The Peltzman hypothesis revisited: An isolated evaluation of offsetting driver behavior

Nearly all of the previous investigations of driver behavior under varying safety conditions have only been able to evaluate the issue of offsetting driver behavior indirectly. Some studies note that engineering estimates of the aggregate improvements of automobile safety enhancements are substantively larger than economic estimates of the same effects. Others have found that non-occupants such as motorcyclists and pedestrians are under greater risk of incurring accidents when automobiles are made safer. Although these observations are consistent with the hypothesis that drivers respond to safety enhancements by driving less safely (having been instilled with a sense of greater protection provided by the safety improvements), they do not provide a direct evaluation of the relationship between driving environment and driver behavior. Thus, factors other than offsetting behavior may be generating these observed results. A few studies have tested the offsetting behavior hypothesis more directly, but only under highly specific safety conditions and narrow measures of driver behavior. This article undertakes an isolated analysis of driver behavior under varying safety conditions, with data which allows for more general changes to safety conditions and broader measures of driver behavior. Estimates obtained from the evaluation of individual accident reports lend support to the offsetting behavior hypothesis and imply that both aggressive behavior and nonaggressive (inattentive) driving behavior are likely to increase as the driving environment becomes safer.