Mental health courts are often presented as one of the justice system’s great compromises. They promise accountability without cruelty and treatment instead of punishment. For politicians and court administrators, they offer something even more valuable: a way to respond to visible suffering without confronting the deeper structures that produce repeated system contact.
The underlying assumption is simple: if people with serious mental illness cycle through jails and prisons, then the cycle must be driven primarily by untreated psychiatric need. Address the illness, and recidivism should fall.
A comprehensive review of mental health interventions in criminal justice settings complicates that assumption in an uncomfortable way: like many things, they only work when they’re done right.
The evidence shows that while mental health courts and diversion programs may improve access to treatment and short-term stability, they do not reliably reduce recidivism when operating in isolation from housing, employment, and broader structural supports.
What the Study Examined
Study: Preventing Criminal Recidivism Through Mental Health Interventions
Source: Systematic review of court-based and diversion interventions
Publication: PMC7280932
The authors reviewed a broad range of criminal justice interventions targeting people with serious mental illness, including mental health courts, jail diversion programs, specialty probation, and mandated treatment models.
Rather than focusing on a single jurisdiction or program, the review synthesized findings across multiple studies, settings, and program designs to evaluate whether these interventions actually achieved one of their primary stated goals: reducing recidivism.
The review distinguished between outcomes related to treatment engagement and outcomes related to system cycling, an important distinction that is often blurred in policy discussions.
What the Data Showed
Across studies, mental health interventions demonstrated mixed and often modest effects on recidivism. Some programs were associated with short-term reductions in arrests or jail days while participants were actively engaged. Those effects frequently weakened or disappeared once court supervision ended.
By contrast, improvements in treatment access and service utilization were more consistent. Participants were more likely to attend appointments, comply with treatment plans, and maintain contact with providers while under court oversight.
The central finding is not that mental health courts fail at treatment access. It is that treatment access alone does not reliably translate into reduced recidivism; there are so many other facets that are way too often ignored. By not providing people basic needs, mental health will quickly deteriorate. It’s an ecosystem that requires careful balance.
The Finding People Will Misuse, and the Finding We Actually Need
The misuse
It would be easy, and dangerous, to misread this evidence as an argument against providing mental health care to justice-involved people.
The authors do not support that conclusion. Mental illness is prevalent in jails and prisons. Treatment improves safety, functioning, and dignity. Those outcomes matter regardless of their effect on recidivism metrics.
The real takeaway
What this review actually shows is that mental health courts are being asked to do work they are structurally incapable of doing alone.
Courts can mandate treatment attendance. They cannot guarantee housing, employment, income stability, or protection from the surveillance dynamics that drive technical violations and returns to custody.
Why This Result Makes Sense
The findings align with what is already known about how recidivism is produced.
- Treatment compliance does not equal life stability
Someone can attend therapy and still be homeless, unemployed, or under constant supervision pressure. - Court involvement adds surveillance even when it adds care
Mental health courts often combine services with closer monitoring, increasing the risk of violations unrelated to new criminal behavior. - Recidivism is shaped by structural exposure, not diagnosis alone
Housing insecurity, poverty, substance use relapse, and enforcement intensity exert constant pressure that treatment alone cannot offset.
Mental illness matters. It is not the sole engine of the revolving door.
What This Means for Risk and Accountability Narratives
Mental health courts allow the system to frame recidivism as a clinical problem rather than a policy outcome. When participants return to custody, failure is attributed to noncompliance, relapse, or lack of insight.
This framing obscures a harder truth.
If the intervention addresses symptoms but leaves underlying conditions intact, recurrence should not be surprising. It is predictable.
Calling this accountability does not change the mechanism producing return.
What This Reveals About Societal Failure
Mental health courts persist because they are politically attractive. They signal compassion without requiring investment in housing, wages, or community infrastructure.
They shift responsibility away from legislatures and onto individuals, providers, and courts, even as structural conditions remain unchanged.
The evidence suggests that recidivism among people with serious mental illness is less about untreated diagnosis and more about how policy choices shape daily survival under supervision.
My Policy Read
Mental health treatment is a legal and human imperative. It should never be optional. But we have to strike balance; treatment only should not be it. And it cannot be used as the system’s alibi.
If policymakers want to reduce recidivism, mental health courts must be paired with guaranteed housing, employment pathways, reduced supervision tripwires, and continuity of care beyond court control.
Without those supports, mental health courts will continue to do what the evidence shows they already do: improve access to services while leaving the revolving door largely intact.
Read The Study:
Preventing Criminal Recidivism Through Mental Health Interventions
https://pmc.ncbi.nlm.nih.gov/articles/PMC7280932/
