Depends what you mean by “work”, I guess…
Alongside our colleagues in the U.K. MOJ’s Justice Data Lab, we’ve spent the last year trying to answer that very question. Then we conducted two large-scale impact evaluations to answer these two questions, using data from the Thinking Skills Programme delivered in prisons across England and Wales.
We decided to define “work” in this context as “is associated with a lower level of future reconvictions” and “is associated with lower levels of future prison misconduct“.
The reoffending study involved a treatment group of 20,293 adults (18,555 males, 1,738 females) who participated in TSP in custody between 2010 and 2019. Proven reoffending was measured over a two-year period from the point of release from custody. The prison misconduct study involved a treatment group of 13,891 adults (12,938 males, and 953 females) who participated in TSP in custody between 2011 and 2019.
Over a two-year period from release, men who participated in TSP were less likely to be reconvicted, were reconvicted less frequently, and took longer to be reconvicted, compared to similar males who did not participate in TSP. These results were statistically significant and the effect sizes were very small. Results indicated that over a two-year period following release, females who participated in TSP were reconvicted less frequently, compared to similar females who did not participate in TSP. These results were statistically significant with very small effect sizes.
Over a 6-month period after starting TSP, men who had participated were less likely to receive a proven adjudication (a hearing in which it was decided prison misconduct had occurred) compared to males who did not participate in TSP and received a proven adjudication less frequently. This was the case for both any type of adjudication and for adjudications classified as violent. These results had very small effect sizes and were statistically significant. Results indicated that over a 6-month period after starting the TSP, females who had participated in TSP received any form of adjudication less frequently compared to those who did not participate in TSP. This result also had a very small effect size and was statistically significant.
In conclusion: yes, they do work, but their independent effect on behavioral outcomes appears to be modest.
However, this is perhaps to be expected as a program like TSP would likely be subject to a “scale-up penalty”. Positive effects of an intervention often decrease as delivery is scaled-up and therefore greater variation in implementation, provision, and target populations mean effects can be diluted. Thus, for a widely-delivered basic core thinking skills programme, it’s actually quite an achievement to find a statistically significant effect at all!
You’re welcome prison psychology.
