Risk compensation and the Illinois seat belt use law

The question of whether drivers alter their behavior due to vehicle safety improvements or public policy changes (laws, enforcement) has been debated for some years. The possibility of risk compensation (offsetting behavior) has been offered: drivers may go faster or operate in a less safe manner in response to such a change. Three 1991 publications purportedly find risk compensation due to mandatory seat belt use laws. However, these conclusions are questionable due to the weaknesses in statistical methods that were used (before/after comparisons, regression). This paper examines whether risk compensation occurred due to the 1985 use law in Illinois. It also compares the results of the before/after method to a preferred technique (ARIMA, developed by Box and others). These approaches are applied to monthly totals and rates of fatalities, injuries classified by level of severity, and total accidents from 1980-1991. Three types of accident are analyzed: vehicle/pedestrian, vehicle/bicycle, and all others. If nonoccupants have suffered adverse consequences, risk compensation could provide the explanation. Much less evidence of offsetting behavior was found. No statistically significant increase in accidents occurred. While other types of safety changes may alter driver behavior, this did not seem to occur in Illinois due to the belt use law.