Mechanism Design for Harm Reduction: Game Theory and Social Choice for Carceral MOUD and Recovery Housing

Individuals released from jails and prisons face extremely high risks of fatal overdose and reincarceration, yet many jurisdictions continue to underprovide medications for opioid use disorder (MOUD), recovery housing, and supervised consumption services. At the same time, recovery residences and diversion courts are expanding without a clear framework for institutional design. This paper develops a mechanism-design model of harm-reduction policy at the interface of criminal justice and community treatment. A public funder chooses a funding regime and certification rules, diversion judges set the stringency of supervision and treatment conditions, recovery residence providers decide whether to operate abstinence-only or MOUD-inclusive housing, and high-risk individuals choose whether to comply or relapse. The model yields a punitive equilibrium, supported by abstinence-only funding and strict conditions, and a harm-reduction equilibrium under MOUD-inclusive funding and flexible conditions. Using effect sizes from Rhode Island’s statewide corrections MOUD program, Massachusetts’ jail-based MOUD pilots, and recent recovery housing evaluations, we show that the harm-reduction equilibrium is Pareto-superior for funders, judges, providers, and high-severity residents, yet the punitive equilibrium can remain risk-dominant because of political and informational frictions. We then embed the game in a computational social choice framework: stakeholders hold multi-dimensional preferences over policy bundles—combinations of funding rules, certification standards, diversion guidelines, and overdose prevention interventions such as supervised consumption sites—and social choice is constrained by justice-based requirements that rule out policies generating avoidable lethal risk or systematic exclusion of MOUD patients from housing and treatment. The analysis characterizes which harm-reduction mechanisms are implementable as equilibrium outcomes of the institutional game while respecting these constrained social preferences, and it identifies simple instruments—MOUD-inclusive funding commitments, performance-based transparency, and structured diversion defaults—that can move jurisdictions from punitive to harm-reduction equilibria within existing legal constraints.