The risk compensation hypothesis suggests that drivers enjoying greater safety will drive more recklessly and thereby impose greater risks on nonoccupants. We provide a test of the risk compensation hypothesis in the context of state seatbelt laws and belt use rates. Fixed-effects models with policy and demographic variables are estimated using annual state data from 1985 to 2002 to test the effect of seatbelt laws and seatbelt use rates on logged fatality rates for occupants, pedestrians, motorcyclists, and all nonoccupants in separate models. Contrary to the risk compensation hypothesis, the results indicate that both occupants and nonoccupants enjoy greater safety due to state mandatory use laws and increased safety belt use rates. Overall, seatbelt laws and the higher belt use these laws induce do not increase nonoccupant risk exposure. If anything, these laws and the accompanying increase in belt use result in safer driving behavior. (Author/publisher)
Abstract