This is the second in a series of posts in which I’ll be serial-blogging my new article, Prison Vouchers, forthcoming in the Penn Law Review. The first post in the series is here. I’ve skimmed some of the comments to that post — some of the commenters’ concerns were addressed right in that first post, and others will be addressed as the series progresses (though if you want to know everything immediately, with all the supporting footnotes, you should look at the full version on SSRN). I’d appreciate comments (especially informed ones, which have a greater chance of making it into the final version, with thanks in the author footnote).
In this post, I discuss the mechanics of how a voucher scheme would work: how choice would operate, why choice isn’t the same thing as privatization, some funding details, and some statutory restrictions.
* * *
Inmates today have little or no choice of prison. In Texas, for instance, convicted defendants can’t ask for a particular prison at the outset; prisoners are assigned based on their supposed needs. There’s no formal way to “bargain” with the court or with the prosecutor. Voluntary transfers are limited. In the federal system, the Bureau of Prisons takes the court’s recommendations into account, but there’s no guarantee that the court will convey the prisoner’s preferences, and generally there’s no systematic way for prisoners to have their preferences satisfied.
In California, there’s a limited amount of choice in jail assignment: through “offender self-pay” programs, minimal-risk offenders can, for a fee and with court approval, opt out of the regular jail system and be housed in the jail equivalent of a “five-star Hilton,” where they get distance from violent offenders, work furlough rights, and perhaps even the right to bring computer equipment. But this California system is both unusual and inegalitarian.
So under the standard regime, prisoners are assigned primarily based on a state correctional employee’s judgment of available space, as well as inmate needs like proximity to family or substance abuse or sex-offender treatment programs.
Prison choice would supersede this mechanism. The process would begin at conviction. The convicted defendant would receive a coupon, good for incarceration for the duration of his term, which he would be required to redeem at a participating prison (sorry, no choice there). The set of participating prisons may or may not include private prisons. In fact, as I discuss below, choice is conceptually independent of privatization, and—even though some arguments for or against vouchers are often made with private providers in mind—one can talk about vouchers even without privatization.
Imagine a convicted defendant awaiting sentencing, presumably with a lot of time on his hands. He can spend his time flipping through a book, perhaps like the Yellow Pages, with ads for different prisons—or perhaps the “flipping” can be done online if he has internet access. To get an initial view of dimensions along which prisons might compete, let’s take a more detailed look at some of the problems prisoners face today. Each of the problems listed below suggests possible reforms that inmates might find attractive. (Prisons could advertise such reforms with color photos, supporting statistics, and inmate testimonials.)
Violence. Though good data is elusive, violence against inmates, both by other inmates and by staff, is a serious problem. This includes prison riots as well as “gang violence, rape, beatings by officers, and in one large jail, a pattern of illegal and humiliating strip-searches.” Violence is fueled by, among other things, overcrowding, idleness, and distance from family and community.
A prison in a voucher system might respond to this problem by adopting a more effective security policy. This could include a “direct supervision” policy, which involves substantially more direct interaction between inmates and corrections officers (rather than the “traditional model” where guards supervise prisoners from behind glass or bars), and which has been said to improve safety. It could also include a focus on conflict resolution and prevention rather than the frequent current emphasis on using force as “a ‘first strike’ response before other tactics are considered or attempted.” It could also include a broader use of surveillance technology to “protect prisoners and staff from violence and from false allegations of misconduct” (to the extent this is consistent with inmates’ preferences for privacy), as well as non-invasive drug- and weapon-detection devices.
Health care. Correctional health care is inadequately funded, understaffed, and often provided by underqualified doctors. Communicable diseases, such as staph infections, tuberculosis, hepatitis C, and HIV, are thus widespread. Care for the mentally ill in prisons and jails—where rates of mental illness are two to four times higher than among the general public—is likewise inadequate.
A voucher prison could offer better medical care (including better screening, testing, and treatment of infectious diseases, and better care for the mentally ill), better staffed facilities, more doctors and nurses per inmate, and partnerships with community health-care providers. A prison could also provide a variety of health insurance plans, offering inmates the chance to opt out of the common cost-control system that requires co-payments for medical care.
High-security segregation. High-security segregation is over-used, often on prisoners who pose little security risk or are mentally ill. Some believe that such segregation is counterproductive, and segregated prisoners have little access to programming that could make them more productive citizens when they’re released.
A voucher prison could limit the use of high-security segregation, have secure therapeutic units for mentally ill prisoners, and offer more human contact and more programs to inmates in segregation, with transfer to the general prison population near the end of their sentence to prepare them for release.
Corrections officers. Corrections officers are underqualified and insufficiently trained to resolve problems without violence, and the general idea of treating inmates with respect—a valuable skill that helps maintain security and control—is undervalued in training. A voucher prison could recruit a more highly qualified and more diverse staff with lower turnover and higher morale.
These are only a few possibilities. Here are some more:
- A prison could offer better activities, like gym equipment and television programming.
- A prison could offer better “programming”: high-school and college-level education, job training (or opportunities for voluntary inmate labor), counseling, and other rehabilitative programs.
- A prison could offer more space per inmate.
- A prison could provide more opportunities for family visits, “providing ample space and time, and even assisting with transportation.”
- A prison could provide cheaper telephone calling plans that would allow more frequent communication between inmates and their families.
- A prison could locate closer to inmates’ home communities.
- A prison could institute an independent system of external monitoring with meaningful enforcement, perhaps through a non-governmental organization modeled after the International Committee of the Red Cross, which inspects “detention facilities in conflict zones worldwide.” This could involve encouraging politicians, judges, citizens, the media, and non-governmental organizations to visit prisons and interview prisoners and staff.
- A prison could develop a meaningful internal grievance system, where complaints are confidential, inmates get copies of their grievances, and prisoners and guards are protected from retaliation.
- A prison could seek accreditation from the American Correctional Association, or develop its own standards (or adopt another organization’s standards) that are more stringent. Regardless of the Prison Litigation Reform Act, prisons could allow inmate lawsuits against them, at least in state court and on contractual grounds.
Many of these policies would be costly for prisons; this is exactly why, rightly or wrongly, they haven’t been provided by the political system. Prisons that adopted such changes would, other things being equal, make less profit per inmate. But if the change is valued enough by inmates, the prisons could make up for the lower per-inmate profit by attracting more inmates. They could also bundle a valued but costly change with a reduction in some other amenities. For instance, a prison with better medical care (or that encouraged greater usage of medical care by abolishing co-payments) might locate further from the inmates’ communities (where real estate is cheaper) or might offer less programming.
The inmate’s choice would be limited to his security level (minimum, medium, or maximum). There may also be certain mandatory conditions attached to the voucher. Perhaps a sexual offender must go to a prison with appropriate programs, and perhaps someone else will need to make the choice for prisoners who are sufficiently mentally ill that they are judged incapable of choosing. But even for the mentally ill, the chooser needn’t be a Department of Corrections bureaucrat. The choice could be made by the inmate’s family or by an appointed legal guardian.
Having made his choice, the convicted defendant would be sent to his requested prison, subject to availability. (A prison system may want to guarantee a spot in certain units to certain prisoners, like those with particular physical or mental illnesses; also, a gang member who has informed on his gang may have to be sent to a “snitch farm” to avoid reprisals. ) As with schools, popular prisons will have waiting lists—which, provided the voucher amount is high enough, would provide an incentive for the prison to grow, since a prison wouldn’t need anyone’s permission to accept new prisoners if it expands (aside from whatever local construction permits may be necessary).
Once a prisoner is in a minimally acceptable place, he may want to stay put, for instance because of social connections he will have forged in prison. So it makes sense to at least offer the choice before incarceration begins. But because prisoners may not have enough information to make a good choice before they’ve served any time, it would probably also make sense to at least offer one transfer after some time, perhaps a year or two, so that they can effectively punish a low-quality prison. One may also offer transfer possibilities at regular intervals—like “open enrollment periods” for health plans or the natural re-enrollment periods in schools based on the school year—perhaps every few years. One could even imagine transfers at will, though this would involve greater administrative and transportation costs.
Current policies on involuntary transfers could be kept exactly the same as before, though unless the purpose of the transfer is to fight gang activity or otherwise to maintain safety, principles of prison choice suggest that Departments of Corrections should still look for volunteers.
Choice Is Not the Same as Privatization
As I’ve mentioned above, arguments about choice are often merged with arguments about privatization. Privatization skeptics may thus also be skeptics about choice. But this needn’t be.
In the school choice debate, a prominent question has been whether religious schools should be allowed to participate. This question presumes that the school choice plan includes private schools, since a religious public school would be facially unconstitutional. More generally, arguments for school choice often include arguments in favor of private schools. This is because the factors that are claimed to make choice work—chiefly flexibility, cost savings, and responsiveness to market incentives—seem to be more present in the private than in the public sector.
But choice needn’t have anything to do with private provision. These are logically distinct policies.
First, let’s ignore private providers.
- Without choice, everyone could be assigned to a particular public school, perhaps their local one. This also describes the current system of incarceration in states without prison privatization, as well as state-funded indigent defense in most places.
- Or one could have a choice program within the public system alone, for instance vouchers limited to public schools. Magnet and charter schools are also a form of public-school choice.
Now let’s suppose some services are provided privately.
- Even then, one could have privatization without choice. For instance, a private company, like Edison, could become the superintendent of an existing (choice-less) public school system. This also describes the W-2 program for welfare in Wisconsin, or the current prison privatization regime, or assigned counsel for indigent defendants, or a regional health facility to which the state has delegated its entire responsibility for its health care duties under the Medicaid statute.
- Or one could have choice within a regime of (partly or wholly) private provision. Food Stamps are a classic example of such a program, since the government plays a minor role in food distribution. Consider also drug or alcohol abuse rehabilitation, or traffic school, where defendants may be sentenced to the program but most or all providers are private. In England and Wales, and also in Ontario, indigent defendants choose their own defense attorneys using public funds. And, of course, school choice proposals generally include private schools, but public schools remain available.
This last option—choice together with private provision—is the one imagined most often, and in fact I’ll often use private-sector examples in what follows. But it should be clear that we can have privatization without choice, and choice without privatization. All four possible schemes exist in the real world. One can thus talk about choice, whether in schools, prisons, or anywhere else, even if one is hostile to private provision.
As with school vouchers, the voucher amount could simply be equal to (or some percentage of) the cost of incarceration at public prisons. If the program is to include private prisons, the percentage would have to be high enough to induce enough prisons to want to participate, so as to get meaningful competition.
But the voucher amount could be determined in more complicated ways. The formula could be based on as many observable characteristics of the inmate as are permissible to consider—for instance, sex, age, security level, nature of the crime, known psychological or medical conditions, and known history of violence.
The voucher amount could even be different depending on what prison it is redeemed at. It could depend on how many inmates the prison already has; thus, if initial inmates are expensive but additional inmates (up to some limit related to the capacity of the prison) are cheaper to serve, the voucher amount at a prison could start high and decline as the number of inmates increases. But this approach would have costs of its own: it would require that the government monitor prisons’ costs, and it would discourage the expansion of successful prisons.
Before talking further about funding, we should determine how (if at all) prisons should be able to pick and choose among inmates. Clearly all prisons can’t choose the inmates they prefer, because what if they all rejected the same inmate? So there must be at least one prison of last resort, perhaps a public prison—though one could also imagine a private firm willing to play this role. If some prisons can choose, they’ll probably be better informed than the government about the characteristics of the inmate, if for no other reason than that some factors that are probably relevant to the cost of incarceration, like race, will probably be impermissible for the government to consider. They’ll thus consider more factors than whatever formula determines the voucher amount.
Prisons that can choose will then systematically reject inmates with insufficient vouchers, which puts the burden of incarcerating them on whoever can’t refuse a prisoner, that is, the public prison of last resort. Perhaps in future rounds of voucher amount determination, the government can correct its past mistakes, but in general prisons will be ahead of the government in the arms race of figuring out an inmate’s true cost of incarceration.
Therefore, it probably makes sense to require participating prisons to take all comers; and if a prison has a waiting list, perhaps prisons should have to admit inmates by lottery. Prisons might still be allowed to serve particular categories of persons—a company might specialize in women’s prisons, or prisons for inmates with particular medical or psychological problems—but at least within those categories, they shouldn’t be able to pick and choose.
Then the only burden on the government is to make the voucher amount for each prisoner category generous enough that, on average, inmates in that category are worthwhile to incarcerate.
So far, we’ve been assuming that—however determined—the voucher amount for a given inmate type at a given prison is a flat fee. But voucher amounts could be even more complicated. One way of solving the problem of prisons rejecting particular inmates would be to institute an auction system, in which prisons would bid on each prisoner, who would then be issued a voucher sufficient to cover a certain number of bids.
This might be administratively too complicated, and would also require a prison of last resort. Or, instead of a flat fee, one could imagine a “per service” voucher amount (by number of medical visits, disruptions, etc.). But, assuming the per-service amount is generous enough to exceed the cost of providing the service, this would give prisons incentives to oversupply the service, which would require the government to incur heavy monitoring costs.
I only mention these briefly to flag the possibility, though from now on I’ll assume that vouchers are structured as a flat fee, with prisons having no ability to reject inmates within the category they’re serving.
The voucher would replace the current arrangements by which private prisons are reimbursed. If an inmate chooses a particular prison, the prison gains the voucher amount as revenue, and if the inmate transfers out, the prison loses that amount.
For public prisons, it’s less clear. Public schools under voucher plans often continue to be funded out of general revenues, without any explicit accounting for how many students attend the school. Whether a public school loses money when it loses a voucher student to a private school thus depends on the details of the school finance system. Some school voucher plans are structured so that public schools don’t suffer at all from losing students. “[P]seudo-choice plans” are common, where “money does not follow students or so little money follows students that a school accepting an extra student cannot cover its marginal costs; plans in which schools are not able to enter, expand, contract, or exit; plans in which schools need to seek approval or financial support from other schools with which they are supposed to compete; and so on.”
Successful schools have had their funding decreased, and vice versa.
The public prison funding system under a voucher system should probably avoid such perverse incentives. If public schools, or prisons, don’t lose money when they lose a “customer,” we shouldn’t expect competition to improve the quality of the public system. If they don’t lose money, this increases the total cost of the system; but it could also be a political concession to, say, public employees, by limiting the public system’s losses. I would suggest that public and private prisons be funded by vouchers in the same way.
The analogy with schools (or food stamps) suggests that prisoners could be allowed to supplement their voucher with their own funds. But this probably isn’t a good idea on ethical grounds—I’ve noted the inegalitarianism of California’s “offender self-pay” program above —despite its possible efficiency benefits.
Suppose the prison choice plan includes private prisons. What regulations would govern them? We may focus on two possibilities:
- anyone may start up a prison, subject to certain security requirements; or,
- (more modestly) the government may choose who may operate a prison, just like now, but the allocation of prisoners to prisons would proceed by choice instead of by bureaucratic assignment.
By analogy to schools, one could imagine a system of “charter prisons”—prisons that are public but that have more independence from the Department of Corrections. Like charter schools, charter prisons could operate whether or not there are private providers. This could improve the possibilities for choice to improve public prisons.
Once the universe of providers is determined as above, prisons could be governed by most of the same statutes and regulations that they have now—for instance, guard training and the like.
One sort of statute that may now be moot would be the sort that requires private prisons to achieve particular cost savings or quality improvements relative to public prisons.
The quality improvement requirement would be replaced by prisoner choice—quality would no longer be defined by a yardstick imposed by someone else (like the Logan quality of confinement index ), and each prison could pursue its own vision of quality, just as each prisoner could have his own view of what constitutes quality. Litigation over quality levels (for instance, if there were a guarantee of floor space or grievance procedures) would take place at the level of contract law—arguably, prisoners would be in a better position litigating as contractual partners than as prisoners.
Nonetheless, nothing prevents some amount of quality regulation (input, output, or outcome) as a floor, just as there is regulation of private schools. Under a voucher system, there is actually an argument for government regulation to provide a ceiling for quality, lest prisons become so good from the prisoners’ perspective that its deterrent value is reduced; and increases in “quality” (from the prisoners’ perspective) that are socially harmful will also have to be regulated. But more on this later.
As for cost savings, this requirement was probably unnecessary anyway, since there’s already a strong incentive to cut costs under any fixed-reimbursement scheme (whether under choice or not). In addition, the government can reduce the voucher amount if it believes there’s enough competition for prisoners at a given amount.
The voucher system could also change how prisons enter and exit, contract and expand. Currently, private prison firms win contracts and build prisons to fulfill them, or build the prisons in advance, hoping to win the contracts to use them. Under a voucher system, firms would still need to get local building or zoning permission, but otherwise they’d only need to convince the pris-oners themselves. As noted above, popular prisons would have waiting lists and be able to expand without having to ask permission from the Department of Corrections; unpopular prisons would contract or close down.
* * *
More next time!