Law enforcement co-response has expanded rapidly as communities seek alternatives to arrest and ED use during mental health crises. Yet outcomes vary dramatically. In some systems, co-response reduces force and improves linkage; in others, it amplifies fear, escalates encounters, and increases involuntary holds. The difference is not intent—it is operational design. This article examines how to structure co-response so it supports crisis stabilisation and continuity rather than becoming enforcement-first by default. For related guidance, see Crisis Response, Stabilisation & Continuity of Care and Mental Health Service Models.
The core risk of co-response: authority drift
Co-response brings together professions with different mandates: law enforcement prioritises public safety and legal authority, while clinicians prioritise therapeutic engagement and least restrictive care. Without explicit design, authority drifts toward enforcement under stress. Psychologically, uniforms, weapons, and command presence increase threat perception for many individuals, particularly those with trauma histories.
Operationally, the key question is not whether police are present, but who decides what happens next—and based on what criteria.
Two explicit expectations for co-response programs
Expectation 1: Clear role separation and clinical lead on disposition
Funders and civil rights monitors increasingly expect co-response models to demonstrate that clinical staff lead care decisions when there is no imminent criminal threat. Programs must evidence role clarity and guardrails that prevent enforcement default.
Expectation 2: Measurable reduction in force, arrests, and involuntary holds
Co-response is justified only if it reduces harm. Oversight bodies expect data showing fewer use-of-force incidents, fewer arrests for behavioral health reasons, and improved diversion durability—not just joint attendance numbers.
Operational Example 1: Dispatch triage that limits co-response to appropriate scenarios
What happens in day-to-day delivery
Dispatch uses a structured triage to determine when co-response is necessary: credible violence risk, weapons present, or active criminal behavior linked to the crisis. For lower-risk calls, clinician-only mobile response is dispatched. When co-response is used, dispatch provides both parties with a shared brief outlining safety concerns, engagement preferences, and clinical goals. The expectation is set that the clinician leads engagement unless immediate safety threats emerge.
Why the practice exists (failure mode it addresses)
The failure mode is overuse of co-response. When police attend most crisis calls “just in case,” enforcement presence becomes normalized and escalatory. Triage exists to reserve co-response for situations where it genuinely adds safety.
What goes wrong if it is absent
Police attend low-risk crises unnecessarily, increasing fear and resistance. Encounters escalate, leading to involuntary holds or arrests that could have been avoided. Systems then see higher force incidents and community mistrust.
What observable outcome it produces
Evidence includes reduced proportion of crisis calls involving police, lower escalation rates in clinician-only responses, and more appropriate use of co-response. Audit trails show triage rationale and deployment decisions.
Operational Example 2: On-scene role discipline and decision authority
What happens in day-to-day delivery
On scene, officers position themselves to manage environmental safety rather than lead interaction. The clinician conducts engagement and assessment, narrating decisions and options aloud to maintain transparency. A clear decision protocol states that, absent imminent criminal threat, the clinician determines disposition (on-scene resolution, crisis facility, ED). Officers intervene only if safety thresholds are crossed.
Why the practice exists (failure mode it addresses)
The failure mode is mixed messaging and authority conflict. When officers and clinicians both direct the interaction, individuals receive contradictory cues, increasing agitation. Role discipline exists to reduce confusion and threat.
What goes wrong if it is absent
Officers default to command-and-control tactics, undermining therapeutic engagement. Clinicians lose authority, and decisions skew toward involuntary options. Operationally, co-response becomes indistinguishable from standard police response with a clinician present.
What observable outcome it produces
Programs can evidence higher on-scene resolution rates, fewer involuntary holds, and improved satisfaction among service users. Documentation shows clinician-led disposition with clear safety rationale.
Operational Example 3: Post-contact continuity that avoids enforcement re-entry
What happens in day-to-day delivery
After the encounter, the clinician—not law enforcement—initiates continuity planning: follow-up contact, appointment scheduling, and barrier resolution. Officers disengage unless there is an ongoing legal matter. High-risk cases are flagged for proactive follow-up to prevent future calls defaulting back to police. Data-sharing agreements support tracking repeat contacts involving law enforcement.
Why the practice exists (failure mode it addresses)
The failure mode is enforcement rebound: without follow-up, individuals re-enter crisis through 911, triggering another police response. Continuity workflows exist to break this cycle.
What goes wrong if it is absent
Police become the primary continuity mechanism by default. This increases criminalisation risk and strains law enforcement resources.
What observable outcome it produces
Evidence includes reduced repeat police-involved crisis calls, improved linkage to behavioral health services, and declining arrest rates for mental health-related incidents.
Governance: ensuring co-response remains a clinical asset
Leaders should monitor use-of-force incidents, arrest rates, involuntary holds, and repeat police contacts for co-response cases versus clinician-only responses. Joint training and after-action reviews should focus on authority drift and decision quality. When governance is explicit and data-driven, co-response can support safety without sacrificing dignity or continuity.