Abstract
Although many previous studies have addressed the impact of drinking and driving laws on motor vehicle related deaths, less attention has been paid to what determines such laws. Accordingly, this study examines factors affecting the probability that a state adopts 6 different restrictions on drinking and driving. Specifically, using a panel of state-level annual data corresponding to the 50 U.S. states for the 1985-2002 period, we estimate linear probability and random effects Probit models, paying particular attention to the endogeneity of the motor vehicle fatality rate. Results reveal strong nation-to-state and state-to-state policy diffusion across different drunk-driving laws.