A model of the argument in Michael Powell's Atlantic essay 'The Cities That Said Yes to Drugs' (https://www.theatlantic.com/magazine/2026/09/city-drug-addiction-harm-reduction-policy/687967/). Radical harm reduction (injection sites, 'safer supply', paraphernalia hand-outs, non-enforcement) makes each act of drug use safer, but it removes every consequence and creates open-air drug markets that pull in users and dealers. The street-using population grows until TOTAL overdose deaths and community disorder rise even though the death risk PER USER has fallen. Enforcement paired with a firm push into treatment does the opposite. One stock of street drug users in a Seattle-sized city (~750k residents); one reinforcing loop (the open drug market attracts more users); two balancing loops (deaths and exits to treatment). The drug supply turns from heroin to fentanyl over 2015-2018 in every scenario, so deaths rise everywhere at first, as they did in reality; policy decides whether the city ends up with a 2x or an 8x increase. Time runs 2015-2035: the decade of the fentanyl era plus a decade of policy consequences.
