Mass Incarceration: Causes and Reforms

For most of its history, the United States incarcerated people at rates similar to other Western democracies. Beginning in the 1970s, that changed dramatically. Incarceration kept rising long after crime began to fall, and the country ended up locking up a larger share of its population than any other nation. How did it happen, and what has it cost the people inside, their families, and public safety?

In this TalksOnLaw interview, Joel Cohen sits down with Professor Rachel Barkow of NYU School of Law, a former member of the U.S. Sentencing Commission and author of Prisoners of Politics: Breaking the Cycle of Mass Incarceration. Barkow explains the scale of mass incarceration, its political and legal roots, and why it fails as a public safety strategy, then turns to the reforms that are working and the structural changes that could do more.

The Scale of the Problem

Barkow describes a system that grew far beyond any historical or international norm, with millions more people on probation or supervised release. The growth did not track crime. An initial rise responded to increasing violence, but when crime leveled off and fell, incarceration kept climbing. Asked whether no one took their foot off the gas, Barkow agrees and adds that the country also "bought a fleet of cars" by investing in police, prosecutors, and prisons that built constituencies for expansion.

Families and Communities

The reach is enormous: roughly one in three American adults has a criminal record, and about one in two has had a family member incarcerated. Incarceration can push already struggling families into poverty. It harms children's school performance and raises the odds of their own involvement in the system. Barkow compares it to family separations at the border: parents are sent hundreds of miles away, visits are rare, and phone calls are expensive because providers profit from families. The burden falls heavily on poor communities and communities of color, and in some cities about half of Black men are under some form of criminal justice supervision.

The Politics of "Tough on Crime"

Barkow traces the political shift to Barry Goldwater's law-and-order rhetoric in the 1960s. Richard Nixon and Ronald Reagan carried it forward, and eventually both parties embraced it. From the start, she says, it was racialized, with "urban crime" serving as a code for civil rights-era unrest. Local television news amplified the fear. Studies find that people who watch more local news are more punitive, whatever the crime rate where they live. A released defendant who commits a new crime makes headlines, while the harm caused by unnecessary detention does not. Nuanced policy loses to simple slogans.

The Laws Behind the Numbers

Federal legislation from 1984 and 1986 through the 1994 crime bill expanded mandatory minimums, especially for drug and gun offenses, and rewarded states for "truth in sentencing" laws requiring prisoners to serve at least 85 percent of their terms. Federal highway funding was tied to suspending driver's licenses for drug convictions, a policy Barkow calls a public safety mistake because people without licenses often cannot get to work. But only a small share of incarcerated people are in federal custody. The main drivers were in the states, where elected prosecutors, legislators, and governors pushed longer sentences and prosecutors charged more cases as felonies.

Why Prisons Fail at Public Safety

About 95 percent of people in prison eventually come home. Long sentences cut people off from family and from a changing world, and Barkow tells of returning prisoners baffled by automatic faucets. Violent, dehumanizing prison environments and untreated mental illness create trauma. Programs that reduce recidivism, including drug treatment, mental health care, education, and vocational training, are cost-effective but chronically underfunded. No prison is judged by its outcomes, and about 70 percent of people released are rearrested within five years. The belief that crime is purely individual choice, the devaluing of incarcerated people's lives, and racial attitudes all make investment politically hard.

The Willie Horton Effect

Barkow recounts Massachusetts' furlough program, which had a 99.9 percent success rate until Willie Horton committed a brutal assault while on furlough. The case became the basis of a notorious 1988 campaign ad against Michael Dukakis. Politicians learned never to risk their own "Willie Horton," and support for parole, clemency, and release programs withered. Policy by anecdote demands zero risk, she argues, which we accept in no other area of life. Ending a program for everyone because of one tragedy makes us less safe, not more. She also stresses that victims and offenders are overlapping populations, and that the current system serves victims poorly too.

Reform Begins with the Budget

Barkow dates the first wave of change to the mid-2000s, and it was driven largely by money. States, which cannot print currency, found corrections consuming their budgets. In Texas, a conservative governor who wanted to avoid a tax increase and new prison construction turned to diversion, drug treatment, and changes for low-level property offenses, giving rise to the "Right on Crime" movement. Other states followed with modest steps, such as easing marijuana penalties and raising the dollar threshold for felony theft. Because crime did not rise, each reform built the case for the next. Studies suggest incarceration explains at most about a quarter of changes in crime, Barkow notes, so sentences can be adjusted without harming public safety.

Three Turning Points

Barkow identifies three larger developments. First, high-profile police killings of unarmed people and the Black Lives Matter movement made mass incarceration a racial justice issue, as Michelle Alexander urged in The New Jim Crow. That created a political cost to "tough on crime" that candidates such as Hillary Clinton had to answer for. Second, in Brown v. Plata, the Supreme Court held 5–4 that California's severely overcrowded prisons violated the Eighth Amendment and ordered the state to reduce its prison population by tens of thousands. California shifted responsibility and funding to counties. Once localities had to pay for incarceration, they found many people did not need it, and the state's prison population fell sharply without the predicted surge in crime. Third, the U.S. Sentencing Commission, on which Barkow served, voted unanimously to lower federal drug guideline ranges by two levels and to make the change retroactive. About 30,000 people received reduced sentences, and earlier data from similar crack-cocaine reductions showed no increase in reoffending.

The Power of Prosecutors

Prosecutors shape both who enters the system and how long they stay, Barkow explains. Their charging decisions, including whether to charge at all, are essentially unreviewable. Research by scholars such as John Pfaff links much of the growth in incarceration to prosecutors bringing more cases as felonies. Prosecutors also drive pretrial detention, which accounts for roughly a quarter of people behind bars, through bail requests. Their sentencing recommendations, and charges carrying mandatory minimums, often determine the outcome. That is why reform-minded candidates, such as Philadelphia District Attorney Larry Krasner, began running for prosecutor on promises to reduce incarceration.

The Limits of Popular Fixes

Marijuana legalization, Barkow cautions, is not the answer. Drug offenses of all kinds account for about a quarter of people incarcerated, and more than half are held for offenses involving violence. Marijuana may matter as a pretext for police stops that lead to other charges, but ending marijuana prosecutions alone will not end mass incarceration. Neither will focusing only on private prisons or nonviolent offenses. Serious reform, she argues, must address sentences for violent crime. People age out of crime, and the adolescent brain is still developing into the mid-20s. The "superpredator" rhetoric of the 1990s, recalled in the story of the Central Park Five, produced decades-long sentences that serve no public safety purpose. Sentence lengths, she says, are "plucked out of thin air."

Diversion and Better Prisons

Diversion lets police and prosecutors give people a break for one-time mistakes and send people with addiction or mental illness to treatment rather than jail. Barkow points to Miami's mental health court, where completing treatment leads to dismissal, as effective and cost-saving. She urges an end to the school-to-prison pipeline and notes that deterrence depends far more on the likelihood of being caught than on the severity of punishment. For those who are incarcerated, preparing for release should start on the first day. Education, vocational training, substance abuse treatment, and meaningful work all reduce crime, and she describes the pay question for prison labor as a balancing act.

Structural Solutions

Barkow is not especially optimistic that ordinary politics will produce more than modest change. The biggest reductions came from institutions insulated from electoral pressure: the Supreme Court in Plata, and a bipartisan Sentencing Commission whose members did not have to face an attack ad. She urges attention to who sits on the federal bench, including judges with criminal defense experience, and more use of expert bodies charged with weighing data and costs.

What to Know Now

The overall picture Barkow describes still holds, though some numbers have changed. Roughly 1.9 million people are now incarcerated, down from about 2.2 million when this conversation was recorded, and the United States now ranks behind El Salvador and a few other countries in its incarceration rate. The U.S. incarcerated population fell to its lowest level in decades during the COVID-19 pandemic, as jails released people and courts slowed, then partially rebounded. Crime has moved independently of incarceration, as she predicted. Homicides spiked in 2020 and 2021, then fell sharply in 2023 and 2024 to some of the lowest levels in years. People on both sides of the debate cite those swings: some argue reforms went too far, while others note that crime fell even as incarceration stayed below its peak.

Some of the specific harms she described have been addressed. Congress restored Pell Grant eligibility for incarcerated students, which took full effect in 2023 and answers the "why pay for prisoners' college?" objection she described. The Martha Wright-Reed Act gave the FCC authority to regulate prison and jail phone and video call rates, and the agency adopted new rate caps in 2024, though in 2025 it replaced them with higher interim caps. Many states have ended driver's license suspensions for drug convictions, and "clean slate" laws in states such as Pennsylvania, Michigan, and New York now automatically seal many old records, reducing the lifelong consequences of a criminal record. The political dynamics she identified remain powerful. Crime remains a major campaign issue, and debates over release, parole, and bail are still often driven by high-profile individual cases.

Federal reform continued with the First Step Act of 2018, which reduced some mandatory minimums, made the Fair Sentencing Act's crack-cocaine reductions retroactive, and created earned-time credits for programming. It also let defendants seek compassionate release directly from courts. The Sentencing Commission lacked a quorum for several years, then resumed work in 2022. Since then it has expanded the grounds for compassionate release (though in Rutherford v. United States (2026) the Supreme Court held that nonretroactive changes in law cannot be an extraordinary and compelling reason for release), reduced guideline ranges for people with no criminal history and limited "status points," applying some of those changes retroactively, and restricted the use of acquitted conduct at sentencing. That record supports Barkow's argument that expert, insulated bodies can move policy. Marijuana reform spread to roughly half the states, and presidential pardons covered federal simple possession, but, as she predicted, the effect on prison populations has been modest.

The prosecutor movement has faced strong headwinds. Krasner won a third term in Philadelphia in 2025, but San Francisco voters recalled Chesa Boudin in 2022 and Los Angeles voters replaced George Gascón in 2024. California voters also passed Proposition 36 in 2024, increasing penalties for some theft and drug offenses and partly reversing earlier reforms. At the federal level, the Justice Department in 2025 directed prosecutors to charge the most serious readily provable offense. On juveniles, the Supreme Court in Jones v. Mississippi (2021) declined to require a finding of permanent incorrigibility before a life-without-parole sentence, while many states have abolished such sentences for children. Crisis-response and diversion programs for mental illness have expanded. Supporters of tougher policies point to the 2020–2021 rise in violence. Reform advocates note the sharp declines in crime since and the continuing costs of long sentences. Barkow's central point remains: lasting change depends on institutions that follow the evidence rather than the headlines.