Automobile Safety Regulation and Offsetting Behavior: Some New Empirical Estimates

Abstract

Suppose that engineers can demonstrate that air bags will reduce the risk of death in an automobile by 25 percent for any given frequency and severity of accidents. Would the installation of these devices necessarily reduce the fatality rate by 25 percent? The answer depends upon the response of drivers to the increased protection from dangerous accidents. If they increase their (speed, recklessness, driving while intoxicated, driving in unsafe conditions, etc.), they may realize substantially less than a 25 percent reduction in expected fatalities. Such offsetting behavior is not irrational: it merely represents a substitution of the marginal benefits of driving intensity for the reduced marginal cost of risk. If offsetting behavior actually occurs, it may be realized in increased risks for bicyclists, motorcyclists, and pedestrians. These externalities could be substantial unless there is a reduction in risk taking among these groups. As a result, the net effect of mandating air bags or any other safety device is far from obvious. There may be no net reduction in fatalities or serious injuries. These theoretical considerations are at the core of Sam Peltzman's classic study (1975) of automobile safety regulation. For policymakers, however, the key question is how much offsetting behavior actually occurs. As Peltzman (1977) acknowledges, offsetting behavior could be trivial or substantial. In this paper, we explore this issue, providing new empirical estimates of the effects of crashworthiness standards established for automobiles over the past fifteen years. These standards have required the installation of lapshoulder belts, energy-absorbing steering columns, head restraints, padded dashboards, crush-resistant passenger compartments, safer windshield mounting, more secure locks, and a variety of other features.