de Figueiredo on Recidivism and Sentencing Enhancements for Drunk Driving

Miguel de Figueiredo has posted Do Sentencing Enhancements for Drunk Driving Decrease Recidivism? a Regression Discontinuity Approach on SSRN. Here is the abstract:

Using micro-level data on drunk driving arrests in Arkansas, this paper exploits discontinuities in sentencing enhancements at various blood alcohol content (BAC) levels to estimate causally the effect of the increasing penalties on recidivism. Specifically, the research design examines defendants just above and just below a BAC level of 0.15 to see if the effect of an increased license suspension by an additional two months has an effect on recidivism. The paper finds that increased penalties in the neighborhood of the discontinuity in the form of license suspensions have no statistically significant effect on drunk driving recidivism, suggesting that the increased penalty does not deter defendants from committing another drunk driving offense in the future. The paper’s focus on suspension of privileges as a form of punishment, coupled with its rich micro-level data (15,973 defendants from two jurisdictions) and research design that enables reliable causal inferences and estimation, make it contrast with numerous studies in the extant literature.