Prison Vouchers — part 5

This is the fifth and last in a series of posts serial-blogging my new article, Prison Vouchers, forthcoming in the Penn Law Review. The first post in the series is here (I recommend reading it for an overview of the entire paper), the second post is here, the third is here, and the fourth is here. 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 speculate on the politics of prison vouchers and conclude.

* * *

The Politics of Prison Vouchers

I have already discussed how, to the extent prison vouchers improve prisons from prisoners’ perspective, the current political dynamic may be completely reversed. Government will actually have to intervene to keep prisons from becoming too “good” (from the prisoners’ perspective).

This may be all the more important in the prison context than in the school context: there seems to be more hope to reform schools politically, without using choice, because at least society at large empathizes with students in failing schools (or at least claims to), whereas prisoners are generally despised, tend to come from communities without a lot of political power, and are themselves (at least felons) often disenfranchised.

This proposal, of course, is subject to the same critique as all reform proposals that propose to remedy a politically insoluble problem. If the problem is politically insoluble, there are presumably political constituencies opposed to remedying the problem. If so, why should this reform, even if it’s perfect, ever be adopted? Section A discusses how prison vouchers could be adopted; Section B speculates on the political fate of vouchers after they are adopted.

An Adoption Coalition

The foregoing discussion gives us clues as to how a prison voucher proposal could succeed politically, or at least how the support for a prison voucher proposal would be different from the support for an arbitrary prison-conditions reform, such that the political infeasibility of the latter need not imply the political infeasibility of the former.

The “cultural cognition” literature suggests how constituencies can be mobilized by a clever packaging of reforms that convinces different groups that their concerns are being taken into account. For instance, Dan Kahan and his co-authors have argued that conservatives during the George H.W. Bush Administration were convinced to support the Clean Air Act Amendments of 1990 because environmental improvements were packaged with an emissions permit trading scheme. Similarly, when nuclear power is packaged as a way to reduce reliance on fossil fuels responsible for global warming, not only do environmentalists become more likely to support nuclear power, but also people who are not generally sympathetic to environmental concerns can become more likely to support action to address global warming.

A similar dynamic could occur with prison vouchers. The left wouldn’t be inclined to support vouchers, but this could change if the program is packaged as a good way to improve prison conditions. Economic conservatives might not be highly interested in improving prison conditions, but they might warm to the idea of vouchers being tested in a new field, both because this would allow greater operation of market forces and because a success of vouchers in prisons may strengthen the case for vouchers in other areas, like education. And while some social conservatives care about prison conditions, and some social conservatives care about vouchers, a leading motivation for social conservatives to support prison vouchers could be that such a program would be the best way for faith-based prisons to operate constitutionally. (A left-right prison-reform coalition isn’t unheard of: it was a bipartisan coalition that enacted RLUIPA, based on reports of the heavy-handed treatment of prisoners’ religious claims.)

And this scenario is possible even if we don’t accept “cultural cognition” theory. Cass Sunstein has disputed Dan Kahan’s account of the Clean Air Act, arguing that conservatives supported emissions trading because the Bush White House was under pressure to pass air quality legislation and emissions trading was the cheapest way to do it. Similarly, there needn’t be anything cultural about voucher support: The left and different wings of the right can support prison vouchers for rational reasons, if they’re convinced that vouchers can improve prison conditions by harnessing market forces and allow for the constitutional operation of faith-based prisons.

This sort of coalition gives an explanation of how vouchers could be enacted—with all the possible beneficial effects for prisoners—even though there isn’t enough of a political constituency to improve prisoner welfare by more direct means. The entire coalition needn’t be motivated by prisoner well-being. And once vouchers are enacted, one no longer needs to “urge” legislators or administrators to act contrary to their political self-interest; rather, the changes happen automatically, through market mechanisms.

Post-Adoption Coalitions and Politics

But if there really is a fundamental (retributivism- or deterrence-minded) constituency opposed to good prison conditions, it won’t stay bottled up. Even if the pro-voucher constituency remains intact, so that vouchers, once adopted, aren’t repealed, we can still expect that the operation of the voucher scheme will be responsive to popular politics.

First, there’s the question of how generously the vouchers are funded. As noted above, school vouchers are generally far below the per-pupil cost of public schools, and it’s unclear how generously prison vouchers would be funded initially. If vouchers indeed improve prison conditions, what would rational voters do?

Most likely, they would decrease funding—when prison spending is more productive, one can take some of the resulting savings and apply it to other social goals, like welfare, education, the military, or tax reduction. Economic conservatives may accept funding decreases to pay for “more worthy causes.” Religious conservatives may accept funding decreases as long as they don’t threaten the viability of faith-based prisons—which may already be cheaper to the extent they’re subsidized by local churches. Retributivism- and deterrence-minded voters who want prison conditions to be bad, though they may not have been part of the coalition that enacted vouchers, will join the coalition to defund them. (Politically, this may manifest itself by retributivist voters’ outrage at certain amenities chosen and enjoyed by inmates.)

These funding decreases will reduce the ultimate extent of the improvement. In the extreme case, perhaps funding will fall so as to keep prisoner welfare constant; all the benefits will thus accrue to the taxpayers, none to the prisoners themselves. This extreme case seems especially likely if prisoner welfare helps no one but prisoners. But perhaps, given that prisoner welfare does have some positive (rehabilitative) spillovers for the rest of society, prisoners may reap at least some part of the gains. (Perhaps society, even retributivists, wanted better prison conditions before, but was unable to implement them because of resource constraints. An improvement in prison productivity would relax those constraints and allow lower levels of prison rape and better medical care.)

But this is all highly speculative. The main point is that the level of funding is endogenous, so defenses of vouchers that rely on a prediction of quality improvements for a given level of funding may be mistaken.

So far, I’ve assumed that the only political avenue open after vouchers is fine-tuning the level of funding. But there are other, regulatory alternatives, which may be more attractive to deterrence- and retributivism-minded voters.

I’ve mentioned above that there’s a continuing role for regulation to prevent prisoners from choosing amenities that are positively harmful, or even amenities that are neutral but dilute the deterrent value of prison. One sensible way of implementing such regulation would be to have an agency—possibly the Department of Corrections or some independent agency —oversee prison offerings and prevent certain amenities from being offered or advertised. There’s probably an overlap between amenities that efficiency-minded voters prefer (where the rehabilitative value outweighs the decrease in deterrence value) and those that retributivist voters prefer (where the amenity is consistent with the moral purposes of punishment). If such a system works, then both deterrence-minded and retributivism-minded voters should have more confidence that, to the extent prisoner welfare increases, it does so in a way that is acceptable. It will therefore not be necessary for retributivists to reduce voucher funding so much as to wipe out all the benefits.

One final point about political dynamics. In other work, I’ve considered the argument that prison privatization increases pro-incarceration political advocacy, on the theory that private prisons, unlike public prisons, benefit from having more bodies and keeping them longer. I’ve argued that this is unlikely: public corrections officers’ unions are already major pro-incarceration lobbyists, and introducing more private prisons could decrease the total amount of political advocacy, since any benefits from lobbying for increased incarceration would have to be shared with the rest of the prison sector.

This conclusion rested critically on the assumption that “targeted” lobbying is difficult—that it is hard for a private prison firm (or for the public corrections officers’ union) to lobby for an increase in incarceration that would benefit it exclusively. If, hypothetically, a single private prison firm (or the whole private-prison industry acting as a bloc) operated all minimum-security prisons while public prisons operated all maximum-security prisons, the private firm would get all the benefit of its lobbying if it advocated a particular pro-incarceration measure that only affected minimum-security prisoners—say increased penalties for white-collar crime. But I considered that possibility to be fairly remote in today’s world, where private firms operate the same range of facilities as the public sector.

However, in a world of vouchers, we might see more and more specialized prisons catering to identifiable niches of prisoners. In such a world, the possibility that private firms might lobby in favor of incarceration might re-emerge as a realistic possibility. Depending on one’s view of self-interested, pro-incarceration lobbying, this possibility might affect one’s view of the desirability of vouchers.


Throughout this Article, I have been taking ideas developed in the education context and applying them to prisons. It’s worth thinking whether prison vouchers can now tell us anything interesting about education.

In today’s political environment, school-voucher proponents and partisans of private education are identified with the right, including the socially conservative right (it helps that most private schools are religious), while the status quo of public schools is identified with the left (though of course the left contains many who favor reforming public education from within). Moreover, reinforcing this dynamic, vouchers are associated with economic arguments, which also (today) tend to be associated with the free-market right.

However, the political valence of prisons is reversed. Prison reformers are usually associated with the left (though privatization proponents, who are usually on the right, do argue that privatization would improve prison quality). The status quo, that is, opposition to prison reform, is associated with the law-and-order right.

Suppose it is true that prison vouchers would improve the well-being of prisoners, particularly as regards prisoner health-care and freedom from assault, sexual and otherwise—short-circuiting the unsympathetic political and judicial processes. And suppose that the negatives I’ve discussed above aren’t too serious, so that vouchers end up actually being a good idea. If so, from a prison-reform perspective, vouchers would have worked a humanitarian miracle. Might the left then reconsider its opposition to vouchers in general . . . including as a tool of school reform?

Perhaps it’s fanciful to think that people would change their minds entirely. After all, education and prisons are different areas, with different policy concerns. It could be that what works in prison reform wouldn’t work in education. But at least, if some people had been opposed to the very idea of vouchers, perhaps on expressive grounds related to a distrust of market-based arguments generally, their opposition could change from a general opposition to vouchers to a specific opposition to school vouchers only. Their opposition might become more sensitive to empirical argument specifically geared toward the school context.

Let’s consider the same issue from the other side. People on the right who are unsympathetic to prisoners may oppose the idea of prison vouchers as giving unwarranted decisionmaking authority to prisoners. Perhaps depriving prisoners of decisionmaking authority is another form of punishment, which one might support on retributivist grounds. Or perhaps this opposition comes from the disconnect between prisoners’ and non-prisoners’ preferences: after all, the social good is not merely composed of the good of prisoners, it’s also composed of the good of victims and people in society as a whole.

But this last point is exactly the argument that the anti-school-voucher left has been making: Where some on the right have treated parent choice as an end in itself, more communitarian arguments have stressed the interests of the children left behind, the “true” interests of children (such as racial balance) that they and their parents may not adequately value, the presence of anti-social values among certain parents, and the interests of society as a whole. Perhaps making the argument against prison vouchers on communitarian grounds would make the pro-school-voucher right better appreciate the communitarian arguments in education.

Let’s move on to the constitutional side. I’ve noted that under a voucher system, the government should have a duty to guarantee a constitutionally compliant spot for those who want one. “Constitutional compliance” means just secularity for Establishment Clause analysis, but for unconstitutional conditions analysis it means observing all the usual constitutional rights, and not offering any “deals” that would be considered impermissible under current doctrine.

So far, this issue rarely, if ever, comes up for schools. Students almost always have the option of going to public school, which, because it’s government-run, is already compliant. But imagine a voucherized world where education is still compulsory but government no longer provides it (or even a non-voucherized world where the government just assigns students to a private school).

It seems that, in such a world, there should be a similar requirement for the government to provide a compliant school experience to any student who wants one, either by running a public school of last resort or by contracting with a private school to provide constitutional rights. This requirement should exist as long as education is compulsory; the government could relieve itself of this duty simply by making education non-compulsory (not an option for prisons, of course).

This seems like it should be the rule, but I doubt that current state action doctrine can get us there. Private schools aren’t state actors on an “inherent public function” theory, nor does the government’s requirement that children attend some school convert every school into a state actor. But the analysis here suggests a possible change to state action doctrine as it relates to education: with respect to their status as custodians of schoolchildren subject to compulsory education, schools should be considered state actors, though they should have enhanced ability to negotiate the waiver of students’ rights in exchange for other benefits as long as a constitutionally compliant alternative is available. The intuitive reason is the same as for prisons—prisons and schools are similar in that, as long as education is compulsory, people are forced to be there.

Even now, the educational system contains both “compulsory” and “voluntary” students, depending on whether they’re above the statutory compulsory education age. A private school, in a fully voucherized world without public schools, might then have some students with whom it must observe due process (based on a constitutional-school-of-last-resort contract with the government), and others students with whom it needn’t.

* * *

Some advocates of vouchers have been extremely optimistic about the ability of vouchers to improve quality of service. The promise of school vouchers, for instance, was to remove education reform from the hands of unresponsive democratic majorities, obstructionist teachers’ unions, and an unsympathetic legal system. As early voucher advocates John Chubb and Terry Moe put it: “Without being too literal about it, we think reformers would do well to entertain the notion that choice is a panacea. . . . It has the capacity all by itself to bring about the kind of transformation that, for years, reformers have been seeking to engineer in myriad other ways.”

I don’t make any such strong claims about prison vouchers. I do believe that prison vouchers, if enacted, could radically change how prisons work; the question is whether it would be for the better. I believe the constitutional effects—making faith-based prisons constitutional and loosening up the unconstitutional conditions doctrine—would probably be positive. But the effects on prison quality are a lot less clear.

The social positives include improvements in the desirable aspects of prisoner welfare, like lower rates of prison rape and better prison medical care, together with the benefits that this would yield for society at large, like greater rehabilitation and thus lower crime rates, less spread of communicable disease, and the like. All of these have been extremely difficult to attain in the current political climate, relying on pro-prisoner legislation, administrative action, or judicial decisions. These benefits are probably substantial.

The negatives include reductions in deterrence from higher prison quality or improvements in the antisocial aspects of prisoner welfare, which can’t be controlled through regulation. These, too, are probably substantial.

If undesirable actions were fully observable, the negative effects of prison vouchers could be controlled by regulation —prevent prisons from competing on ease of escape by penalizing escapes, and prevent prisons from making cell phones available by mandating cell phone jamming technology. If the undesirable actions are those of inmates, not prisons, the regulation could either act on inmates directly (sex offenders can’t transfer into a prison that doesn’t have sex offender treatment), or control inmates indirectly by acting on prisons (prisons get reduced, or zero, voucher revenue from receiving a sex offender if they don’t have sex offender treatment).

The problem arises when undesirable actions are unobservable. It’s hard to tell how much contraband gets into a prison. We don’t know who’s a gang member or gang leader, so we can’t control how gang members move around if they can transfer easily. So to some extent, the negative effects of prison vouchers can be controlled, and to some extent, there will remain some uncontrollable residual that we would just have to live with.

The questions is whether these uncontrollable residual negative effects outweigh the positives. Perhaps they do; perhaps not. I hope this thought experiment stimulates further thought along these lines to investigate whether prison vouchers are an interesting reform proposal after all.