Does San Francisco’s Community Justice Center Reduce Criminal Recidivism?

Community courts are expressly oriented toward improving outcomes for offenders by addressing factors—like substance use disorders, mental health issues, and unemployment—that are often linked to criminal behavior; this is done by incorporating access to treatment and services within the criminal case management process itself. The community court model is a novel approach for processing criminal offenders. The goals of community courts are largely similar to those of traditional criminal courts (reducing crime, ensuring safety, and protecting due process rights), but the community court approach differs in two important ways. First, community courts are expressly oriented toward improving outcomes for offenders by addressing factors—like substance use disorders, mental health issues, and unemployment—that are often linked to criminal behavior; this is done by incorporating access to treatment and services within the criminal case management process itself. Second, community courts emphasize ties to a specific neighborhood within a city. In 2009, San Francisco opened a community court, the Community Justice Center (CJC), located in and designed to serve the Tenderloin and adjacent neighborhoods. This report examines whether the CJC reduces the risk of criminal recidivism (as proxied by rearrest) when compared to more traditional approaches for addressing arrestees. We use a differences-indifferences (DD) design that exploits temporal and geographic variation in CJC eligibility. For the four police districts that touch the CJC catchment area, we examine one-year rearrest rates among those arrested for eligible offenses inside and outside the catchment area both before and after the CJC opened. Our analysis finds that the probability of rearrest for those originally arrested outside the CJC catchment area increased over time, while the probability of rearrest for those originally arrested inside the CJC decreased over time. After controlling for a number of arrestee-level factors (including criminal history) as well as month-and police district–level fixed effects, the DD estimator from our preferred models ranges from –8.2 to –7.1 percentage points, which corresponds to an 8.9 percent to 10.3 percent reduction in the probability of being rearrested within one year. These findings support the hypothesis that the CJC reduces criminal recidivism and are robust to a number of sensitivity analyses. Given these results, it is plausible that the CJC may have produced other benefits to participants, their intimates, and the community. Future research could examine how CJC participation influences treatment outcomes as well as emergency room visits. It is also imperative to incorporate information about the costs of …