America Is Obsessed With Punishment—and It’s Making Us Less Safe

The Mental Health-Crime Connection

The relationship between untreated mental illness and criminal justice involvement is documented in decades of research. It is not a fringe position or an advocacy claim. The Bureau of Justice Statistics, the National Alliance on Mental Illness, and peer-reviewed public health literature converge on a consistent finding: people with untreated serious mental illness, substance use disorders, trauma histories, and developmental conditions that affect decision-making are overrepresented in jails, prisons, and court dockets at rates far exceeding their presence in the general population.

Approximately two in five people in U.S. jails and prisons have a history of mental illness. Substance use disorders, which are themselves classified as mental health conditions, affect an even larger share of incarcerated individuals. Post-traumatic stress disorder, major depression, schizophrenia, bipolar disorder, and conditions affecting impulse control and executive function are common features of the incarcerated population — not because mental illness causes crime in any simple causal sense, but because untreated mental illness intersects with poverty, housing instability, inadequate healthcare access, and exposure to violence in ways that increase the probability of criminal justice contact.

These conditions do not excuse criminal conduct. They explain it. And they point unmistakably toward interventions that incarceration cannot provide.

The system’s response to this reality has been largely to ignore it. Standard criminal processing does not begin with a mental health assessment. Jails and prisons are not designed as treatment environments. Sentences calibrated to punishment do not address the conditions that generated the conduct. And when incarcerated individuals are released — often without connection to services, housing, or ongoing care — the conditions that contributed to criminal justice involvement are still present, frequently worsened by the incarceration itself.

Beyond Punishment: A Public Health Approach

When crime is understood as a public health problem rather than a moral failure alone, different responses become available and more defensible.

Diversion programs redirect individuals from standard criminal prosecution into treatment, supervision, and community support. The Vera Institute has documented their operation and outcomes across jurisdictions: when properly structured, diversion programs reduce rearrest rates, connect participants to services they would not otherwise access, and reduce long-term system costs. They do not eliminate accountability; they change its form from incarceration to treatment compliance and community supervision.

Mental health courts represent a more structured intervention within the criminal justice system itself. Operating in most states including Michigan, mental health courts handle cases involving defendants with diagnosed mental illness through a specialized process that links court supervision to treatment participation. A 2020 meta-analysis in Psychiatric Services found significant reductions in recidivism associated with mental health court participation, with effects persisting over follow-up periods of a year or more. The Bureau of Justice Assistance’s mental health courts program has expanded that model, and the evidence base supports continued expansion.

Early intervention is a third category with strong evidentiary support. Many serious mental health conditions first emerge during adolescence and early adulthood — precisely the period when many individuals first encounter the criminal justice system for lower-level offenses. Accessible mental health care at that stage of development can interrupt a trajectory that might otherwise lead from minor offenses to more serious criminal conduct and eventual incarceration. The system currently lacks the infrastructure to deliver that care at scale. The consequence is predictable.

The Economic Case

The public safety argument for a public health approach is straightforward. The economic argument reinforces it from a different direction.

Per-inmate incarceration costs in the United States range from roughly $25,000 to over $60,000 per person per year in most states, with several high-cost jurisdictions — California, New York, Massachusetts — exceeding $100,000 annually. These figures cover direct operational costs; they do not account for downstream costs to families, communities, or public systems that absorb the effects of incarceration. Michigan’s annual corrections spending exceeds $2 billion.

Comprehensive community-based mental health treatment, including intensive outpatient services, supported housing, and case management, costs substantially less than those per-person incarceration figures even at the high end of the treatment cost range. The cost comparison is not close, and it becomes less close when the analysis includes recidivism: incarceration alone, without treatment of underlying conditions, produces high rates of return to the system. Each return generates another round of incarceration costs. A model that treats the condition is not just cheaper per episode. It is cheaper over time because it reduces the number of episodes.

There is also the cost to families. Parental incarceration is one of the most significant adverse childhood experiences documented in the research literature. Children with incarcerated parents face elevated risks of their own mental health problems, educational disruption, poverty, and eventual criminal justice involvement. The generational cost of a punishment-only approach to mental-illness-related crime is not confined to the individual being incarcerated. It extends to households and communities that absorb those costs for years after the sentence ends.

Mental Health, Policing, and the Point of First Contact

The criminal justice system’s encounter with people in mental health crisis typically begins not in a courtroom but on a street, in a home, or during a welfare check call. Research and advocacy organizations have documented that a significant share of police use-of-force incidents — including fatal encounters — involve individuals experiencing psychiatric emergencies. That concentration is not incidental. It reflects a structural reality: in most American jurisdictions, law enforcement is the default response to behavioral health crises, because no other infrastructure exists to respond.

Crisis Intervention Team programs train officers to recognize and de-escalate mental health crises rather than defaulting to force. Co-responder models pair law enforcement officers with mental health clinicians who can assess the situation and connect individuals to appropriate care. Both approaches have shown promise in reducing the probability that a mental health crisis becomes a use-of-force incident or a criminal charge. They represent the kind of systems-level change that requires sustained investment, not just policy language.

Challenges and What Honest Reform Requires

A public health framing of crime does not require abandoning accountability or minimizing the real harm that criminal conduct causes to victims and communities. It requires distinguishing between responses that reduce the probability of future harm and responses that feel punitive but do not. Incarceration for individuals whose criminal conduct stems from untreated serious mental illness often falls in the second category. It addresses the conduct without addressing the condition. The result is release into the same circumstances that produced the original offense, with the additional destabilization of a criminal record and a gap in whatever fragile social supports existed before.

Honest reform requires several things simultaneously. It requires accessible mental health care before criminal justice contact — in communities, in schools, in primary care settings. It requires diversion infrastructure at the point of first contact — trained first responders, co-responder programs, diversion pathways that do not require a felony charge as a prerequisite. It requires mental health assessment at every stage of criminal processing, from arrest through release planning. And it requires community-based support systems for individuals with mental health needs who are returning from incarceration, because release without connection to services is not a reentry plan. It is a countdown to the next arrest.

None of this is currently the default. In most jurisdictions, including most Michigan counties, the infrastructure for the front-end interventions described here is inadequate relative to the documented need. That gap is a policy choice, maintained by funding priorities that reflect a consistent preference for incarceration over treatment even when the evidence does not support that preference on either safety or cost grounds.