988 / 911

When someone calls 911 about a mental health crisis, the dispatcher has historically had three choices: police, fire, or EMS. None of them is a mental health professional. The call goes to whichever category the dispatcher judges least wrong, which is usually police.

Mental health dispatch integration changes what happens at that decision point. A licensed clinician — either working inside the 911 call center or reachable through a direct transfer — takes the call. The clinician assesses the situation, de-escalates by phone, connects the caller to resources, and decides whether any field response is needed at all. Most calls end there. The clinician handles it. Nothing is dispatched.

The 911 system was established in 1968 around three response categories: police, fire, and EMS. It was not designed for behavioral health calls. A person in psychiatric distress, a family member calling about a child in crisis, a neighbor reporting someone in public distress — these calls enter 911, and 911 routes them to whichever of three options the dispatcher judges least wrong.

Monica Johnson, then the national director of 988 and behavioral health at SAMHSA, described the gap: “911 wasn’t set up to address mental health needs. Either callers end up in a frenetic emergency room, waiting for hours and sometimes days to get care, or they end up interacting with law enforcement, which can lead to tragedy or trauma.”

Cities that consider mental health dispatch integration typically already have something. They have 911. They may have 988. They have dispatcher training. Some have mobile crisis teams. The question is whether any of that solves the routing problem — and in most cases, it doesn’t.

The LEAP/Center for American Progress/NYU Policing Project joint report identified the routing problem as the gap that existing tools leave open: each tool adds value, but none places clinical judgment inside the dispatch decision itself.

Not every call that enters 911 is appropriate for clinical routing. Not every call that seems appropriate is. Programs with the clearest documented outcomes have defined specific eligibility criteria — specific enough for dispatchers and clinicians to apply consistently, flexible enough for the ambiguity that characterizes most behavioral health calls.

The evidence base for mental health dispatch integration is consistent in direction and limited in verification. No randomized controlled trial or independent evaluation with comparison groups has specifically studied the embedded dispatch clinician model. The outcome figures that programs report — Austin’s 90%+ diversion rate, Dakota County’s 83% resolution without police, Sacramento’s 90%+ phone resolution — are program-reported, not independently verified. That distinction matters when these numbers are cited in budget hearings and a skeptic checks the source.

What follows maps what the evidence shows, at what tier, and where the gaps are.

Mental health dispatch integration exists in some form in every U.S. region — from Minnesota counties operating under a state mandate to Houston’s decade-old embedded clinician model to Sacramento’s 911/988 interoperability system. The programs for which public documentation is available are concentrated in mid-sized to large cities; rural implementation exists through Minnesota’s statewide mandate but with minimal published documentation at the county level.

The LEAP/CAP/NYU joint report documented the distinction that matters most for comparing programs: systems using embedded clinical professionals produce materially higher capture rates than systems relying on dispatcher training or routing protocols alone.

National polling shows broad, consistent public support for the concept of routing mental health calls to mental health professionals. The documented positions of police chiefs, police union leaders, and elected officials across the political spectrum add institutional weight to that polling. Where disagreement enters is typically over implementation details — who pays, how it connects to existing police work, whether specific designs are adequate.

Mental health dispatch integration touches more organizational actors than its apparent simplicity suggests. Placing a clinician inside a 911 call center changes the working environment of dispatchers, the call volume of police, the staffing demands of mental health organizations, and the accountability structure of everyone connected to the routing decision. The LEAP/CAP/NYU joint report identified the institutional relationships at 911 centers, police departments, and mental health organizations as the primary implementation variables that determine whether a program reaches its designed call volume.

Mental health dispatch integration has documented failure modes, structural fragilities, and honest gaps in what the current evidence can tell decision-makers. Four failures are documented with named programs and identified causes.

No two cities have built mental health dispatch integration identically. The variation reflects genuine design choices: where clinical capacity lives, who makes routing decisions, what calls are eligible, and how the call center connects to the broader behavioral health system. What follows maps those choices — the options available, the tradeoffs each produces, and what the documented programs chose.

Decision 1: Which Dispatch Model?

This is the foundational design choice, and the one with the most documented outcome variation. The LEAP/Center for American Progress/NYU Policing Project joint report identified three principal models, with a fourth emerging from practice.

Model 1A — 911 Call Center-Led Dispatch (Dispatcher-Routing). The 911 call-taker decides whether a call routes to the behavioral health team using a decision tree, specific call codes, or enhanced training. No clinical staff are present in the 911 center. The call-taker makes the routing decision and dispatches through the same CAD system used for police.

Used in: Denver (early STAR dispatch), Olympia, Chicago, Cincinnati, Dayton, St. Petersburg, San Francisco.

The advantage is efficiency: no transfer, no warm handoff, no second phone conversation. The disadvantage is that it places the clinical routing judgment with a call-taker who is not clinically trained, and the cultural change required — getting dispatchers to route away from police under uncertainty — is the hardest part of implementation.

The LEAP/CAP/NYU report found dispatchers in this model “significantly hesitant” to route calls away from police, and the Wayne State University 2025 study found that 911 professionals coded just over half of mental health-component calls correctly. Columbus’s early Right Response Unit, using this model, captured approximately 10% of eligible calls — compared to 80–90% in embedded clinician programs.

Some of the most experienced dispatchers in documented programs developed creative workarounds — reframing questions, stretching definitions, or entering call types after they’d already decided who to send — to route calls to clinical teams despite protocol constraints rather than because of them. This pattern, documented in the LEAP/CAP/NYU report, illustrates the fundamental limitation of the training-only model: the routing decision’s quality depends on individual dispatcher initiative rather than structural clinical judgment.

Model 1B — Warm Transfer to External Crisis Line (Interoperability). 911 dispatchers identify eligible calls and do a warm transfer to a 988 counselor or external crisis line. The dispatcher stays on briefly to hand off context, then exits.

Used in: King County (Seattle), Sacramento County.

King County’s Dustin Freeman described the model: “When someone is in a mental health crisis, there’s no crime happening, and 988 has more resources to help people.” South King County 911 dispatchers transfer calls where callers don’t have weapons and aren’t actively threatening harm. King County invested in both technical interoperability systems and cultural relationship-building between dispatcher communities and 988 call center staff — recognizing that the technical connection is necessary but not sufficient.

Sacramento County’s 911/988 dispatch technology enables smooth handoffs; 988 counselors resolve over 90% of transferred calls by phone, referring only a small fraction to in-person teams.

The structural advantage is institutional independence: the mental health organization maintains its own staff and protocols without being inside police communications infrastructure. The structural disadvantage is that the model depends on the dispatcher recognizing a call as behavioral health before transferring it — the same hesitancy problem as Model 1A, without clinical staff present to override it. Additionally, the warm transfer adds time and can lose callers who don’t want to be transferred or who disconnect during the handoff.

The LEAP/CAP/NYU report documented that Model 1B produces higher capture rates than Model 1A (because transfers go to a clinical professional, not a social worker with limited call-handling capacity) but lower capture rates than Model 1C (because the dispatcher still makes the initial identification decision).

Model 1C — Embedded Professional-Led Dispatch (Co-located Clinician). A licensed mental health professional sits physically inside the 911 call center. When a call-taker recognizes a potential crisis call, they do a warm handoff — staying on the line while introducing the caller to the embedded clinician. The clinician handles the call clinically: assessing, de-escalating, resolving by phone, connecting to resources, or dispatching a field team. In more advanced implementations (Austin, Durham CCD), the clinician can also monitor the call stream proactively and flag eligible calls.

Used in: Austin, Houston, Dakota County, Durham CCD unit, Philadelphia, Louisville, St. Louis, Des Moines, Waco.

Austin puts 24 of its 71 Integral Care crisis employees directly in the dispatch center, working around the clock. Houston’s CCD program, launched in 2015, was one of the earliest embedded models and has processed tens of thousands of calls over a decade. Dakota County staffs its 911 center with mental health professionals from the county’s Crisis Response Unit specifically for this function.

The advantage is clinical quality: the person making the routing decision has clinical training. The 80–90% of calls that can be resolved over the phone by a clinician never need a field response at all. Austin’s dispatch-embedded clinicians resolved more than 90% of their FY2025 calls without a police response. Dakota County’s model produces 83% resolution without law enforcement field response. The LEAP/CAP/NYU report identifies this as the model with the highest documented capture rates: the clinician makes the routing decision, eliminating the hesitancy and capability gap that limits dispatcher-only models.

The disadvantage is cost and the warm handoff itself, which adds time and can lose callers who don’t want to be transferred. The 24-hour staffing requirement — to cover all shifts — is the single largest budget driver in these programs. Austin employs 24 embedded dispatch clinicians specifically for call-center work; achieving that staffing pool from an already scarce licensed clinician workforce takes deliberate investment and time.

Model 1F — Dual Mechanism (Automated CAD + Embedded Clinician). Some jurisdictions have added automated computer-aided dispatch flagging to embedded clinical staff. The CAD system flags call types that match behavioral health criteria; the embedded clinician handles flagged calls. The dual mechanism catches calls that either system would miss independently.

Durham’s Crisis Call Diversion unit uses this approach — automated behavioral health call type flagging plus CCD clinicians embedded in the 911 center who handle the flagged calls. The LEAP/CAP/NYU report documented that dual-mechanism systems catch calls that single-mechanism systems miss, because automated flagging catches calls where the dispatcher wouldn’t have made a behavioral health routing decision, and the clinician catches calls that the automated system might not flag based on call type codes but that present behavioral health patterns in the caller’s description.

The documented outcome gradient. Programs producing the highest documented phone-resolution rates — Austin at 90%+, Dakota County at 83%, Sacramento at 90%+, Nebraska at 96%+ — use embedded or interoperability models with clinical staff involved in the routing or resolution decision. Columbus’s dispatcher-training-only design produced approximately 10% capture. All figures are program-reported, not independently verified.

The gap between 10% and 90% is not a gap in program quality or commitment. It is a structural difference in who makes the routing decision. When a dispatcher without clinical training decides whether a call is behavioral health, many calls with behavioral health components don’t get identified. When a licensed clinician in the call center makes that assessment, the recognition rate is fundamentally higher — because the clinician has the training to hear behavioral health indicators the dispatcher is not trained to notice.

Decision 2: Who Are the Embedded Staff?

The qualifications and organizational affiliation of the clinical staff in the call center shape what they can do, what they can access, and what the program costs.

Licensed clinicians (master’s degree or above). Austin, Houston, Dakota County, and Philadelphia all embed licensed clinicians — licensed professional counselors, licensed social workers, marriage and family therapists. Licensed clinicians can provide clinical assessment, use evidence-based de-escalation techniques, identify diagnostic presentations, and in some organizational structures, bill Medicaid for qualifying encounters.

Kedra Priest, Practice Administrator of Crisis Services at Integral Care, noted the specific professional demand: dispatch clinicians must be comfortable handling crises at the phone level without the full information a field visit provides. This is a distinct professional profile, not simply any clinical professional.

Behavioral health navigators and trained counselors. Some programs — Baltimore, some 988-integrated models — use staff with counseling training short of independent clinical licensure. These staff can provide de-escalation, resource connection, and warm handoffs but cannot perform formal clinical assessment or bill Medicaid for most encounter types.

The staffing tradeoff. Licensed clinicians cost more and are scarcer. The HRSA designates more than half of U.S. counties as behavioral health workforce shortage areas. Programs that require licensed staff for 24/7 coverage are staffing multiple overlapping shifts from an already constrained pool. Some programs use licensed clinicians for the dispatch layer and less credentialed navigators for follow-up functions, separating the clinical decision at dispatch from the connection work that follows.

Decision 3: Where Does the Program Live Organizationally?

The institutional home of the embedded clinical staff shapes their accountability, their access to clinical records and billing infrastructure, and their relationship to 911 leadership.

Inside a mental health authority (contracted model). Austin’s EMCOT operates through Integral Care, Travis County’s designated local mental health authority. Integral Care employees work inside Austin’s 911 call center under a contract. Integral Care’s clinical records system means dispatch clinicians can check whether a caller is already connected to services — a meaningful clinical function unavailable in models without records access. The contract creates a shared accountability structure: clinicians are Integral Care employees, but they work inside a city operation.

Dominick Nutter, Austin’s Emergency Communications Director, explained the operational logic: “A 911 call comes in and we have our standard case entry questions… as they start explaining the situation, our team will decide if it’s something that is appropriate for police, fire or mental health response… The big thing is, does the person have a weapon and is there any danger? In that case, then mental health response wouldn’t be appropriate. If it’s something that’s appropriate for mental health response, then the call is transferred to a counselor.”

The advantage: the mental health organization brings clinical infrastructure, billing capability, clinical supervision, and workforce. The vulnerability: the contract is subject to annual city appropriation. Austin’s Prop Q failure — which cut funding by $1 million — operated directly through this structure.

Inside a standalone city department. Durham’s Crisis Call Diversion unit is part of the Community Safety Department, a standalone city department whose director reports directly to the city manager. CCD clinicians are city employees, accountable through city management, with a departmental budget rather than a contract line.

The advantage: institutional standing comparable to police and fire. The Department’s budget — approximately $5 million for the full HEART program including CCD — is a departmental line item rather than a program contract. The disadvantage: building a full organizational infrastructure — HR classifications, clinical supervision, training programs — from scratch inside city government requires significant investment before the program is operational.

Inside the 911 center itself. Philadelphia’s embedded behavioral health navigators are employees of the 911 center, reporting to emergency communications leadership. Michele Mallette, Philadelphia’s Chief of Staff, described the operational goal: “as we divert calls over to the clinicians, it helps with our improved response time. If responders aren’t going to those calls, they can focus on other calls where the public needs them.”

The advantage: accountability sits squarely inside the dispatch organization; no interagency seam to manage. The disadvantage: behavioral health professionals employed by a public safety dispatch center may have less clinical autonomy and less connection to the broader behavioral health service system for referrals and records access.

Multi-agency collaboration. Houston’s CCD operates through a collaboration among HPD, Houston Fire, and the Harris Center for Mental Health and IDD. The telecounselors are not 911 employees but work at the Houston Emergency Center full-time under this shared structure. Wayne Young, CEO of The Harris Center, explained: “This collaboration not only assists deputies in making informed decisions but also significantly reduces the role of law enforcement in mental health crises.”

The multi-agency structure preserves the clinical organization’s records infrastructure and billing capability while placing staff inside the 911 environment. The challenge is governance: when something goes wrong, or when the program needs to expand or change, multiple organizations with different priorities and accountability structures must reach agreement.

Brent Anderson, Dakota County 911 operations director, identified the institutional relationship as foundational: “I think because we’ve had such good, open communication, and we talk things through, I think that’s the foundation of why things are working.”

Decision 4: Coverage Hours

Coverage hours are the operational variable with the most direct effect on a program’s actual reach.

Austin, Durham’s CCD unit, and Albuquerque’s Triage Specialists have achieved 24/7 coverage. Houston’s CCD has operated 6 AM to 10 PM since 2015, leaving a documented overnight gap for more than a decade. Columbus Right Response Unit operated 8-hour daily coverage for years before adding embedded clinical staff. Durham’s CCD had no overnight coverage through mid-2025; Durham City Manager Bo Ferguson estimated the 24/7 expansion would push HEART’s annual budget from approximately $5 million to $6.5 million.

The workforce math of 24/7 coverage is direct: programs that cover 12 hours per day need approximately twice as many licensed clinical staff to cover all shifts. Programs that have achieved 24/7 — Austin with 24 embedded dispatch clinicians, Albuquerque with Triage Specialists on all shifts — made that staffing investment incrementally across multiple budget cycles. Austin moved from initial embedded clinicians in 2019 to full 24/7 coverage in 2023 — a four-year ramp that required sustained budget commitment across city administrations.

The funding mechanism behind overnight gaps is structural, not incidental. Angela Kimball of Inseparable identified the core problem: police and fire departments are funded for capacity — officers are paid whether or not a call arrives in their shift. Crisis programs are funded for encounters — Medicaid reimburses billable clinical contacts but does not cover a licensed clinician sitting in the 911 center between calls at 2 AM. A program that bills Medicaid for active encounters still needs separate funding for the readiness infrastructure that makes 24/7 coverage sustainable. Programs without a dedicated budget line for that readiness cost find organic expansion to overnight hours difficult even when their daytime program is performing well.

The behavioral health calls that arrive overnight are not less severe. Substance use crises, acute psychiatric distress, and welfare checks do not follow business hours. A program that stops at 10 PM leaves its most resource-scarce hours — when alternative clinical services are also closed — covered by the police default that existed before the program launched. Houston’s decade-long overnight gap illustrates that strong daytime performance does not create automatic pressure to expand hours; it requires a specific budget decision.

Decision 5: Safety Screen Design

Every documented program has a safety screen upstream of clinical routing. The design variation is in who applies it and how — and that variation has direct consequences for capture rates and responder safety.

Dispatcher applies the screen first. Austin’s model requires the initial call-taker to complete a safety screen before any transfer to the EMCOT clinician. Dominick Nutter, Austin’s Emergency Communications Director, described the protocol: “A 911 call comes in and we have our standard case entry questions… as they start explaining the situation, our team will decide if it’s something that is appropriate for police, fire, or mental health response… The big thing is, does the person have a weapon and is there any danger? In that case, then mental health response wouldn’t be appropriate. If it’s something that’s appropriate for mental health response, then the call is transferred to a counselor. One of our telecommunicators will talk to the counselor first so the counselor is aware of the situation, and then they will bring in the caller.”

The advantage: clear accountability; the dispatcher handles the safety determination before any clinical involvement. The risk: dispatcher judgment about what constitutes an active threat varies. Some ambiguous calls may be screened out that an embedded clinician could have handled safely. The safety screen decision is a judgment call about an unknown situation, and dispatcher conservatism — defaulting to police for anything unclear — can limit program reach on exactly the ambiguous calls where clinical expertise matters most.

Clinician assesses safety in real time. Programs where clinicians monitor the call stream proactively — Austin’s more mature model, Durham’s CCD — allow the clinician to participate in the safety assessment alongside the dispatcher. The clinician can hear the caller directly and apply clinical judgment to the safety question rather than relying solely on the dispatcher’s interpretation of the caller’s words. This requires the clinician to be available and engaged with incoming calls continuously, which is possible when call volume is managed and staffing is adequate.

Protocol-driven exclusions. Durham’s HEART program excludes “anything with a weapon or threat of violence” as a defined protocol rule rather than a dispatcher judgment call. Albuquerque’s ACS uses a 1–5 priority scale, with Priority 1 calls (active safety threats) remaining with police regardless of any behavioral health component. These protocol-driven exclusions reduce the judgment burden on individual dispatchers and create consistent, auditable routing decisions.

The self-reported safety record across documented programs is strong. Durham HEART reported that responders felt safe 99% of the time across nearly 25,000 calls. Minneapolis reported no serious responder injuries in more than 16,000 calls. Albuquerque ACS required police backup on less than 1% of FY2025 calls. All of these figures are program-reported, and “felt safe” is a different measure than formal safety incident tracking. As programs expand to 24/7 operations and broaden their eligible call types, the safety screen will be tested on a wider range of presentations than early programs encountered.

Decision 6: Eligible Call Type Definition

Programs define eligibility in two primary ways: explicit lists or dynamic clinical assessment.

Explicit call-type lists. Austin began with a narrow defined list — suicidal ideation, parent-child conflicts, psychosis/paranoia presentations — and has expanded it over time as the program built credibility and dispatchers developed confidence. Durham’s 911 center routes approximately 1,500 different call natures using pre-coded question protocols; a defined subset routes to CCD clinicians.

Explicit lists give dispatchers clear decision points and allow programs to track capture rates against eligible volume. The limitation: real calls don’t always match listed categories. Baltimore’s fidelity gap — call-takers diverted only when callers explicitly stated a diagnosis — illustrates what happens when eligibility depends on caller vocabulary rather than clinical judgment.

Dynamic clinical assessment. Programs with clinicians physically present in the 911 center can assess eligibility more dynamically. The clinician hears what the dispatcher hears, flags calls the dispatcher might not have identified as behavioral health, and makes eligibility determinations based on clinical judgment rather than a pre-coded category. This is one mechanism behind the higher capture rates in embedded models.

Decision 7: Connection to Records and Systems

A clinician embedded in a 911 center is meaningfully more effective with access to clinical records than without.

Austin’s EMCOT clinicians “check records to see if a person already is connected to resources” as part of their standard call process — made possible because EMCOT operates within Integral Care’s clinical records system, which contains case management histories for Integral Care’s existing service population. This function allows the clinician to see whether a frequent caller has an active case manager, a scheduled appointment, or a recent crisis contact — information that changes the clinical response. A caller with an existing therapist at a known agency gets a different response than a caller with no documented service connection.

The RTI International Durham call analysis documents why records access matters: approximately 15% of Durham 911 calls were behavioral health or quality-of-life calls, while only 1–2% were CAD-coded as mental health at dispatch — because the presenting complaint doesn’t trigger a behavioral health flag even when behavioral health is the underlying driver. A clinician who can look up a caller’s history may recognize a welfare check as a repeat contact from a high-frequency caller with a documented mental health history. That recognition changes the clinical response and often changes whether any field dispatch is needed at all.

Programs where clinical staff are employees of the 911 center rather than a clinical organization may lack records access entirely. The call reaches a clinician who has no information beyond what the dispatcher has — the same information constraint the dispatcher was working with. The Houston multi-agency collaboration, involving the Harris Center for Mental Health and IDD as a partner organization, preserves the clinical organization’s records infrastructure alongside the 911 center placement.

Same-day follow-up calls — used by Austin — require records infrastructure to be useful. The clinician making the follow-up call needs to know what happened on the original call, what resources were offered, and whether the person accepted them. Without records, the follow-up call starts from scratch.

Decision 8: Political and Implementation Dynamics

The most technically sound program designs have failed at launch because of political or operational factors. Design decisions don’t exist only on paper — they shape the political feasibility of the program in the environment it operates in.

Building dispatcher and police support before launch. Durham’s Community Safety Director Ryan Smith presented data directly to patrol officers before HEART launched, answering their questions about what cases would and would not be diverted. Austin’s clinicians built a feedback relationship with dispatchers before the formal program was operational — the “coach reviewing game film” feedback loop began informally before it was institutionalized. Albuquerque’s APD Chief Harold Medina characterized ACS at launch as “already helping to free up our officers so they can respond to high-priority calls.”

The LEAP/CAP/NYU report identified the dispatcher buy-in sequence that documented programs share: leadership alignment first, followed by dispatcher training, followed by relationship-building between dispatch staff and clinical staff through ride-alongs, joint debriefs, and feedback sessions on how specific diverted calls turned out. In Dayton, dispatcher comfort with the Mediation Response Unit improved significantly after dispatchers joined the team on ride-alongs — a pattern the report describes as more durable than classroom training alone. The mechanism is straightforward: dispatchers who have watched clinical staff handle a call understand what the program does. Dispatchers who have only been told what the program does remain uncertain about edge cases.

Framing for the political environment. Austin Police Association President Michael Bullock testified before City Council: “It’s time that we work towards getting law enforcement out of mental health. We have never claimed to be the experts, but yet we have been charged with the responsibility of responding to mental health crisis.” He characterized the fourth-option model as “positive steps towards identifying better ways to divert these calls.” Albuquerque’s ACS launch explicitly described the program as allowing police to “focus on violent crime” and EMS to focus on “urgent, life-threatening situations.” Durham City Manager Bo Ferguson: “It was always critically important for me that this not be perceived as something that we were taking away from the police department.”

These framings reflect a documented pattern across the programs with the strongest police buy-in: dispatch integration positioned as freeing police to focus on their core function encountered less organized resistance than programs positioned as alternatives to police. The operational result is identical in both framings — fewer calls routed to police — but the political dynamics of the description shape whether police organizations become institutional supporters or institutional opponents.

Building measurement infrastructure. Programs that built public measurement from the start — Baltimore’s quarterly public dashboard under consent decree oversight, Durham’s monthly Community Safety Department data snapshots, Albuquerque’s quarterly departmental reports — created accountability structures that serve the program’s political sustainability. When a program can demonstrate its outcomes publicly, skeptics must engage with documented data rather than with anecdotes about what might go wrong. Austin’s same-day follow-up call is a specific design choice that also produces continuity data — whether the person was reached and what their status was hours after the initial call — that the program uses to demonstrate impact beyond diversion rates.

Design Comparison: Two Ends of the Spectrum

Programs documenting the highest diversion rates — Austin at 90%+, Dakota County at 83%, Sacramento at 90%+, Nebraska at 96%+ — sit at or near the high-capacity end. Columbus at 10% sat at the minimal end before adding embedded clinical staff. All figures are program-reported.

The distinction between these models is not a matter of program quality or commitment — it is a matter of who makes the routing decision and with what information. When a clinician with clinical training and records access makes the routing decision, more eligible calls are captured and more are resolved by phone. When a dispatcher with enhanced training makes the routing decision, fewer calls are captured and the structural hesitancy documented in the LEAP/CAP/NYU report shapes the outcome.

Decision 9: Policy Pathway — City-by-City vs. State Mandate

Every jurisdiction documented above built dispatch integration through local initiative — a city or county decision to invest in clinical capacity at the 911 center. Minnesota chose a different design: statewide legislative mandate.

The Minnesota model. Minnesota’s 2021 statute requires dispatchers to refer mental health calls to trained crisis responders where available. By 2023, 85 of the state’s 87 counties were compliant — implementing the dispatch integration function without waiting for city-by-city political processes to mature.

The mandate design has specific advantages. Political feasibility at the city or county level varies dramatically; a city council that wants to launch a program but faces organized opposition can point to state law as the obligation. The mandate creates floor-level consistency: every county must address the routing problem, even if the specific implementation varies. And because compliance is tracked statewide, the policy produces data at scale — thousands of mental health routing decisions across 85 counties — that voluntary adoption cannot replicate.

The mandate design has specific limitations. State law can require that calls be referred to crisis responders “where available” — but it cannot guarantee that trained crisis responders exist in every county. Rural counties that pass through the routing decision to a crisis line operated from a distant urban center are technically compliant but operationally limited. And the mandate produces compliance variation: Dakota County’s highly documented, high-performing embedded model sits within the same compliance category as smaller counties operating with far less clinical infrastructure.

The Virginia model. Virginia funds crisis services through telecommunications fees — a dedicated revenue stream modeled on how 911 itself is funded. The result is approximately 20-second average wait times and a 92.3% call answer rate for mental health crisis calls across the state, according to The Center Square reporting. The funding mechanism is the design innovation: by attaching crisis service funding to the same fee structure as 911, Virginia created a sustainable revenue stream outside the annual general fund appropriation process that has destabilized programs in Austin and Baltimore.

What the state-level models offer local decision-makers. A city or county considering dispatch integration in a state with a mandate can focus on implementation quality — what embedded model to build, what staffing to invest in, how to connect to records systems — rather than the prior political question of whether to build anything at all. A city or county in a state with dedicated funding can access that funding stream rather than depending entirely on general fund appropriation. Both state-level design choices change the operating environment for local programs.

Mental health dispatch integration programs draw on multiple funding streams, and no single stream covers the full cost of clinical readiness at a 911 call center. The structural gap — between what encounter-based funding covers and what 24/7 clinical staffing requires — is documented across the field and explains most coverage-hours limitations.

Angela Kimball of Inseparable identified the mechanism: police and fire departments are funded for capacity — officers are paid whether or not a call arrives in their shift. Crisis programs are funded for encounters — Medicaid reimburses billable clinical contacts but does not cover a licensed clinician sitting in the 911 center between calls at 3 AM. Programs that rely primarily on encounter-based revenue cannot sustain the overhead necessary for round-the-clock clinical readiness.

The most effective public framing for mental health dispatch integration has not been reform framing. It has been public safety efficiency framing: police are not trained for this, they are wasting time on calls they cannot resolve well, and putting clinical expertise at the dispatch point fixes that problem. This framing works across political contexts because it does not require anyone to concede anything about police. It requires acknowledging that police are doing work they are not equipped for and should not be expected to do.