Policies that promote diversion from incarceration for juveniles are unlikely to be intended to be health policy, but could have a significant impact on justice-involved youth. These policies and programs emphasize alternatives to incarceration for non-violent offenders, and often involve intensive case management, resource-linkage, and family interventions.
An evaluation of these policies could help inform their implementation and downstream effects. They seem logical in theory, but anecdotally there are concerns that for some programs, violations can lead to incarceration for much lower-level offenses. On the other hand, if effective, these policies could also have an impact on school attendance, continuity in health care, and improved family dynamics.
I think this type of policy would have the greatest impact on low-income, minority youth, who are more likely to be come involved with the justice system and more likely to be sentenced to out-of-home placement compared to their white and higher-income counterparts. There are many outcomes that I think would be interesting. Some health-related outcomes could be mental health related (e.g. self-reported, or scales measuring depression, anxiety, stress, PTSD, or substance abuse) or simply attendance at visits with their primary care provider. I imagine we would need much longer term studies to evaluate changes in metabolic markers or other markers of chronic disease, since this would be a young cohort. Non-health related outcomes that I would also be interested in following would be completion of high school and involvement in the adult criminal justice system.
I wouldn’t feel comfortable randomizing youth to diversion or not, however I could imagine comparing cohorts from different states or cities—one with diversion policies in place and one without—and then following them prospectively. I would want to assess baseline characteristics, and then follow up maybe every 6 months or every year and have them complete questionnaires asking for the above outcomes. Because this isn’t randomized, it will be subject to a fair amount of bias, especially because whether or not a state/city has a diversion policy in place is likely an indicator of their other social policies, so it will be difficult to isolate the effect of this policy.