Forrest Stuart, Caylin Louis Moore, and Patrick Janulis examine San Francisco’s Urban Alchemy program in their 2026 PNAS Nexus article, “Civilianized safety practitioners reduce crime and disorder: Evidence from a quasiexperimental evaluation in San Francisco.” Using data from 2018 through 2023, they compared 41 intersections with practitioner deployments against 583 comparison intersections through a generalized synthetic control design. Unarmed workers provided a consistent presence, intervened in conflicts, and connected people to services. The analyses found reductions in total recorded incidents and behavioral health-related incidents. Drug and violent crime reductions were statistically significant when comparing daytime and nighttime changes, but not in the daytime analysis alone. Property crime showed no detectable change.
The study makes an important contribution to community-safety scholarship by evaluating a workforce whose authority rests on relationships rather than coercion. Its findings support serious consideration of paid civilian prevention roles, including opportunities for people with experience of incarceration and homelessness. Nevertheless, this was not a randomized experiment, and administrative records cannot fully distinguish reduced harm from changes in reporting or police contact. Municipalities should therefore pair implementation with independent evaluation, resident perspectives, and measures of service access.



Leave a Reply