Alcohol Ban and Crime: The ABC's of the Bihar Prohibition

Abstract

While alcohol regulation policies may help control crime, the effects of such policies are highly dynamic and may be difficult to understand using the annual crime data normally available in developing countries. We use district-level monthly crime data from the Indian states of Bihar and Jharkhand and a difference-in-differences reasearch design to provide evidence that a prohibition imposed in Bihar led to a 0.21 standard deviation reduction in the incidence of violent crimes but had no significant impact on non-violent crimes. These results illustrate the idea that intoxication may reduce risk aversion, thereby increasing the probability of committing a violent crime. Our results are concentrated in areas where the ban may have had a larger impact - due to either higher alcoholism at the baseline or where the ban may be easier to implement - and are robust to a number of alternate specifications, including the use of a synthetic control group.