BAC or Bacchus? Location versus quantity of drinking

Introduction: State drunk driving laws in the United States are designed to deter impaired driving. These laws and in particular the BAC (blood alcohol concentration) limits have been found to be quite effective in terms of reducing the alcohol involved motor vehicle accidents and fatalities. It is not clear though whether these reductions are due to the decreases in alcohol consumption or due to reductions in driving after drinking. This paper tries to identify the possible underlying mechanisms through which the BAC laws may have reduced the alcohol related accidents. Methods: We use data from the 2001-2002 National Epidemiologic Survey on Alcohol and Related Conditions (NESARC). These data include information about incidences of impaired driving, the typical location of drinking (in public versus elsewhere), as well as the typical quantity of drinking. We jointly estimate heavy drinking and impaired driving incidences. We also estimate joint models of heavy drinking and typical location of drinking. Results: We find that the .08 BAC limit significantly reduce the probability of impaired driving and drinking in public whereas they do not have any significant effect on heavy drinking. Discussion: These findings confirm that the BAC limits serve as an appropriate policy instrument to correct the externality of drunk driving. The results are robust to different definitions of heavy drinking. Other state policies, such as monetary fines, jail sentences, and licence suspension or revocation penalties do not seem to be as effective as the BAC limits.