Drug courts, mental health courts, veterans courts, and domestic violence courts have accumulated decades of evidence suggesting they reduce recidivism and improve outcomes for participants. Policymakers across the political spectrum point to them as proof that criminal justice can be both more humane and more effective.

Yet despite this evidence base, problem-solving courts remain stubbornly marginal. They handle a small fraction of eligible cases, operate as boutique programs in jurisdictions that adopt them, and frequently fail to reproduce pilot results when expanded. The gap between demonstrated efficacy and system-wide adoption has persisted for more than thirty years.

Understanding why successful models resist scaling reveals something important about criminal justice reform itself. The obstacles are not primarily ideological or financial—they are structural features of how court systems are organized, resourced, and evaluated. Examining these barriers clarifies what genuine scaling would require, and why incremental expansion often produces watered-down programs that underperform their origins.

Problem-solving courts operate on a fundamentally different infrastructure than traditional criminal courts. A typical drug court requires dedicated judges, prosecutors, and defense attorneys who commit to staying with cases for twelve to eighteen months. It requires treatment providers, case managers, drug testing capacity, and sanctions options that range from incentives to graduated responses.

This coordination is not incidental—it is the mechanism through which these courts produce results. The weekly status hearings, the team staffings before each court session, the real-time information sharing between treatment providers and the bench: these are the active ingredients. Stripping them away to reduce costs does not produce a cheaper version of the same program. It produces something different, usually resembling traditional probation with extra paperwork.

The resource calculus also extends beyond the court itself. Problem-solving courts depend on community-based treatment capacity, stable housing resources, employment services, and behavioral health infrastructure. In jurisdictions where these supports are thin or absent, courts cannot simply be established by judicial order. The ecosystem must be built alongside the court, which requires sustained investment across agencies that typically do not coordinate their budgets.

This explains why pilot programs so often thrive in carefully selected jurisdictions with strong partnerships, then falter when transplanted. The visible court is the tip of an iceberg of coordination that is invisible in evaluation reports but indispensable to outcomes.

Takeaway

A program's visible features are rarely what makes it work. When replication fails, the problem usually lies in the unglamorous infrastructure that supported the original.

Criminal court systems are organized around throughput. Judges are evaluated, implicitly or explicitly, on case disposition rates. Prosecutors manage massive caseloads by processing guilty pleas efficiently. Public defenders operate under crushing volume that leaves minutes per client. The dominant logic is speed, and the dominant metric is closure.

Problem-solving courts invert this logic. They keep cases open rather than closing them. They require judges to see the same participants repeatedly, sometimes weekly, for a year or more. They demand individualized attention that cannot be routinized or delegated. In a system measured by velocity, these courts look like congestion.

This tension creates persistent pressure to narrow eligibility criteria, shorten program lengths, and reduce the frequency of judicial contact. Each of these accommodations responds to real system constraints, but each also erodes the program's therapeutic mechanism. A drug court that sees participants monthly instead of weekly, or graduates them in six months instead of fifteen, is no longer operating the intervention that produced the evidence base.

Scaling therefore encounters not merely resource limits but a fundamental incompatibility between problem-solving methodology and the production demands of high-volume courts. Expansion without protecting these courts from caseload pressure tends to produce programs that carry the name but not the substance.

Takeaway

Institutional metrics shape behavior more powerfully than stated intentions. Reforms that conflict with how success is measured will slowly deform until they align with it.

Implementation research on problem-solving courts consistently documents a phenomenon called program drift. As models expand beyond their original sites, elements that researchers identified as essential—intensive judicial monitoring, graduated sanctions and incentives, clinically appropriate treatment matching, team-based decision-making—are selectively dropped.

The drift is rarely intentional. Expanded programs adapt to local constraints, accommodate staff turnover, and adjust to political pressures. Each adjustment seems reasonable in isolation. But the cumulative effect is that third-generation programs often retain the structure of a problem-solving court while omitting most of its operative features. Evaluations then show diminished outcomes, which undermines political support for the broader model.

This pattern is exacerbated by how these courts are typically funded. Federal grants support startup costs but assume local absorption over time. Local budgets rarely sustain the training, quality assurance, and team development that maintain fidelity. Without ongoing investment in practice standards, the program gradually reverts toward the path of least institutional resistance.

The field has responded with best-practice standards, certification frameworks, and fidelity instruments. These tools help, but they cannot substitute for the organizational conditions that make fidelity possible: protected caseloads, committed personnel, and evaluation systems that measure practice quality rather than just case counts.

Takeaway

Scaling a program without scaling its supporting conditions does not expand success—it dilutes it until the evidence that justified expansion no longer applies.

Problem-solving courts illustrate a broader truth about criminal justice reform: effective interventions are rarely portable without their context. What looks like a court model is actually an integrated system of resources, relationships, and institutional protections that together produce results.

Genuine scaling requires more than replicating the visible elements. It requires building the infrastructure that made those elements work, protecting the program from the throughput pressures of conventional courts, and sustaining the investment that maintains fidelity over time.

Without these conditions, expansion produces the appearance of reform without its substance. The lesson is not that these courts cannot scale, but that scaling is itself a design problem—one that deserves the same rigor as the original innovation.