Thread by @seenafazel

What effects do prisons have on repeat offending? Our new paper examines this across 44 Swedish prisons. V brief thread. Paper #OA: journals.plos.org/plosone/articl… 1/

We compared reoffending rates in people who were in different prisons at the same security level (between-individual), adjusting for measured confounds, and also examined repeat offending in the same person released from different prisons (within-individual). Two approaches. 2/

To do this, we selected a reference prison with the lowest reoffending risk within each of the 3 security levels. Then we followed up released prisoners for reoffending using national crime register for an average of 8 yrs. 3/

Findings were consistent across the 3 security levels. Found no differences in reoffending risks when conducting within-individual analyses. Eg in high security, hazards ranged up to 1.6 between individuals but not when the same person was released from different prisons 4/

So what does it mean? Previous work suggested that the prison environment, eg in its ‘climate’, contributes to reoffending. These findings do not support that. Individual-level risk factors are more relevant to reduce recidivism - supported by other evidence. 5/

Finally, the study suggests that research designs in prison research need to more fully account for individual level factors when studying recidivism. In other studies, we used sibling controls to investigate risk factors: thelancet.com/journals/lanps… 6/

And another study using within-individual approach to investigate the effects of medication on reoffending. Highlights potential role of treatment for psychiatric disorders and substance misuse: jamanetwork.com/journals/jama/… 7/

And this is a systematic review of recidivism rates around the world. Most countries report 2-yr reconviction rates of 30-60%. Rearrest and reimprisonment rates also high:
wellcomeopenresearch.org/articles/4-28/…