Citation

Jofre-Bonet M, Sindelar JL. NBER Work. Pap. Ser. 2002; 2002(online): w9038.

Copyright

(Copyright © 2002, National Bureau of Economic Research)

DOI

unavailable

PMID

unavailable

Abstract

Drugs and crime are known to be correlated, but the direction of causality and the magnitude of the relationship have not been well established. We take a new approach to estimating this relationship and examine a little used, multi-site dataset of 3,500 inner-city drug users entering treatment. We analyze the change in crime and in drug use pre and post treatment, controlling for other covariates. We take first differences to address omitted variable problems. For our sample, we find that treatment reduces drug use which, in turn, reduced drug decreases crime. Reduced drug use due to treatment is associated with 54% fewer days of crime for profit, ceteris paribus. Our evidence suggests that, reduced drug use is causally related to reduced crime. This finding is robust to different specifications and subsamples. Our findings broadly suggest that drug treatment may be an effective crime-fighting tool. Given the huge and growing expense of the criminal justice system, drug treatment might be cost-effective relative to incarceration.