Prison Myths

excerpted from the book

Crime and Punishment in America

Why solutions to America's most stubborn social crisis have not worked - and what will

by Elliott Currie

Henry Holt, 1998, paper


What the experience of the past twenty-five years would seem to tell us, then, is that we have paid a steep price for an approach to violent crime that is badly out of balance. More than any other developed country in the world, we have relied on the jails and prisons as our first defense against crime; yet we still maintain the developed world's worst level of violence. And we have systematically depleted other public institutions in order to pay for our incarceration binge-a self-defeating course that helps to ensure that violent crime will remain high despite ever more drastic efforts to contain it.

Not everyone, however, would agree with this assessment. Indeed, we often hear a very different story. In the face of the reality that we use incarceration far more than most other nations and have dramatically increased it in recent years, we hear that America is "soft" on crime. In the face of sobering evidence that mass incarceration has had little impact on violent crime, we're told that "prison works." And in the face of evidence that this reactive strategy has been not only ineffective but enormously costly, we hear that "prison pays." The problem, we're told, is not that we have overemphasized incarceration at the other expense of other | approaches to crime, but that we haven't incarcerated enough.

Many Americans believe that the main reason we remain a frighteningly violent country is that we are shockingly lenient with criminals. That would seem, at first glance, a difficult position to maintain in the country that boasts the developed world's highest imprisonment rate. And, in fact, the idea that serious, violent criminals are treated leniently in the United States is a myth.


"For every dollar we spend to keep a serious criminal behind bars," John DiIulio writes, "we save ourselves at least two or three. The $16,000 to $25,000 a year it takes to incarcerate a felon is in fact a bargain, when balanced against the social costs of the crimes he would commit if free." James Q. Wilson, similarly, states it as a matter of undisputed fact that "all of the estimates of the cost of the prison population suggest that the benefits in terms of crimes avoided exceeded the cost by a factor of at least two to one."

Once again, the historical record would seem to contradict this position. Twenty-five years of the "prison experiment" have sucked up tens of billions of public dollars, but violent crime has stubbornly refused to respond in any consistent fashion. Yet the idea that prison, in spite of this record, is a "bargain" is on its way to becoming a fixture of the public discussion of crime in America.

The reasons aren't hard to find. In the late 1980s, as violence kept rising in the face of huge expenditures on prisons, a new set of constituencies began to weigh in against indiscriminate incarceration. Local and state officials, in particular, along with some farsighted business people, began looking at the prison spending explosion and wondering whether it was a bad deal in sheer economic terms. State officials began to realize that beyond a certain point building prisons meant they couldn't build colleges. Corporate executives began worrying about where they were going to get a competent work force if the government starved schools in favor of prisons. These concerns were rarely an expression of bleeding-heart leniency, but more often of hard, businesslike calculation.

Enter the "prison pays" argument-designed to show that prison not only stops crime but saves money in the process. Like the "prison works" argument, with which it is often yoked, this "myth of costlessness" is frequently supported by studies of great complexity and mathematical ingenuity. They involve calculating how much, in dollar terms, various kinds of crimes cost society and then estimating how many crimes are prevented by locking offenders up-and essentially multiplying the two to arrive at an estimate of the economic benefit of incarceration. Sometimes these calculations are quite sophisticated, even useful. Sometimes they are utterly fantastic.

A notorious 1994 study from the California governor's Office of Planning and Research, for example, claimed that the state's "three strikes and you're out" law, by locking up tens of thousands of new offenders, would save the state vast amounts of money. That was a surprising conclusion, since most analysts, including the state's own Department of Corrections, were predicting that it could cost the state several billion dollars. It turns out that the governor's researchers arrived at their more optimistic view in part by predicting that the three-strikes measure would prevent large numbers of homicides every year-more, in fact, than now occur in California.

.. we now spend [approximately] 40 billion a year on locking people up ...

As a recent report from New York's Vera Institute of Justice puts it, incarceration has a number of "unintended consequences"-consequences that, even in the simplest economic terms, undercut its benefits.

To take just one example, the rapid increase in the numbers of women behind bars-most of them mothers, many of them imprisoned on relatively minor drug charges or for property crimes related to their addiction-has left us with a growing problem of parentless children. Where on the cost-benefit ledger have we entered the costs of their substitute care, or their increased risks of delinquency, welfare dependency, or drug abuse in the future? The unprecedented rise in the incarceration of women will also put new pressures on already strained prison health-care systems. As James Marquart of the University of Texas and his colleagues write, "The full impact of the growth in the female prisoner population and concomitant medical costs has yet to be fully explored or considered. Incarcerating more women, coupled with their unique health demands, will be a costly crime control policy."

Other hidden costs will be generated by the rapid increase in the proportion of older inmates, which has resulted in part from the growing number of "lifers" under mandatory-sentencing schemes like "three strikes" in California. Between 1986 and 1995, the proportion of the state prison population serving life sentences or sentences of twenty years or more increased by nearly half, from 17 to 25 percent; by the year 200O, there may be more than 50,000 state prison inmates over the age of sixty-five, with correspondingly greater-and more expensive-needs for medical care.

Finally, the dwindling availability of serious vocational training or education in many prison systems means that most ex-inmates leave prison today even less able to fit into an ever more demanding labor market than they were when they went in; they've acquired the stigma of prison without increasing their capacity to function on the outside. (As the British criminologist David Garland notes, this is a deeper meaning of the idea that prison "incapacitates offenders.) To the extent that incarceration aggravates the already severe labor-market problems of their mostly low-income, poorly educated inmates, it will increase the costs to the public sector of dealing with them on the outside-through public assistance, drug treatment, and emergency health care, as well as in lost taxes and economic productivity.

This points to a more general, and even more troubling, issue. It is undeniable that the experience of going to prison reduces recidivism among some offenders. But it is equally undeniable that it increases recidivism among others. Indeed, the tendency for incarceration to make some criminals worse is one of the best established findings in criminology, and it has been recently confirmed by highly sophisticated research. In a study of delinquent youth in Massachusetts, for example, John H. Laub and Robert Sampson have shown that going to prison often increased the chances of their committing further crimes as adults, mainly because it reduced their prospects of getting a stable job. Obviously, to the extent that prisons increase the propensity of some people to commit crime, their economic benefits will shrink and their real costs will go up. But this problem is ignored in most discussions of the economic benefits of incarceration. Just as the "benefits" of incarceration are routinely exaggerated, in short, so too are the "costs" of imprisonment greatly underestimated.

In the end, what the studies of the costs and benefits of imprisonment most clearly tell us is that critics of indiscriminate incarceration have been right all along. No matter how the numbers are manipulated, the results confirm that prison makes sense for some offenders but not for others. It makes a great deal of sense for truly violent people, but its utility dwindles to the vanishing point for minor property offenders, not to mention minor drug offenders, who are horrendously expensive to incarcerate even by the most generous "economic" calculation.

The economic inefficiency of incarceration for nonviolent offenders is especially troublesome given the thrust of the new conservative penal policy, which is not only to incarcerate those offenders but to sentence them to long terms. Even leaving aside any questions about the way costs and benefits are calculated in some of these studies, they show that the net costs of the lengthy imprisonment of minor offenders are staggering. Take the Miller et al ... [study] The average burglary, they estimate, costs society $1,400. Suppose the average burglar commits ten burglaries a year. Society would thus "save" $14,000 a year by locking him up. Locking him up for a minimum of twenty years on a third conviction, as California's three-strikes law mandates, would "save" $280,000. But it would also cost (at least) $21,000 a year, meaning that the social deficit each year is roughly $7,00O, or a net loss of $140,000 over twenty years. (Note that this calculation excludes the cost of prison construction.) And that is for the most serious of property crimes. If we do the same exercise with larceny the losses are ~ stratospheric. Ten thefts a year would cost society, on average $3,700. A year's incarceration thus generates a net cost of more than $17,000. Thirty years under a Draconian habitual-offender sentence would cost society half a million dollars. Even if we assume an average of twenty-five thefts a year, the annual net expense of incarcerating the thief remains about $12,00O, and the "lifetime" expense over $350,000.

Indeed, this is actually an underestimate of the net costs of incarcerating minor offenders, because it assumes they would continue, if on the street, to commit crimes at the same high rate even as they get older. But we know from a host of careful studies that they will not; instead, they are likely to "age out" of high-rate offending, and by middle age many will have radically slowed if not stopped altogether. ("Aging-out," in fact, is one of the most predictable phenomena in criminology.) Meanwhile, the cost of keeping these offenders in prison begins to rise inexorably as their advancing age leads to increased expenses for medical care. It is very difficult, in short, to come up with a credible calculation in which prison "pays" for most nonviolent offenders. As we sweep more and more of them behind bars, for ever longer sentences, we are increasingly pouring money into an enterprise with small and constantly diminishing returns. The social programs of the 1960s I were often criticized for "throwing money at problems"; this is throwing money at problems, with a vengeance.

1995 study of New Jersey inmates by John DiIulio and Anne Piehl

DiIulio and Piehl acknowledge, putting ordinary drug offenders behind bars has very little effect on the rates of drug-related crime. They argue, indeed, that "the best estimate of the incapacitation effect (number of drug sales prevented by incarcerating a drug dealer) is zero." Hence the economic benefit of prison for most of these offenders is also zero. And because of the war on drugs, there are a lot of such people in prison; fully 27 percent of their New Jersey sample of inmates said that in the four months before being imprisoned their only offense was drug sales. Piehl and DiIulio found that when drug offenders were added to the calculation, prison even for the offender at the median level of social cost turns out to be "cost-ineffective"; that is, for more than half the offenders behind bars in New Jersey, it is more expensive to society to keep them behind.

Crime and Punishment in America

Index of Website

Home Page