Diversion programs redirect defendants away from traditional prosecution and incarceration toward treatment, community service, educational programming, or supervised release agreements. They operate at multiple points in the system: pre-arrest, pre-charge, post-charge, and post-conviction. The variety of program types makes definitive summary statistics difficult. The consistency of the findings across that variety is remarkable.
Pre-arrest diversion, where law enforcement officers connect individuals in crisis directly to treatment instead of booking them, has been studied most extensively through CAHOOTS in Eugene, Oregon, and the STAR program in Denver. CAHOOTS handled 24,000 calls in 2019, fewer than 1 percent of which required police backup. Denver's STAR program completed its first six months handling 748 calls with zero criminal charges. Law enforcement, mental health providers, and the people who encountered these programs were broadly positive. They worked because they connected the right response to the presenting need: clinical intervention for clinical problems.
Drug courts, the oldest and most extensively studied form of diversion, have a research base spanning over 30 years. A RAND meta-analysis found that drug court participation reduced drug use by 27 percent and recidivism by 26 percent compared to traditional prosecution. The fiscal analysis is consistent across jurisdictions: drug courts cost more upfront per participant than traditional prosecution because they involve intensive supervision and treatment. They cost far less when reincarceration rates are incorporated. The RAND estimate of net savings runs between $5,000 and $12,000 per participant.
Pre-trial diversion agreements, which offer defendants the opportunity to avoid prosecution by completing conditions, typically work well for first-time offenders charged with nonviolent offenses. Completion rates in well-designed programs run between 70 and 85 percent. Participants who complete the program avoid a criminal record, maintaining access to employment and housing that a conviction would foreclose. The return on that investment compounds over years of tax contributions, reduced public benefit usage, and reduced future justice system contact.
What diversion programs cannot do is replace appropriate accountability for serious offenses. This limit is sometimes used to argue against diversion as a concept, but it confuses the tool with the misapplication of it. Surgery is not ineffective because it cannot treat a cold. Diversion is designed for the specific population it serves.
The systemic argument for expansion is straightforward. Roughly 40 percent of people incarcerated in state prisons were convicted of nonviolent offenses. These are, collectively, the population most likely to respond to structured intervention and least likely to pose serious public safety risks during a supervised community sentence. Keeping them incarcerated is not primarily a public safety strategy. It is an expense. A very large one.
The political space for diversion expansion has widened considerably over the past decade, partly because the fiscal argument crosses ideological lines. Republican-led states including Texas, Georgia, and Utah have expanded diversion capacity explicitly as a cost-containment strategy. The outcomes have been consistent with what the research predicted.