Federal prison residential drug treatment reduces substance use and arrests after release

Objective: The effectiveness of federal prison-based residential drug and a lcohol treatment programs was evaluated using event history procedures that addressed the problem of selection bias and included a wide range of contr ol variables. Methods: The sample comprised 760 treatment subjects and 809 comparison subjects. Treatment subjects were from 20 different prisons of m edium, low, and minimum security levels. Comparison subjects were drawn fro m over 30 prisons. Results: The results indicated that individuals who ente red and completed in-prison residential treatment were less likely to exper ience the critical postrelease outcomes of new arrests and substance use du ring the first 6 months following release. Conclusions: Without controlling for selection bias, the effects of treatment would most likely have been a ttenuated. The results have greater generalizability than other studies of prison-based treatment. This study occurred within a multisite context of 2 0 programs serving both male and female inmates and operating within differ ent security levels and different geographic regions.