The Impact of Cognitive-Behavioral Therapy on the Recidivism of High Risk Probationers: Results from a Randomized Trial

Community corrections are being used with increasing regularity for the supervision and management of serious and violent offenders. Attempts to increase the frequency and severity of conditions of supervision have not resulted in meaningful decreases in crime rates among this population. Some encouraging results, however, have been observed when a treatment component is integrated into supervision protocols. This dissertation first examines the theories and current research that inform this shift in strategies. Secondly, we evaluate for the first time, a cognitive-behavioral therapy intervention developed to reduce recidivism within a high-risk, male probation population. This dissertation begins with a review and synthesis of the literature, both in criminology and psychology, regarding the development of cognitive-behavioral techniques designed to reduce recidivism. Next, the unique characteristic of the intervention being evaluated are set out in Chapter 3. The logistics and characteristics of the randomized trial itself are discussed in Chapter 4. This section includes an overview of the risk forecasting procedures used to identify the experimental sample and the randomization scheme employed. In the following section, the impact of the cognitive-behavioral intervention delivered in Philadelphia is evaluated. Using techniques standard within experimental research, a significant reduction in the prevalence of non-violent offending and some forms of drug use are identified. An instrumental variable analysis is then used to better specify effect sizes in light of relatively high treatment dilution. Finally, implications for future research and public policy are discussed in Chapter 6. After 12 months, there were some significant and meaningful differences within the measures of prevalence of offending. Fewer offenders assigned to the treatment group (33.9%) than control (40.5%) were charged with an offense of any kind (p=.041). Therefore, assignment to the Life Skills program caused a 7.5% decrease in the number of offenders committing non-violent crimes. Significant reductions were also noted in the proportion of urinalysis screenings that were positive for PCP and time-to-failure for non-violent offending. Using randomization as an instrumental variable to compensate for treatment dilution, the reduction in the prevalence of non-violent offending was estimated at 18.8%. This research contributes to the broader literature by reinforcing the hypothesis that an integrated treatment-control supervision strategy is a viable approach for probation agencies seeking to both increase levels of control and reduce recidivism. Specifically, the results reported here represent the first, randomized outcome evaluations of an innovative form of cognitive-behavioral therapy with that specific goal. Secondly, the integration of these findings into the literature using meta-analytic techniques may better inform our understanding of the actual effects and promises of community-based recidivism-reduction programming. Finally, the innovations in experimental design and implementation developed during this project may serve as both an inspiration and a caution for other experimental criminologists