The announcement last summer that in 2009 the number of Americans behind bars had increased for the 37th year in a row provoked a fresh round of national soul-searching. With its prisons and jails now holding some 2.4 million inmates—roughly one in every 100 adults—the United States has the highest incarceration rate of any free nation. As a proportion of its population, the United States incarcerates five times more people than Britain, nine times more than Germany, and 12 times more than Japan. “No other rich country is nearly as punitive as the Land of the Free,” 

But a highly significant fact went largely unremarked amid the hubbub: The population of the nation’s state prisons, which house all but a relative handful of convicted felons, decreased by nearly 3,000. Although the drop was slight in percentage terms, it was the first since 1972. (State prisons held 1.4 million inmates at the end of 2009 and federal prisons more than 200,000, while the number held in local jails, mostly for minor crimes, averaged about 770,000 over the course of the year, and the majority had yet to face trial.) In California, which has the nation’s largest state prison system, with nearly 170,000 men and women behind bars, the prison population fell for the first time in 38 years. The national prison population—including those held in federal facilities—grew by less than one percent, the slowest rate in the last decade. These changes mean it is very likely that we are seeing the beginning of the end of America’s long commitment to what some critics call “mass incarceration.”

If that shift does occur, it will not be because the United States has solved its crime problem. In fact, if there were a close correlation between crime rates and incarceration, the prisons would have begun emptying out in the late 1990s, when crime in most of its forms began to decrease.

How did we get here? Soaring crime rates, especially in the inner cities, are the most obvious part of the explanation. From 1960 to 1990, the overall U.S. crime rate increased more than fivefold, the frequency of violent crime nearly quadrupled, and the murder rate doubled. Drug use increased. The upsurge was widely blamed on lenient punishment, particularly for violent repeat offenders. Legislatures responded by passing “get tough” measures, including sentencing guidelines (which required prison sentences for some offenders who in the past might have been put on probation), so-called three-strikes-and-you’re-out laws (which mandated prison terms for repeat offenders), mandatory minimum sentences (forcing judges to impose fixed sentences regardless of mitigating factors), and truth-in-sentencing measures (requiring inmates to serve a greater proportion of their imposed sentence before becoming eligible for parole). These policy changes increased both the probability of going to prison if convicted and the length of prison terms.

Many liberal critics, pointing out that two-thirds of those imprisoned in federal and state facilities are African Americans and Hispanics, contended that “mass incarceration” is little more than a reworked form of racial and social domination—“the new Jim Crow,” as Michelle Alexander, a law professor at Ohio State University, put it in the title of her recent book.

But virtually all those who study the matter now agree that imprisonment has reached often counterproductive levels, particularly in the case of drug possession and other nonviolent crimes. The prominent conservative scholar James Q. Wilson, whose book 

 (1975) set the national crime control agenda during the 1980s, recently wrote, “This country imprisons too many people on drug charges with little observable effect.” In my travels around the country I have conducted an unscientific survey of prison administrators, and nearly all of them say that 10 to 15 percent of their inmates could be safely released.

If there were a close correlation between crime rates and incarceration, the prisons would have begun emptying out in the late 1990s, when crime in most of its forms began to decrease.

What we are seeing today is a growing recognition that our approach to dealing with convicted criminals is simply too costly. Not only is the price too high, but the benefits are too low. The states now spend an estimated $50 billion on corrections annually, and the growth of these outlays over the past 20 years has outpaced budget increases for nearly all other essential government services, including transportation, higher education, and public assistance.

California, where I was involved in the corrections system in various capacities under reform-minded governor Arnold Schwarzenegger, pours 10 percent of its massive state budget into correctional facilities. Between 1985 and 2005, it built 21 new prisons—more than one a year. The state’s prison population surged, and so did costs: The state spent nearly $10 billion on corrections last year, or about $50,000 per prisoner. (The national average is $23,000.) Now that California is grappling with a budget crisis, it is clear that it cannot continue on this course. The evidence for the rest of the country may be less dramatic, but it is no less clear.

These vast sums are not buying as much as many people think. Mass imprisonment has helped reduce crime rates, but most specialists agree that the effects have been considerably smaller than proponents claim and that we are now well past the point of diminishing returns. Confinement behind bars accounted for at most about a quarter of the substantial decline in crime that occurred during the 1990s (mainly, most researchers believe, by preventing imprisoned offenders from committing fresh crimes against the general public rather than by promoting a deterrent effect).

More important, that decline may well be reversed if we don’t do a better job of planning for the reentry of prisoners who have finished their sentences. There is a very simple and immutable “iron law” of imprisonment: Almost everyone who goes to prison ultimately returns home—about 93 percent of all offenders. (A relative handful die in jail; the rest have life sentences or are on death row.) Although the average offender now spends 2.5 years behind bars, many terms are shorter, with the result that 44 percent of all those now housed in state prisons are expected to be released within the year. This year, some 750,000 men and women will go home. Many—if not most—will be no better equipped to make successful, law-abiding lives for themselves than they were before they landed in prison.

These vast sums are not buying as much as many people think. Mass imprisonment has helped reduce crime rates, but most specialists agree that the effects have been considerably smaller than proponents claim and that we are now well past the point of diminishing returns.

Today’s offenders are different from those of the past. They are still overwhelmingly male (though the female proportion of the population has climbed to nine percent), African American or Hispanic, and unskilled. But the offenders leaving prison now are more likely to have fairly long criminal records, lengthy histories of alcohol and drug abuse, significant periods of unemployment and homelessness, and a physical or mental disability. Their records are more likely to include gang activities and drug dealing. In short, the average offender today leaves prison at a greater disadvantage (and more primed for trouble) than his predecessors did. Yet fewer participate in prison rehabilitation and work programs than a decade ago. When I was co-chair of California’s Expert Panel on Rehabilitation in 2007, the panel found that California spent less than $3,000 per year, per inmate, on rehabilitation programs, and that 50 percent of all prisoners released the year before had not participated in a single program.

Even as the states were cutting back in-house prison programs most severely, in the decade from 1985 to 1995, Congress and state legislatures were passing dozens of laws closing off many job opportunities to ex-offenders and restricting their access to welfare benefits and housing subsidies. Former inmates are now commonly barred from working in some of the economy’s fastest-growing fields, including education, childcare, private security, and nursing and home health care. Such legal barriers sometimes protect us from dangerous felons, but they also make it hard for men and women who want to go straight to get their feet on the ground.

It should not come as a surprise to learn that we have a corrections system that does not correct. The U.S. Bureau of Justice Statistics reports that two-thirds of released prisoners are rearrested for at least one serious new crime, and more than half are re-incarcerated within three years of release. The two-thirds rearrest rate has remained virtually unchanged since the first recidivism study was conducted more than 40 years ago. Former prisoners account for an estimated 15 to 20 percent of all arrests among adults. That means that thousands of Americans are being victimized every year by criminals who have already done time without experiencing “correction.”

At the same time, we are beginning to recognize that our overreliance on locking people up has an especially malign effect on poor urban neighborhoods, where up to 20 percent of the adult male population may be behind bars at any given time. Not only do the men come home with diminished prospects that hurt the whole community, but as criminologist Todd Clear shows in 

(2007), their absence weakens the family and social networks they need when they come home and hurts those left behind. It is no accident that the sons and brothers of men who go to prison are more likely to follow the same path. These trends help cause crime rather than prevent it.

Prison is where some people belong, many for long periods of time. But we need policies that do not produce more crime in the long run.

Budget cutters may rejoice at the chance to gut corrections budgets, and liberal critics of “mass incarceration” may celebrate any policy that shrinks the prison population, but cutting corrections budgets will prove hugely counterproductive if we act without giving serious thought to how we will deal with the offenders who are released. Until recently, for example, Kansas was a model of forward-thinking prison policy. In 2007 the state legislature funded a range of programs—involving education, drug treatment, and subsidized housing—to help former inmates reintegrate. The approach appeared to work: The number of ex-offenders returning to prison dropped by 16 percent between 2007 and 2009. But then came the economic crisis and cutbacks. According to state legislator Pat Colloton, recidivism rates quickly spiked. Kansas is back where it was in 2007.

It should not come as a surprise to learn that we have a corrections system that does not correct. The U.S. Bureau of Justice Statistics reports that two-thirds of released prisoners are rearrested for at least one serious new crime, and more than half are re-incarcerated within three years of release. The two-thirds rearrest rate has remained virtually unchanged since the first recidivism study was conducted more than 40 years ago.

To avoid throwing away much of the progress we have made in reducing crime, it is more imperative than ever that we pursue alternatives to prison and new ways to ease inmates’ reentry into civilian life. The good news is that after decades of false starts, researchers have finally begun to zero in on the things that can make a difference in at least some cases. The news was good enough to help persuade the conservative Bush administration to push through the $330 million Second Chance Act in 2007, giving government agencies and nonprofits the tools to get some of these efforts off the ground. The money was to be doled out over time. The bad news is that amid today’s intensified financial strains, Congress may be reluctant to continue funding this effort to enhance prisoner reentry programs.

Rehabilitation programs reduce recidivism if they incorporate proven principles and are targeted to specific offenders. Research demonstrates that offenders who earn a high school equivalency diploma while behind bars are more likely to get jobs after release. Those who receive vocational skills training are more likely to get jobs and higher wages after release. And those who go through intensive drug treatment programs in prison are less likely to relapse outside of it. If we could implement effective programs, we could expect to reduce recidivism by 15 to 20 percent. To put it in concrete terms: About 495,000 of the 750,000 prisoners who will be released this year are likely to be rearrested within three years. With effective programs, we could reduce the number of repeat offenders by nearly 100,000. We could do even better if these efforts were linked to improved services in the community upon release. Such efforts would pay for themselves by reducing future criminal justice and corrections costs. Economist Mark A. Cohen and criminologist Alex Piquero found in a recent study that a high-risk youth who becomes a chronic offender costs society between $4.2 and $7.2 million, principally in police and court outlays, property losses, and medical care. We either pay now or pay later—and we pay a lot more later.

Advocates of rehabilitation constantly struggle against the widespread view that “nothing works.” In part, this view grows out of an experience that began in the 1980s, when horrendous prison crowding in southern prisons, economic woes, and court rulings spurred some unusual experiments. When federal courts ordered states either to build new facilities or find some other way to punish offenders, the states began experimenting with alternative sanctions. Georgia, for example, developed an intensive supervision program (ISP) for probationers that yielded some evidence that it reduced recidivism rates—and also appeared to save the state the cost of building two new prisons. By the mid-1990s, virtually every state had passed some kind of legislation for intermediate sanctions.

Probation and parole departments across the country implemented a variety of ISP programs, including boot camps, day reporting centers, and electronic monitoring. The hope was that some offenders who normally would have been bound for prison could be “diverted” from expensive prison cells to intensive community programs that could keep a closer watch on them and offer more support services. Other offenders could be released early into community programs. But as I discovered when I was codirector of the RAND Corporation’s national evaluation of ISPs in the early 1990s, despite all the good intentions, most of the ISP dollars wound up being used to fund more drug testing, parole agent contacts, and electronic monitoring rather than enhanced social services. The main result was that offenders who violated court conditions by using drugs, for example, were identified more quickly and sent into custody.

Within a decade, ISPs went from being “the future of American corrections,” as one probation officer enthused to a Washington Post reporter in 1985, to what seemed to be a failed social experiment. Most of the programs were dismantled by the late 1990s. Some advocates of the prison buildup pronounced that alternatives to prison had been tried and did not work. But the RAND study found that in places where efforts were actually implemented according to the original design, they were rather effective. Offenders who participated in drug or alcohol treatment, community service, and employment programs had recidivism rates 10 to 20 percent below those of nonparticipating offenders.

Today, we have even more refined knowledge of what works. The most popular approach involves using something akin to a medical technique, focusing on individual cases. Called the risk-need-responsivity (RNR) model, it uses risk assessment tools to size up each person and match him or her to the right program. The treatment efforts are behavioral in nature (with rewards and punishments) and geared to place the sharpest focus on higher-risk offenders. There is a heavy emphasis on cognitive behavioral and “social learning” techniques—ranging from anger management training to sessions devoted to weaning offenders away from their negative and antisocial attitudes. All of these efforts use peers and family members to reinforce their messages. And, as several studies show, they work. Criminologist Edward J. Latessa of the University of Cincinnati studied the results of RNR efforts in Ohio’s 38 halfway house programs and found that they cut the recidivism of high-risk offenders by as much as 20 percent. Several states, including Maine, Illinois, and Oregon, are now using the RNR model.

Community partnerships are another approach that hold great promise. An excellent example is the Boston Reentry Initiative, a city interagency program that brings together law enforcement, social service agencies, and religious institutions to start working with inmates while they are still incarcerated. On the day the prison doors swing open, a family member or mentor is on hand to meet each released prisoner, and social service agencies are prepared to begin working to help the former inmate get a fresh start. The initiative focuses only on the highest-risk offenders leaving prison. They are offered opportunities for work and treatment, but for those who fail to take advantage of them and slip back into crime, the program calls for swift arrest and fast-track prosecution. In a sense, the Boston Reentry Initiative is the ISP experiment all over again—but this time backed with treatment resources, mentorship, and community collaboration. The results have been impressive. Harvard researchers found that participants had a rearrest rate 30 percent lower than that of a matched comparison group.

It is no longer justifiable to say that nothing works. There is scientific evidence that prison and parole programs can reduce recidivism. It is not easy and it is not inexpensive, but it is possible. To retreat now would be to pull the rug out from under hundreds of programs that are contributing to the decades-long war against crime, which, whatever its shortcomings, has been one of the nation’s great success stories, vastly improving the lives of ordinary citizens and the vitality of cities. One of the surest ways we know to keep crime down is to prevent those who have committed crimes in the past from doing so again.

That is not to say that criminality is a problem that can always be solved. People go to prison for a reason, and in many cases there is very little or nothing that anyone can do to change the choices they will make in the future. Rehabilitation programs are not for every prisoner, and we should not waste money on those who lack motivation. But it would be foolish not to help those who wish to change. Effective rehabilitation and reentry programs that help offenders go home to stay are good for them, and good for the rest of us, too.

 is the Adelbert H. Sweet Professor of Law at Stanford University and codirector of the Standford Criminal Justice Center. She is the author of several books, including 

When Prisoners Come Home: Parole and Prisoner Reentry 

(2003), recently reissued in paperback, and she is coeditor of the new book 

Inmates_Orleans_Parish_Prison.jpg

For decades, America's chief answer to crime has been to put more criminals behind bars for longer. That expensive strategy is yielding diminishing returns. It's time for a closer look at ways of helping ex-offenders steer away from crime. 

Joan Petersilia