Can Swift, Certain, and Fair Programs to Reduce Substance-Impaired Driving Be Successfully Replicated?

• In Event: Drugs: Markets and Policy Considerations

Abstract

Efforts to reduce substance-impaired driving among justice-involved individuals traditionally prioritize sanction severity over certainty and celerity, which is problematic because hyperbolic discount rates and time inconsistencies in decision-making are particularly acute among substance-involved populations. In 2008, North Dakota adopted a program for substance-involved offenders, replicating the 24/7 Sobriety model established in South Dakota. The program required participants to abstain from alcohol or other drugs and undergo very frequent testing, focusing on swift, certain, and moderate violation sanctions to deter recidivism. Those failing or missing a test were subject to a brief stay in jail. We evaluate variation in implementation choices at the state and county-level to inform the process of adaptation to new jurisdictions and evaluate a set of process indicators of site and participant-level compliance. Exploiting variation across counties in the timing of program implementation within North Dakota, we estimate differences-in-difference and triple-differences models to identify the effect of 24/7 Sobriety on arrests for driving under the influence of an intoxicating substance (DUI). We find that test violation rates are higher in North Dakota than South Dakota, but that DUI arrests declined by approximately 26% after the program became operational (IRR=0.742; p<.05).

Authors

• Greg Midgette, University of Maryland

• Nancy Nicosia, RAND Corporation

• Beau Kilmer, RAND

• Paul Heaton, University of Pennsylvania