Crime is a corrosive force in American society, creating and exacerbating social mistrust, political division, and racial animus. Doing something serious about it ought to count as a major objective, especially for those concerned with the plight of the poor and socially marginalized. The recent drop in crime rates, as welcome as it is, has simply returned us to the level of the early 1980s. Much deeper reductions in crime and disorder are necessary, most of all for the sake of poor city dwellers, who are the most victimized by crime and by the crime-avoidance behavior of those who choose to locate elsewhere, depriving the people left behind of nonpoor neighbors, job opportunities, and convenient access to goods and services.
America's current criminal-justice policy, dependent upon incarceration as its only real punishment, cannot get us to the level of safety that should be a civic birthright for all Americans. What is more, the benefits we have reaped from our increasingly high incarceration rates have come at a staggering cost. And the burden has been felt most acutely by those on whose behalf the struggle against crime and drug abuse most urgently needs to be fought: poor (largely black and Latino) Americans.
It isn't only liberals who have noticed this. A number of longtime defenders of more and longer imprisonments have recently announced second thoughts about the trajectory of American prison policy. John DiIulio, the most academically distinguished proponent of harsher sentencing laws, has declared that "the nation has 'maxed out' on the public safety value of incarceration." Glenn Loury, the prominent black neoconservative, has stated that "accumulating evidence demonstrates that the punitive antidrug crusade of the last decade is, in fact, not producing benefits commensurate with its substantial costs." Even James Q. Wilson, the senior conservative guru on crime control policy, has declared the current prison-focused drug strategy a hopeless failure.
But policies do not change simply because well-meaning intellectuals provide compelling critiques. This is especially true in the area of crime, where the problem is unquestionably serious and the policy response--incarceration of huge numbers of people--is so deeply entrenched politically, morally, and institutionally.
This astonishing expansion of state power, ironically, has come about due to a lack of trust in government. Voters do not trust legislators or judges, whom they see as the people who got us into the crime mess in the first place. Legislators have responded by tying the hands of judges with "three strikes" laws and mandatory minimum sentences. We have, in effect, put the expansion of incarceration on automatic pilot.
Over-incarceration is also driven by the absence of effective alternatives to prison. There is literally no state in the Union with a well-developed set of institutions, other than prisons, to which judges can send criminals and expect that substantial punishment, incapacitation, or rehabilitation will be administered. Virtually everywhere, in the eyes of offenders, victims, police, and prosecutors, a probation sentence is tantamount to a free ride. Under these circumstances, a legislator who wants to punish and restrain known criminals is like a handyman with only a hammer. It would be nice to have a screwdriver, but you work with what you've got, and after a while, everything starts to look like a nail. Meanwhile, for lack of real alternatives to prison, critics of incarceration are put in the untenable position of supporting the release of criminals with, effectively, no sanction whatsoever.
In order to change the politics of criminal justice, therefore, we must first change its instruments. This requires that advocates of reductions in incarceration admit that convicted criminals need to be punished and restrained, and that they develop institutions that can be punitive and rehabilitative without the social and economic cost of prison. While it is possible to make a good career merely denouncing over-incarceration (or, on the other side of the aisle, drug abuse), denunciation does not constitute policy.
The Purpose of Punishment
Legal punishment can serve important purposes other than crime control. A crime denies both the validity of the norm it violates and the importance of the victim it damages; punishment reaffirms those things and reduces the incentive for private vengeance. Still, we do want punishments to reduce the level of criminal activity, and this can be accomplished in only a limited number of ways: by deterrence (creating disincentives for future offenses), by incapacitation (reducing or eliminating a person's capacity to re-offend for some period), and by rehabilitation (changing either the offender's objective circumstances or his value system in ways that make crime less attractive).
How does incarceration measure up as an anticrime strategy? Prisons are unmatched at incapacitation, and for most people, the threat of prison is also an excellent deterrent. We also have the administrative capacity to build prisons, staff them, and keep prisoners inside them.
Those statements, however, exhaust the virtues of incarceration. As rehabilitation, prison has proven in the past to be no better than a wash, even in the days when prison administrators were putting energy and money into programs designed to reduce recidivism. Today's programmatically leaner and organizationally meaner prisons may well be worse than that.
There is a basic problem with the use of prison for rehabilitation, which fans of turning the prison system into a massive drug-treatment program ignore. Behavior change tends to be highly situational. Getting people not to use drugs in prison isn't very hard, but it's not that useful, either; after all, the point is to get them not to use drugs when they're back in the community. True, a few high-quality in-prison treatment programs have been shown to reduce future drug use and recidivism, when supplemented with good post-release follow-up. But most of what passes for drug treatment in prison isn't very high quality, and there's solid evidence that good post-release follow-up alone can do much of the job.
Nor can prisons deliver much in the way of valuable job training or work experience. Once again, the problem is providing an experience in prison that is close enough to the world outside to have value there. (For example, Arizona's famous and apparently quite successful use of prisoners to staff the reservations line for a major hotel chain--thus skirting the federal ban on interstate commerce in prison-made goods by providing interstate services instead--has been more discussed than imitated and is no longer in operation, for reasons the Arizona Department of Corrections will not specify.)
Similarly, the fad for "boot camps" for offenders reflects a misunderstanding of what happens in a real boot camp such as Parris Island. The Marine camp, like its criminal-justice parody, involves physical challenge, yelling, discomfort, and stress, and is designed to strip its subjects of their old values, behavior patterns, and identities. At the end of that process, those who don't drop out will have a new identity, role, and status, which many enlistees desired initially and an even greater number will come to value during the process, and which present them with supports for their self-respect and a certain amount of respect from others. That is to say, at the end of Marine boot camp, the subject is a Marine and likely to be proud of it.
By contrast, the criminal-justice "boot camp" does fine on providing stress and tearing down previously acquired identities, but it provides no new identity capable of commanding respect or supporting self-respect. Ultimately, if it succeeds, the subject is an ex-convict and ashamed of it.
The great exception to this discouraging litany is literacy. Being a skill rather than a habit, literacy is portable from the prison setting to the outside world. Anne Morrison Piehl, at the Kennedy School of Government, has gathered impressive evidence that improvements in reading scores correlate closely with reductions in the probability of re-arrest, for the simple reason that employment competes with crime as an outlet for time and energy, and employment is more available to those who can read than to those who cannot. With the development of computer-assisted literacy labs, there is no reason prisons can't start teaching this skill, but rehabilitation is so out of favor both with prison administrators and with legislatures that little progress is being made on this front.
These inadequacies of prisons--and their staggering costs in money and suffering--suggest that we would want to use incarceration sparingly if we had another way to deter and incapacitate offenders.
We would still want to use prisons for those whose criminal activity on the outside is frequent and serious enough to merit the maximum available incapacitation, even at a cost of $25,000 per year. But picking out those truly deserving cases from the mass of routine, relatively low-threat offenders is harder than it looks. To be reasonably sure that the most persistent and dangerous offenders are behind bars, we have to lock up large numbers of people who wouldn't do much damage if they were out. When the choice is between prison and nothing, as it effectively is right now, it makes sense to imprison 10 offenders for every one we actually need to keep behind bars. But surely it would be better to develop other punishments capable of reducing crime--perhaps not as certainly as prison, but more cost-effectively.
Prison would still be useful as a deterrent and, of course, as the backup sanction for those who refused to comply with the terms of nonprison sentences. But prison sentences designed to maximize deterrence could justifiably be short. While the threat of prison time deters, the threat of five years in prison doesn't deter five times as much as the threat of one year in prison. Deterrent value is subject to diminishing returns.
The present system is particularly unsuitable for repeat offenders. As James Q. Wilson and Richard Hernstein have noted, repeat offenders tend to be reckless and impulsive. Controlling their behavior, therefore, depends more on our ability to deliver sanctions swiftly and predictably than it does on the severity of those sanctions. The current system of erratically imposed draconian sentences may make offending unattractive to most people. But if crack-using burglars reacted well to a small threat of a big sanction, they wouldn't be crack users or burglars.
Much of the political discussion of crime control has centered on increasing the severity of punishment, in part because that's something legislators can do relatively easily. But if the problem of crime is largely one of repeat offenders, then increasing severity (through mandatory sentencing, "truth in sentencing," three-strikes laws, or other such schemes) will prove to be largely a dead end.
In theory, probation and parole--the community corrections system--could be far more effective with repeat offenders. Community corrections officers could respond to these offenses with great certainty (since probationers and parolees are subject to intrusive surveillance) and much more swiftly than the regular judicial system (since probationers and parolees don't have the full panoply of rights accorded criminal defendants). Moreover, community corrections officers can back up what would otherwise be merely good advice--about drug use, time management, choice of associates, treatment for behavior problems, job-seeking, job-holding, and the acquisition of job-related skills--with promises and threats of immediate pleasant and unpleasant consequences.
The current community corrections system rarely does that. Instead, it largely reproduces the behavior of the rest of the criminal-justice system: It makes many demands, spottily monitors compliance, and punishes detected deviations slowly and unpredictably, but often quite severely.
Nonetheless, community corrections remains the most plausible place to begin. For instance, probation and parole would need to deliver a form of punishment that effectively reduced crime--imposing on offenders intense levels of supervision, clear rules of behavior, and swift and certain sanctions for breaking these rules. Probationers and parolees might, for instance, be required to abstain from the use of illicit drugs, obey curfews and requirements to stay away from specific places (such as drug markets and the residences and workplaces of their former victims), and comply with orders to seek treatment, perform community service, meet financial obligations and family responsibilities, and show up for school or work.
In fact, the current system already imposes many such rules, though with insufficient attention to skills acquisition, work, and time-and-place restrictions. What's chiefly lacking is the capacity to detect violations and to respond appropriately. Face-to-face interviewing, the cornerstone of what little supervision efforts probation departments currently make, is simply not an effective violation-detection technology. Promising alternatives include drug testing, electronic position monitoring, and the creation of data links between community corrections and other public and private agencies with which community corrections clients interact.
What we know about drug testing is instructive. Contrary to the strongly held belief that drug use by addicts is involuntary behavior and therefore not subject to control by its consequences, the evidence is that frequent testing with predictable sanctions can rapidly decrease the rate of drug use. In Maryland a group of probationers starting on a testing-and-sanctions program had a positive-test rate of about 40 percent; four months later, the rate among the same group was down to 7 percent. Connecticut is reporting an even lower rate among a group of parolees with known cocaine and heroin problems. Older programs in Coos County, Oregon, and Lansing, Michigan, have demonstrated similar results. Testing-and-sanctions programs outperformed mandated treatment in a random experiment run as part of the District of Columbia drug court. Given the importance of probationers and parolees as drug market customers--our estimate is that about 60 percent of the total volume of heroin and cocaine consumed in America is sold to community corrections clients--testing-and-sanctions programs present a major opportunity to shrink the size of the drug problem as well as a potentially important innovation in community corrections.
Why a Failing Policy Succeeds
Typically, a policy article would at this point describe in greater detail just what a more sensible alternative to over-incarceration would entail. But the more pressing ques-tion is not about policy but about politics: Why, politically, does our present unsensible policy persist?
There are several plausible answers, but three, we think, deserve particular attention because they explain not only how we got into this mess, but also how to get out.
First, there is the American public's across-the-board mistrust of government, which we mentioned earlier. Curiously, this mistrust exists alongside a continuing expectation that government solve social problems. The way voters and politicians have squared this particular circle is through the creation of counter-discretionary government: indexation of Social Security payments, for instance, and three-strikes laws that require government to act but also tie its hands.
Second, most decision makers are conservative, in the sense that they mistrust theoretically clever ideas and find evidence of success more compelling. They pick not from an infinite array of policy alternatives, but from a much more limited set of choices that are known, legitimate, and proven. Thus, state legislators who want to argue for the kinds of punishments we've been talking about--coerced abstinence and highly supervised probation and parole--will be asked where those approaches have been tried and what happened there. And if they don't have answers, they will lose politically.
The situation is analogous to what conservatives faced in welfare reform in the 1970s and early 1980s. Conservatives could make a compelling critique of Aid to Families with Dependent Children, but they had no alternative that had proven itself on the ground. Welfare defenders had the advantage: They could claim that imposing substantial new obligations on recipients was "risky" and "unproven" and would likely lead to dangerous and unforeseen consequences. What dramatically changed the politics of welfare was the emergence of state-level experimentation, actively encouraged by the Reagan and Bush administrations. While many of these experiments showed fairly minimal results, others partly succeeded, and alternative policies thus came to be known and legitimated.
Third, the policy of over-incarceration persists because administrative and budgetary arrangements work against more efficient alternatives. A governor with rising prison bills can't solve the problem by improving probation and parole because, in most states, probation (by far the larger part of community corrections) is under the control of judges, not the executive branch, and paid for out of county, not state, budgets. It would make sense for the states to offer financial incentives for probation departments to do their job well, perhaps some fraction of the avoided cost of imprisonment. But no such arrangement exists anywhere in the country. As a result, states wind up spending $25,000 per year to lock people up because counties won't spend an extra $1,000 on a good drug-testing program.
Similarly, on the prison side of the equation, it is the states that pay for prisons, but prosecutors and criminal court judges are, almost everywhere, county officials. When a prosecutor demands a 10-year prison sentence and the judge imposes it, the county doesn't pay the $250,000 of public funds they just jointly decided to spend. Counties whose prosecutors and judges overuse prisons are in effect being subsidized by the rest of the state.
Meanwhile, at the state level, when a legislature considers new sentencing laws, no one is forced to pay attention to costs. Annual budget calculations reflect only the cost of imprisoning whomever is behind bars this year, not the future cost commitment represented by sentencing practices. And annual budgeting puts probation--like any other preventative program--at a disadvantage. Its costs come now; its benefits, in reduced imprisonment, come later.
Thus, budgetary and institutional arrangements keep the politics focused on everything but the two primary objectives: reducing crime and reducing cost.
Academic arguments will not get America's overincarceration policies changed. But the political obstacles to change that we've identified are problems that careful strategic thinking and intelligent institution-building can address. Here are some suggestions:
Be punitive. The most important strategic move that opponents of mass incarceration can make is to shift from criticizing imprisonment to supporting punitive alternatives. It is not enough to argue that there would be less crime if we had better job-training programs or a more generous welfare state. The public wants to ensure that known criminals receive some punishment and that they will not pose a threat to public order or safety. If liberals can't convince Americans that these ends can be satisfied through nonincarceration alternatives, the public will continue to support politicians who promise to "lock 'em up."
Use federalism. Liberals typically see America's decentralized political system as an obstacle to better public policy, but the same "laboratories of democracy" that cooked up school choice and welfare reform can be used to get punitive alternatives to incarceration on the political agenda.
Set up models. The most important step is to find just a handful of states willing to put a system of intensive parole and probation into practice. If liberals in a few states can successfully employ alternatives to prison, other states are likely to follow. Center-left foundations could substantially fund organizations in a few carefully selected states. Conservatives were very shrewd to concentrate their welfare reform and school choice efforts in Wisconsin, a state of manageable size and with a political culture open to experimentation. Picking a single state also allowed conservatives to get a lot of bang for their buck. Liberals should be just as focused.
Fix the budgetary and institutional incentives. Seemingly bland administrative reforms could have substantial policy consequences. Encouraging states to put all criminal-justice expenditures on three- to five-year budget cycles--and to centralize budgeting and responsibility for probation, parole, and imprisonment--is not just good government; it's a way to push decision makers toward more efficient anticrime policies.
There are ample substantive arguments for beefing up the probation and parole systems, even if doing so would not reduce imprisonment. But there is a good political argument as well. Pushing for intensive probation and parole puts liberals on the side of order and authority--providing a mechanism for responding to the public's fear of crime and disorder in its own terms, rather than condescending to it by preaching about crime's "root causes." Only if liberals get behind forms of punishment that are more effective than prison will they be taken seriously when they call for lower levels of incarceration. The political alternatives are simple: increasing stringency outside prison walls, or housing a practically and morally indefensible number of Americans inside them. ¤