Quantifying CAHOOTS: Mobile crisis response program diverts and prevents police response

Mobile Crisis Response Programs (MCRPs) have emerged as alternatives to police for emergency mental and behavioral health calls, yet their effectiveness in diverting calls from police remains contested. CAHOOTS, the United States' longest-running MCRP, has been subject to widely varying ad hoc diversion rate estimates, from 3% to 20%. We use a systematic difference-in-differences design to estimate the causal effects of CAHOOTS on police call volume by introducing three estimators. The Substitution Diversion Rate (SDR) quantifies MCRP responses that directly replace police responses and estimate it as 18% for CAHOOTS. Our results show in addition to substituting for police response, CAHOOTS prevented police calls. The Prevention-adjusted Diversion Rate (PDR) accounts for this, and we estimate at 23% for CAHOOTS. Finally, the Overlapping-mandate Diversion Rate (ODR) quantifies the proportion of mutually answerable calls handled by an MCRP. We estimate CAHOOTS’ ODR at 50%. Our findings provide evidence that MCRPs significantly reduce police response call volume through both direct substitution and incident prevention.