How To Create A Community Response To A Mental Health Crisis

The modern mental health crisis response problem has a clear origin point in American communities: the deinstitutionalization movement of the 1960s and 1970s. The closure of large psychiatric institutions, motivated by a combination of civil libertarian concern for patient rights, fiscal conservatism, and genuine clinical critique of the institutional model, released hundreds of thousands of people with serious mental illness into communities that had not built the infrastructure to support them.

The Community Mental Health Centers Act of 1963 promised a national network of community mental health centers that would provide the care previously delivered in institutions. That network was never fully funded. The result was a massive gap: people with serious mental illness, many of them unhoused, cycling through emergency rooms, jails, and streets rather than receiving consistent treatment and community support.

Into this gap, by default, fell law enforcement. Police became the de facto first responders to mental health crises not because they were equipped for it but because they were available 24 hours a day and had no way to decline a call. The consequences have been severe and extensively documented. People in mental health crisis are 16 times more likely to be killed during a police encounter than other civilians. Many more are arrested, jailed briefly, and returned to the same conditions — a cycle that treats crisis as a law enforcement problem rather than a health problem.

The White Bird Clinic founded CAHOOTS in Eugene, Oregon in 1989, making it the longest-running alternative mental health response model in the United States. Understanding how it actually works operationally is essential for communities considering building their own systems.

CAHOOTS teams consist of a medic (EMT or nurse) and a crisis worker with mental health training. They are dispatched through the standard 911 system — the dispatcher routes calls to CAHOOTS rather than police when the caller's described situation fits CAHOOTS criteria: mental health crisis, welfare check, substance use call, public intoxication, or similar non-violent situations. CAHOOTS teams have no arrest authority and carry no weapons. They can call for police backup if a situation turns violent or a person is an imminent danger.

The outcomes documented over three decades are consistent. In 2019, CAHOOTS handled 24,000 calls — about 17% of all calls to Eugene's combined emergency dispatch. Of those, fewer than 150 (less than 1%) required police backup. The cost per CAHOOTS call is substantially lower than the cost of a police response, and dramatically lower than the cost of an emergency room visit or a night in jail, which are the most common outcomes when crisis response is handled by law enforcement.

What CAHOOTS has not solved — and does not claim to have solved — is the longer-term problem of connecting people to consistent mental health treatment. Its crisis workers are skilled at de-escalating acute situations and making referrals; the fragmentation of the mental health system means those referrals often lead to months-long wait lists or services that don't match the person's actual situation.

Denver's Support Team Assisted Response (STAR) program, launched in 2020, represents a second-generation design informed by CAHOOTS. STAR uses a co-responder model: a licensed clinician and a paramedic ride together, dispatched specifically to mental health and substance use calls in designated geographic areas.

A peer-reviewed evaluation of STAR's first six months of operation found that it handled 748 calls without requiring police backup in any of them, and with zero citations or arrests. The program operated at a cost of approximately $1 million annually — a fraction of what police response to those calls would have cost.

What STAR did differently from CAHOOTS in its initial design: it operated only during daytime hours in its first phase, focusing on building a strong evidence base before expanding. This staged approach allowed the program to demonstrate effectiveness to skeptical city council members and police union representatives before requesting expanded resources. This strategic patience proved useful for the program's political sustainability.

Professional crisis response teams are necessary but not sufficient for a genuine community mental health response. The deeper work happens in the social infrastructure around crisis — the relationships, check-in systems, and mutual knowledge that allow communities to catch people before crisis becomes acute.

Peer Support Networks: The most cost-effective mental health support is often delivered by people with lived experience of mental health challenges who are trained as peer specialists. Peer support programs, which pay trained peers to provide support, advocacy, and connection to services for others going through mental health challenges, have strong evidence bases and can be embedded in neighborhoods rather than clinical settings. The challenge is funding: peer support is often last to be funded and first to be cut in budget cycles.

Natural Community Connectors: Every neighborhood has people who naturally know what is happening with their neighbors — hairdressers, barbershop owners, religious leaders, longtime corner store owners, building superintendents. Mental health first aid training for these natural connectors is among the highest-leverage investments a community can make. Mental Health First Aid, an 8-hour evidence-based training program, teaches non-clinicians to recognize crisis signs and connect people to help. Training 50 natural connectors in a neighborhood creates a distributed early-warning network that no professional system can replicate.

Warm Lines: Crisis hotlines handle acute crisis; warm lines handle the earlier stage — when someone is struggling but not yet in crisis. Warm lines are staffed by trained peers available to talk without the clinical stakes of a crisis line. Several states and regions operate warm lines; many communities do not, leaving a gap in the prevention-intervention continuum.

Housing-First Approaches: Any serious community mental health response must grapple with the intersection of mental health and housing. A substantial proportion of people who cycle through crisis response systems are unhoused. Crisis response that returns people to unsheltered conditions — regardless of how skillfully the acute crisis was handled — is treating a symptom while the cause continues. Communities that have made significant investments in Housing First programs (providing stable housing before requiring sobriety or treatment compliance) consistently show reductions in psychiatric crisis calls, emergency room utilization, and incarceration for mental health-related incidents.

The politics of transitioning from police-dominant mental health response to community-centered response are difficult. Police unions often oppose diverting calls away from officers, framing it as both a safety risk and a staffing challenge. Some community members are skeptical that unarmed responders can handle volatile situations safely.

Building political will requires a combination of evidence presentation, coalition development, and strategic pilot design. On evidence: the documented outcomes from CAHOOTS, STAR, and similar programs provide strong arguments, but they must be translated into local terms — what would it mean for this city's specific call volumes, costs, and outcomes?

Coalition development requires bringing together mental health advocates, faith communities, family members of people who have experienced mental health crises, and progressive law enforcement leadership (sheriffs and police chiefs who recognize that crisis response is not a core law enforcement competency and who are burdened by having to provide it). These coalitions often find unexpected allies: emergency room physicians who see the revolving door of crisis patients, jail administrators overwhelmed by mentally ill inmates, fiscal conservatives who recognize the extreme cost inefficiency of using emergency rooms and jails as de facto mental health facilities.

Strategic pilot design — starting small, in a defined geographic area or call type category, with rigorous outcome measurement — allows advocates to build an evidence base within their own community rather than relying solely on external examples. The STAR model's deliberate first-phase scope limitation reflected this logic.

A growing number of cities are experimenting with what might be called the peer responder model — mental health crisis teams composed primarily of people with lived experience of mental health crisis, incarceration, or homelessness, rather than clinicians. Organizations like the Peer Wellness Center in San Francisco and various programs emerging from defund-the-police advocacy have argued that credentialed clinical training is less important in crisis response than lived experience and relational trust.

The evidence on peer responder programs is less developed than on clinician-based programs like CAHOOTS, but the theory of change is compelling: people in crisis often respond better to someone who can credibly say "I have been where you are" than to a clinician, however skilled. Peer responders also build trust within communities that are systematically distrustful of clinical and governmental institutions — Black communities, Indigenous communities, communities with histories of coercive psychiatric treatment.

The question of how to train, supervise, and support peer responders is not yet fully resolved. The emotional demands of crisis response are substantial, and peer responders — particularly those whose lived experience includes recent trauma — require strong supervision and support structures to sustain the work without being retraumatized.

There is a deeper argument for community mental health response that goes beyond the pragmatic evidence about outcomes and costs. How a community treats its most vulnerable members in their most vulnerable moments is a direct expression of the community's actual values — not the values it claims in mission statements but the values it enacts under pressure.

A community that defaults to sending armed officers to people in psychiatric crisis is expressing, implicitly, that crisis is a threat to be managed rather than a human experience to be supported. A community that builds genuine crisis response infrastructure — peer supporters, trained community connectors, unarmed responders, housing, ongoing care — is expressing that belonging means something real, that the social contract extends to the moment of breakdown.

This is not a marginal policy choice. It is a community identity choice. And communities that have made it consistently report something that goes beyond the measurable outcomes: a different relationship between residents and the institutions that are supposed to serve them — more trust, less fear, more willingness to reach out for help before crisis becomes catastrophe.

The work of building a community mental health response is technical, political, and relational. The technical work is designing dispatch systems and training protocols. The political work is building coalitions and winning budget fights. The relational work — which must underpin both — is building a community that actually knows its members well enough to notice when someone is struggling.