The United States has approximately 2 million people in jails and prisons, roughly 25 percent of the world's incarcerated population despite representing about 4 percent of the world's people. This scale of incarceration is historically recent: incarceration rates quadrupled between 1975 and 2008, driven by policy choices including mandatory minimums, three-strikes laws, truth-in-sentencing provisions, and the war on drugs. These choices were made with the explicit goal of reducing crime through the deterrent and incapacitating effects of incarceration. Evaluating whether they achieved that goal, and at what cost, requires engaging the evidence honestly.
What Incarceration Does and Does Not Do for Public Safety
The research on incarceration and crime is nuanced. Credible evidence shows that incarceration reduces crime through incapacitation at low levels, keeping high-rate, violent offenders incapacitated prevents the crimes they would otherwise commit while incarcerated. But the returns diminish sharply as incarceration levels rise. At current U.S. incarceration levels, the marginal incapacitation effect of adding another person to prison is small because the additional person is increasingly likely to be a low-rate, nonviolent offender whose incapacitation removes few crimes from the community.
The evidence for deterrence, the idea that the threat of incarceration reduces crime by making potential offenders weigh consequences, is similarly complicated. The deterrent effect of certainty of apprehension is reasonably well-supported. The deterrent effect of sentence length, the mechanism behind mandatory minimums, is not. Research consistently finds that longer sentences produce little additional deterrent effect beyond what shorter sentences produce, because most potential offenders neither accurately know the sentence lengths they face nor discount future consequences enough for marginal increases in sentence length to affect their behavior.
The Collateral Costs of Mass Incarceration
The direct costs of incarceration, approximately $35,000 per person per year at the state level, are substantial but not the full accounting. Incarceration disrupts families, removes parents from children, and concentrates those effects in specific communities with high incarceration rates in ways that produce measurable intergenerational harm. Children of incarcerated parents face elevated risk of poverty, educational disruption, behavioral health problems, and incarceration themselves. Communities with high incarceration rates experience labor market disruption, reduced civic participation, and the demographic consequences of removing large proportions of working-age adults for extended periods.
Formerly incarcerated individuals face significant barriers to reintegration: restricted access to public housing, professional licenses, student loans, and many employment categories. These restrictions reduce economic opportunity in ways that increase recidivism risk, producing a cycle that serves neither individual rehabilitation nor public safety goals. Criminal records that follow individuals for decades for offenses committed years or decades earlier make the incapacitation logic of incarceration extend far beyond the period of confinement itself.
What the Evidence Supports
The evidence most consistently supports alternatives that produce comparable or better recidivism outcomes at significantly lower cost. Drug courts, which divert people with substance use disorders from incarceration to treatment with judicial supervision, reduce recidivism by approximately 8 to 14 percent in well-designed evaluations and at a fraction of incarceration cost. Community supervision, including probation and parole, is less costly than incarceration and shows comparable recidivism outcomes for most nonviolent offenders. Cognitive behavioral therapy delivered in correctional or community settings reduces recidivism by 10 to 30 percent in meta-analyses. Investing in these alternatives, and in the education, employment, and housing interventions that reduce recidivism over the long term, would address public safety more cost-effectively than continued incarceration expansion.
