Drunk driving legislation and traffic fatalities: New evidence on BAC 08 laws

This article reexamines the effectiveness of blood alcohol content (BAC) laws in reducing traffic fatalities. Differences-in-differences estimators of U.S. state-level data with standard errors corrected for autocorrelation show no evidence that lowering the BAC limits to 0.08 g/dL reduced fatality rates, either in total or in crashes likely to be alcohol related, or in states that passed BAC 08 in laws either in advance of or in response to federal pressure. Other legislations, including administrative license revocation and primary seat belt laws, are found effective in reducing fatalities in all specifications. Endogeneity tests using event analyses confirm the differences-in-differences estimates.