem. Build a safer car, and fewer people will die in accidents. Peltzman showed that a regulation – such as requiring safer cars – may itself affect the driver’s behavior, and in perverse ways.

If drivers demand more safety, then they may indeed drive the safer car with care. In that case, the regulation may have the desired effects. But it will be superfluous, since the drivers would have wanted to buy safe cars on the market. Regulators would not have had to force automakers to produce them; the profit motive would have seen to that.

At times, regulation benefits us. The driver may not consider that his ill-maintained engine generates air pollution, since he does not suffer most of this pollution himself. Requiring annual inspections of engines may clear the air.

Even here, regulators could borrow a few tricks from the market. They could issue among drivers the number of permits to pollute (by skipping inspections) that corresponds to the level of pollution that they are willing to tolerate. Drivers who find it easy to reduce engine emissions – perhaps because they have new cars – would sell their permits to drivers who cannot afford to clean up. Thus the government would cut pollution to the tolerable level as cheaply as possible. Peltzman’s point – that regulators must consider behavior – holds again.

In 2007, Kazakhstan’s government canceled the ban on right-hand steering wheels just before elections, noted an English newspaper, The Telegraph. The reason was presumably political. But an economic reason exists as well. -- Leon Taylor, tayloralmaty@gmail.com



Good reading

Sam Peltzman. The effects of automobile safety regulation. Journal of Political Economy 83(4): Pages 677-726. August 1975. Online at www.jstor.org. Fun to read.


References

Gethin Chamberlain. Kazakhstan election a 'foregone conclusion'. The Telegraph. August 12, 2007. Online.