Criminal Justice Reform: What Research Shows About What Works

The United States has the highest incarceration rate among wealthy nations, imprisoning approximately 2 million people at any given time. The scale of American incarceration is a product of policy choices made primarily in the 1970s, 1980s, and 1990s, including mandatory minimum sentences, three-strikes laws, truth-in-sentencing requirements, and dramatic expansion of drug offense prosecution. The consequences of these choices, for individuals, families, communities, and public budgets, have been substantial and are increasingly well-documented. Understanding what criminal justice reform should prioritize requires engaging with the evidence on what actually drives crime rates and what interventions reduce recidivism, the rate at which formerly incarcerated individuals return to criminal behavior. The evidence on these questions has grown considerably over the past two decades, and it challenges several assumptions that drove policy expansion during the incarceration build-up. Deterrence, the idea that harsh punishments reduce crime by increasing the cost of criminal behavior, has strong theoretical support but a more complicated empirical record. Research on the marginal deterrent effect of longer sentences, compared to moderate sentences, is consistently weak. Studies of mandatory minimums, which lengthened sentences for specific offenses, do not show corresponding reductions in those crimes. Research consistently finds that the certainty of punishment, the probability of being caught and convicted, is a more powerful deterrent than its severity. This finding suggests that investments in effective policing and prosecution may reduce crime more effectively than mandatory sentence enhancements. Incarceration itself can increase recidivism by exposing individuals to criminal networks in prison, disrupting employment and family relationships, reducing legitimate economic prospects through criminal records, and removing people from communities for periods long enough to destroy their social capital. Longer sentences do not consistently reduce recidivism more than shorter sentences in well-designed studies. This creates a policy problem: if incarceration increases recidivism and longer sentences do not reduce it more than shorter ones, the justification for the current scale of incarceration cannot be built primarily on public safety grounds. Community-based supervision, including probation and parole, has better recidivism outcomes than incarceration for many categories of offenders. Well-implemented cognitive behavioral programs delivered in community settings have consistent evidence of reducing recidivism. Employment support, housing assistance, and substance use treatment reduce recidivism more reliably than supervision intensity alone. Concentrated supervision without supportive services produces fewer benefits than more holistic approaches. Drug policy is central to the incarceration story. Drug offenses account for a substantial share of federal and state prison populations, and the racial disparities in drug law enforcement have produced the dramatic racial disproportionality in incarceration rates. Black Americans are incarcerated at rates five times those of white Americans, a disparity that exceeds what can be explained by offense rates and reflects documented racial disparities in policing, prosecution, and sentencing. Reform of drug policy, including decriminalization of possession, expansion of diversion programs, and reduction of mandatory minimums for drug offenses, has accumulated support across ideological lines as evidence on effectiveness has grown. Diversion programs, which redirect people from the criminal justice system into treatment and services, show consistent benefits for specific populations, particularly those with mental health conditions and substance use disorders. Mental health courts, drug courts, and crisis intervention programs have accumulated positive evidence for reducing recidivism, improving treatment engagement, and producing cost savings compared to standard prosecution and incarceration. These programs reflect the understanding that the criminal justice system is often dealing with problems, addiction, mental illness, poverty, trauma, that it is not equipped to solve through punishment. Reentry programs, designed to support people leaving incarceration in securing housing, employment, and services, have mixed evidence. The most effective programs provide comprehensive, individualized support and maintain connections during the transition period. Employment programs with employer partnerships and skills training show the most consistent effects on employment and recidivism. Housing programs that provide stable housing immediately upon release show substantial benefits for stability and reduced recidivism. The fiscal argument for reform is significant. The average annual cost of incarceration in the United States exceeds 35,000 dollars per person in many states and is substantially higher in others. Investment in prevention, treatment, and community-based alternatives can produce comparable or better public safety outcomes at substantially lower cost. Several states have restructured sentencing policies based partly on fiscal analyses showing that diversion and treatment are more cost-effective than incarceration for many offense categories. Evidence-based criminal justice reform is not a politically simple project. Crime and safety are emotionally charged issues, and policymakers face electoral incentives that do not always align with evidence-based approaches. But the evidence that the current scale of American incarceration is not justified by its public safety benefits, and that alternatives produce better outcomes at lower cost, is now substantial enough to ground serious reform conversations.