|Practice area:||Justice and Legal Affairs | Public Policy | Rehabilitation|
|Client:||Ministry of Justice (UK)|
|Keywords:||qualitative analysis quantitative analysis stakeholder surveys and consultations|
London Economics undertook an evaluation of the impact of a cognitive-behavioural offending behaviour programme (Enhanced Thinking Skills, or ETS) on the reconviction outcomes of treated prisoners. By combining new data collected as part of the SPCR survey and existing management information for the first time, the causal treatment effect (average treatment effect on the treated, ATT) was estimated using propensity score matching, a retrospective quasi-experimental research design. Controls were matched using a radius matching algorithm based on the odds ratio of the propensity score, predicted using logistic regression. This yielded a control group with no statistically significant difference from the treated group across a set of 42 variables pre-selected as being simultaneously predictive of the probability of treatment and the probability of reconviction (consistent with the theoretical and empirical literature). This methodology has since been established as a blueprint for future evaluations of interventions at the Ministry of Justice.