Human Rights Watch Maps Non-Police Crisis Response Models as Federal Support Faces Uncertain Future

Human Rights Watch released a comprehensive examination of non-police mental health crisis response programs across the United States on March 30, arriving at a moment when the federal infrastructure supporting these community-based alternatives faces restructuring that could threaten their sustainability.

The report, developed in partnership with New York Lawyers for the Public Interest and the Center for Racial and Disability Justice, identified more than 150 programs nationwide that remove police as primary responders to mental health crises. From this landscape, researchers profiled eight programs embodying rights-based approaches: peer-led teams, consent-centered interventions, and crisis workers trained in de-escalation rather than enforcement.

The Intersection of Crisis, Race, and Disability

The urgency driving these alternatives is documented in stark data: people with serious mental health conditions are 16 times more likely than those without such conditions to be killed by police. Black individuals showing signs of acute psychiatric distress are shot and killed by law enforcement at a rate ten times higher than Black people in the general population.

These disparities reflect not only failures of police training—officers receive on average fewer than 15 hours of mental health and de-escalation instruction—but also the fundamental mismatch between law enforcement approaches and the needs of people experiencing mental health crises. Police tactics prioritizing authority and compliance can escalate rather than resolve situations where someone is confused, disoriented, or unable to follow commands.

The report examines how diagnostic disparities compound these dangers. Black students are three times more likely than non-Black peers to be labeled with "emotional disturbance," while Black and Latinx youth under 18 face double the rate of psychotic disorder diagnoses compared to white children. These patterns persist into adulthood, where provider bias, misinterpretation of trauma, and lack of cultural competency contribute to over-diagnosis and misdiagnosis of Black and Latinx individuals at rates substantially higher than white populations.

Even when people of color access mental health services, they encounter systems where hospitals serving predominantly Black neighborhoods operate with fewer resources, fewer specialists, and higher documented rates of medical negligence. While facing barriers to adequate voluntary care, Black individuals are simultaneously overrepresented in involuntary psychiatric holds: in New York, Black people comprise 17.7 percent of the state population but represent 38 percent of Involuntary Outpatient Commitment recipients.

Eight Models of Rights-Based Response

The programs highlighted operate without armed officers, police-style uniforms, or coercive tactics. Mental Health First in Oakland deploys three-person volunteer teams—a nurse or EMT, a mental health expert, and a security person to handle potential police encounters—operating Fridays and Saturdays from 2 p.m. to 2 a.m. Run by the Anti-Police Terror Project, a Black-led coalition focused on eradicating police violence in communities of color, the program responds to calls through a dedicated phone line.

Oakland's second featured program, MACRO (Mobile Assistance Community Response of Oakland), takes a different structure: operated by the Oakland Fire Department since 2022, two-person teams pair an EMT with a community intervention specialist, available daily from 6:30 a.m. to 8:30 p.m.

Cambridge Holistic Emergency Alternative Response Team represents grassroots organizing translating directly into service provision. Launched in 2021, HEART currently runs a peer-led support line offering emotional assistance and resource navigation on Tuesdays and Thursdays, 10 a.m. to 4 p.m., with plans to expand into mobile crisis response.

San Diego County's Mobile Crisis Response Teams demonstrate how programs can scale to 24/7 coverage: beginning as a pilot in January 2021, MCRT expanded countywide by December and achieved round-the-clock operations by April 2022. Operated by Telecare Corporation under county contract, each team includes a licensed clinician, master's-level case manager, and certified peer recovery specialist. Calls arrive through the San Diego Access and Crisis Line or 988.

New Orleans launched its Mobile Crisis Intervention Unit in June 2023, contracting with nonprofit Resources for Human Development to staff teams typically composed of a peer support specialist and behavioral health professional supervised by a licensed clinician. Operating 24/7, MCIU responds to non-violent, weaponless mental health calls routed through 911 dispatch, with police able to request assistance when appropriate.

The Elm City COMPASS program in New Haven emerged from community-driven development, partnering with nonprofit Continuum of Care for staffing and operations. Teams pairing social workers with peer support specialists operate three shifts spanning 8 a.m. to 3 a.m., responding to non-emergency mental health crises and providing follow-up care through New Haven's 911 system.

Denver's STAR (Support Team Assisted Response) program began in June 2020 as a partnership between the Department of Public Health and Environment, WellPower community mental health center, and Denver Health and Hospital Authority. Recognized as a "fourth responder" alongside police, fire, and EMS, teams pair WellPower clinicians with DHHA paramedics or EMTs, operating daily 6 a.m. to 10 p.m. Community members can specifically request STAR-only response when calling Denver's 911 or non-emergency lines.

Franklin County, Ohio's Netcare Access Community Mobile Team launched in 2023 serving Columbus and surrounding areas. Funded primarily by the county's Alcohol, Drug, and Mental Health Board, teams of licensed clinicians and certified peer recovery specialists operate 24/7, accessible through 988 or a direct program line.

Growth Amid Structural Limitations

The expansion of these models reflects accelerating momentum: as of 2024, all 50 states reported having mobile crisis teams, totaling 2,448 programs with 12 states planning to add at least 89 more in 2025. Federal support has come primarily through a five percent set-aside of the Community Mental Health Block Grant awarded by the Substance Abuse and Mental Health Services Administration, supplemented by state funds, Medicaid, and private insurance where reimbursement structures permit.

Yet significant gaps persist. Only 34 states provide services statewide. Only 28 states operate mobile crisis teams 24 hours a day, seven days a week. Most crisis responses still rely on police-led models or co-responder teams where clinicians accompany law enforcement—structures that embed officers at the center of crisis intervention and retain their authority to use force.

Research on Crisis Intervention Team programs, which provide officers approximately 40 hours of specialized mental health training beyond the standard 15 hours, shows limited impact on reducing use of force, arrest rates, or injury outcomes for individuals experiencing crises. A 2014 systematic review found that while CIT training may improve officers' knowledge and attitudes, effects on arrests, force deployment, and safety remain inconsistent and limited.

The mere presence of armed officers can trigger heightened fear and distress, particularly for individuals confused or unable to comply with commands. In co-responder models, clinicians typically deploy only after police determine a scene presents no threat of violence, reinforcing law enforcement as primary authority. Emergency medical personnel often wait for police to declare a scene safe before providing care, creating windows where harmful interactions occur.

Federal Restructuring Threatens Infrastructure

The Trump administration's early 2025 announcement that SAMHSA would dissolve and merge into the newly created Administration for a Healthy America raises questions about sustainability. The proposed consolidation would combine the Mental Health Block Grant and Substance Abuse Prevention and Treatment Block Grant into a single Behavioral Health Innovation Block Grant.

Although administration officials assert that "critical programs" will continue under AHA, the loss of SAMHSA as a standalone agency could undermine federal financial support for non-carceral, rights-based mental health responses. The July 24, 2025, Executive Order on "Ending Crime and Disorder on America's Streets" signals interest in expanding coercive and punitive mental health interventions at federal, state, and local levels—a directional shift potentially at odds with voluntary, consent-centered crisis models.

These federal developments arrive as programs navigate operational challenges independent of political climate. Many operate limited hours, citing staffing and funding constraints. Training programs require development, funding mechanisms remain unclear as Medicare and Medicaid reimbursement varies by state, and private insurance coverage proves inconsistent. Some states have established peer specialist certification; others lack formal credentialing structures.

Integrating new team members into existing clinical workflows demands coordination: psychiatrists accustomed to independent practice need support adapting to collaborative models, electronic health record systems may lack documentation pathways for community health worker activities, rural clinics operating on razor-thin margins struggle to justify adding staff without guaranteed reimbursement.

The Case for Legal Mandate

The Americans with Disabilities Act may provide legal leverage independent of federal funding streams. The Department of Justice found the City of Minneapolis and Minneapolis Police Department discriminated by dispatching police to mental health calls, noting that "many behavioral health-related calls for service do not require a police response, but MPD responds to the majority of those calls, and that response is often harmful and ineffective."

The DOJ stressed that because mental health calls "often involve no violence, weapon, or immediate threat of harm," such situations "could be safely resolved with a behavioral health response, such as a mobile crisis team." Under ADA requirements, governments must make "reasonable modifications" to policies and practices if necessary to avoid discrimination based on disability—potentially including creation and expansion of first-responder entities staffed by mental health practitioners and peers.

This legal framework operates independently from Congressional appropriations or executive branch reorganizations, offering a path for communities to compel alternatives even as federal support infrastructure undergoes restructuring.

Implementation Questions Remain

The Human Rights Watch report explicitly states it does not endorse specific programs or evaluate their effectiveness in practice. Researchers did not conduct independent audits, interview service users, or verify programs' claims and outcomes. Instead, the analysis relied on publicly available materials, stated commitments to non-carceral approaches, and interviews with program staff and advocates.

Many profiled programs launched after 2019 and remain in early implementation stages. Longer-term studies under formal protocols are necessary to assess sustained performance and whether programs follow stated principles over time.

Effectiveness metrics themselves present challenges: measuring crisis de-escalation, voluntary engagement, participant satisfaction, and connections to ongoing support requires different frameworks than traditional emergency response metrics. Programs emphasizing consent may see lower immediate resolution rates than coercive interventions while producing better long-term outcomes—but demonstrating those results demands longitudinal data collection most programs lack resources to conduct.

The report arrives during what researchers describe as "significant political backlash against rights-based approaches to mental health care," heightening urgency around identifying and defending existing alternatives. Whether the profiled programs represent replicable models or context-specific successes dependent on particular combinations of local government support, community organizing, and available funding remains unclear.

What Consent-Centered Response Requires

Beyond removing police from initial response, rights-based crisis intervention demands structural changes to how systems define and respond to distress. Traditional approaches frame crisis as pathological deviation requiring professional management, potentially overlooking how the person experiencing crisis understands their own needs and what forms of support they find meaningful.

Programs emphasizing self-determination must navigate tensions between respecting individual autonomy and addressing situations where someone's behavior raises safety concerns. When does de-escalation require intervention despite lack of consent? How do teams trained in voluntary engagement respond when family members, bystanders, or emergency callers demand action the individual in crisis refuses?

These questions have no universal answers. The Gerstein Crisis Centre in Toronto, examined in an earlier Human Rights Watch report, offers voluntary crisis beds where people can stay up to five days with no diagnosis or treatment required—but even that model operates within legal frameworks permitting involuntary hospitalization under certain circumstances, creating ongoing tensions between stated principles and practical authority.

Cultural competence adds another layer of complexity. International human rights standards require health services be "respectful of the culture of individuals, minorities, peoples and communities, sensitive to gender and life-cycle requirements." Achieving this demands staff self-aware of their own cultural biases, workforce representative of communities served, and ongoing collaboration with diverse populations rather than one-time community input sessions.

Harm reduction principles—acknowledging the dignity of people who use drugs, supporting them without judgment, providing tools to reduce health and social risks without requiring abstinence—integrate into some crisis models but face resistance in jurisdictions where substance use remains heavily criminalized or where funding sources prohibit serving people actively using.

The Road From 150 Programs to Systemic Change

The gap between 150 identified programs and nationwide transformation is vast. Most programs concentrate in urban areas; rural communities face particular challenges given smaller populations, geographic distances, and limited mental health infrastructure. Fairbanks, Alaska implemented non-police crisis teams despite a population just over 30,000, but replication depends on local political will, available funding, and capacity to recruit and retain qualified staff in isolated areas.

California leads with 28 programs, followed by New York with 15, reflecting both larger populations and state policy environments more receptive to alternatives. Other states have single programs or none, leaving millions of Americans in jurisdictions where police remain sole option for mental health emergencies.

Funding fragmentation complicates sustainability. Some programs draw from city or county general funds, others from state allocations, federal grants, private foundations, or voter-approved sales taxes dedicated to mental health services. Reliance on multiple sources provides some resilience against any single stream drying up but creates administrative complexity and uncertainty when grants expire or political priorities shift.

The Human Rights Watch researchers emphasize their survey was not comprehensive and the landscape continually evolves. Programs launch, expand, contract, or fold based on funding availability, political support, and operational challenges. Tracking this shifting terrain exceeds the capacity of any single organization, leaving significant uncertainty about total program numbers, coverage, and outcomes.

Accessible and transparent data management—identified as a rights-based design component—remains aspiration more than reality for most programs. Publishing call volume, response times, disposition of cases, demographic breakdowns of who receives service, and follow-up outcomes enables accountability and allows other jurisdictions to learn from experience, but many programs lack systems to collect such data comprehensively, let alone share it publicly.

Transformation or Tokenism?

The growth from minimal non-police crisis capacity in 2019 to more than 150 programs by 2023 represents meaningful expansion, particularly given entrenched institutional resistance to changing emergency response systems. Whether this constitutes fundamental transformation or marginal adjustment to overwhelmingly police-centered crisis intervention depends partly on metrics chosen and partly on time horizons.

If the measure is number of people experiencing mental health crises who receive support from peers and clinicians rather than armed officers, the programs represent substantial progress. If the measure is structural change to power dynamics determining who gets labeled as crisis, who gets coerced into treatment, and whose distress receives compassionate response versus punitive intervention, the programs exist as small counter-examples within systems largely unchanged.

The report's release timing—as federal infrastructure supporting these alternatives faces reorganization—frames the programs as fragile achievements potentially vulnerable to political winds rather than stable features of American crisis response. Yet legal mandates under the ADA operate independently from federal funding, and community organizing sustaining programs like Mental Health First or Cambridge HEART persists regardless of executive branch priorities.

For the 122 million Americans living in designated behavioral health shortage areas, for Black individuals facing disparities at every stage from diagnosis through involuntary commitment, for people with disabilities killed by police at rates 16 times higher than non-disabled populations, these programs offer something traditional systems have not: response treating mental health crisis as human experience requiring empathy and voluntary support rather than threat requiring containment.

Whether that represents pathway to liberation, as the report's title suggests, or incremental improvement within fundamentally unchanged systems remains contested. What seems clear is that the programs profiled provide evidence alternatives to police-centered crisis response can operate at scale, across diverse communities, serving thousands of people annually—not as theoretical possibility but as documented reality.

The challenge ahead is not proving such models can exist. They do. The challenge is determining whether they will be defended, expanded, and integrated into emergency response infrastructure, or whether they will remain isolated exceptions while millions continue receiving crisis intervention designed for compliance rather than care.