Wilkinson v. Austin - 04-495 (2005)
OPINION OF THE COURT
WILKINSON V. AUSTIN
545 U. S. ____ (2005)
SUPREME COURT OF THE UNITED STATES
REGINALD A. WILKINSON, DIRECTOR, OHIO DEPARTMENT OF REHABILITATION AND CORRECTION, et al., PETITIONERS v. CHARLES E. AUSTIN et al.
on writ of certiorari to the united states court of appeals for the sixth circuit
[June 13, 2005]
Justice Kennedy delivered the opinion of the Court.
This case involves the process by which Ohio classifies prisoners for placement at its highest security prison, known as a “Supermax” facility. Supermax facilities are maximum-security prisons with highly restrictive conditions, designed to segregate the most dangerous prisoners from the general prison population. We must consider what process the Fourteenth Amendment to the United States Constitution requires Ohio to afford to inmates before assigning them to Supermax. We hold that the procedures Ohio has adopted provide sufficient procedural protection to comply with due process requirements.
The use of Supermax prisons has increased over the last 20 years, in part as a response to the rise in prison gangs and prison violence. See generally U. S. Dept. of Justice, National Institute of Corrections, C. Riveland, Supermax Prisons: Overview and General Considerations 1 (1999), http://www.nicic.org/pubs/1999/014937.pdf (as visited June 29, 2005, and available in Clerk of Court’s case file). About 30 States now operate Supermax prisons, in addition to the two operated by the Federal Government. See Brief for United States as Amicus Curiae 2. In 1998, Ohio opened its only Supermax facility, the Ohio State Penitentiary (OSP), after a riot in one of its maximum-security prisons. OSP has the capacity to house up to 504 inmates in single-inmate cells and is designed to “ ‘separate the most predatory and dangerous prisoners from the rest of the . . . general [prison] population.’ ” See 189 F. Supp. 2d 719, 723 (ND Ohio 2002) (Austin I) (quoting deposition of R. Wilkinson, pp. 24–25).
Conditions at OSP are more restrictive than any other form of incarceration in Ohio, including conditions on its death row or in its administrative control units. The latter are themselves a highly restrictive form of solitary confinement. See Austin I, supra, 724–725, and n. 5 (citing Ohio Admin. Code §5120–9–13 (2001) (rescinded 2004)). In the OSP almost every aspect of an inmate’s life is controlled and monitored. Inmates must remain in their cells, which measure 7 by 14 feet, for 23 hours per day. A light remains on in the cell at all times, though it is sometimes dimmed, and an inmate who attempts to shield the light to sleep is subject to further discipline. During the one hour per day that an inmate may leave his cell, access is limited to one of two indoor recreation cells.
Incarceration at OSP is synonymous with extreme isolation. In contrast to any other Ohio prison, including any segregation unit, OSP cells have solid metal doors with metal strips along their sides and bottoms which prevent conversation or communication with other inmates. All meals are taken alone in the inmate’s cell instead of in a common eating area. Opportunities for visitation are rare and in all events are conducted through glass walls. It is fair to say OSP inmates are deprived of almost any environmental or sensory stimuli and of almost all human contact.
Aside from the severity of the conditions, placement at OSP is for an indefinite period of time, limited only by an inmate’s sentence. For an inmate serving a life sentence, there is no indication how long he may be incarcerated at OSP once assigned there. Austin I, supra, at 740. Inmates otherwise eligible for parole lose their eligibility while incarcerated at OSP. 189 F. Supp. 2d, at 728.
Placement at OSP is determined in the following manner: Upon entering the prison system, all Ohio inmates are assigned a numerical security classification from level 1 through level 5, with 1 the lowest security risk and 5 the highest. See Brief for Petitioners 7. The initial security classification is based on numerous factors (e.g., the nature of the underlying offense, criminal history, or gang affiliation) but is subject to modification at any time during the inmate’s prison term if, for instance, he engages in misconduct or is deemed a security risk. Ibid. Level 5 inmates are placed in OSP, and levels 1 through 4 inmates are placed at lower security facilities throughout the State. Ibid.
Ohio concedes that when OSP first became operational, the procedures used to assign inmates to the facility were inconsistent and undefined. For a time, no official policy governing placement was in effect. See Austin I, supra, at 726–727. Haphazard placements were not uncommon, and some individuals who did not pose high-security risks were designated, nonetheless, for OSP. In an effort to establish guidelines for the selection and classification of inmates suitable for OSP, Ohio issued Department of Rehabilitation and Correction Policy 111–07 (Aug. 31, 1998). This policy has been revised at various points but relevant here are two versions: the “Old Policy” and the “New Policy. ” The Old Policy took effect on January 28, 1999, but problems with assignment appear to have persisted even under this written set of standards. 189 F. Supp. 2d, at 728–736. After forming a committee to study the matter and retaining a national expert in prison security, Ohio promulgated the New Policy in early 2002. The New Policy provided more guidance regarding the factors to be considered in placement decisions and afforded inmates more procedural protection against erroneous placement at OSP.
Although the record is not altogether clear regarding the precise manner in which the New Policy operates, we construe it based on the policy’s text, the accompanying forms, and the parties’ representations at oral argument and in their briefs. The New Policy appears to operate as follows: A classification review for OSP placement can occur either (1) upon entry into the prison system if the inmate was convicted of certain offenses, e.g., organized crime, or (2) during the term of incarceration if an inmate engages in specified conduct, e.g., leads a prison gang. App. 42–43. The review process begins when a prison official prepares a “Security Designation Long Form” (Long Form). Id., at 20. This three-page form details matters such as the inmate’s recent violence, escape attempts, gang affiliation, underlying offense, and other pertinent details. Id., at 20, 38–45.
A three-member Classification Committee (Committee) convenes to review the proposed classification and to hold a hearing. At least 48 hours before the hearing, the inmate is provided with written notice summarizing the conduct or offense triggering the review. Id., at 22, 58. At the time of notice, the inmate also has access to the Long Form, which details why the review was initiated. See Tr. of Oral Arg. 13–17. The inmate may attend the hearing, may “offer any pertinent information, explanation and/or objections to [OSP] placement,” and may submit a written statement. App. 22. He may not call witnesses.
If the Committee does not recommend OSP placement, the process terminates. Id., at 62, 65. See also Brief for Petitioners 9. If the Committee does recommend OSP placement, it documents the decision on a “Classification Committee Report” (CCR), setting forth “the nature of the threat the inmate presents and the committee’s reasons for the recommendation,” App. 64, as well as a summary of any information presented at the hearing. Id., at 59–65. The Committee sends the completed CCR to the warden of the prison where the inmate is housed or, in the case of an inmate just entering the prison system, to another designated official. Id., at 23.
If, after reviewing the CCR, the warden (or the designated official) disagrees and concludes that OSP is inappropriate, the process terminates and the inmate is not placed in OSP. If the warden agrees, he indicates his approval on the CCR, provides his reasons, and forwards the annotated CCR to the Bureau of Classification (Bureau) for a final decision. Id., at 64. (The Bureau is a body of Ohio prison officials vested with final decisionmaking authority over all Ohio inmate assignments.) The annotated CCR is served upon the inmate, notifying him of the Classification Committee’s and warden’s recommendations and reasons. Id., at 65. The inmate has 15 days to file any objections with the Bureau of Classification. Ibid.
After the 15-day period, the Bureau of Classification reviews the CCR and makes a final determination. If it concludes OSP placement is inappropriate, the process terminates. If the Bureau approves the warden’s recommendation, the inmate is transferred to OSP. The Bureau’s chief notes the reasons for the decision on the CCR, and the CCR is again provided to the inmate. Ibid.
Inmates assigned to OSP receive another review within 30 days of their arrival. That review is conducted by a designated OSP staff member, who examines the inmate’s file. Id., at 25. If the OSP staff member deems the inmate inappropriately placed, he prepares a written recommendation to the OSP warden that the inmate be transferred to a lower security institution. Brief for Petitioners 9; App. 25. If the OSP warden concurs, he forwards that transfer recommendation to the Bureau of Classification for appropriate action. If the inmate is deemed properly placed, he remains in OSP and his placement is reviewed on at least an annual basis according to the initial three-tier classification review process outlined above. Brief for Petitioners 9–10.
This action began when a class of current and former OSP inmates brought suit under Rev. Stat. §1979, 42 U. S. C. §1983, in the United States District Court for the Northern District of Ohio against various Ohio prison officials. We refer to the class of plaintiff inmates, respondents here, collectively as “the inmates.” We refer to the prison officials, petitioners here, as “Ohio.”
The inmates’ complaint alleged that Ohio’s Old Policy, which was in effect at the time the suit was brought, violated due process. In addition the inmates brought a claim that certain conditions at OSP violated the Eighth Amendment’s ban on cruel and unusual punishments, but that claim was settled in the District Court. The extent to which the settlement resolved the practices that were the subject of the inmates’ Eighth Amendment claim is unclear but, in any event, that issue is not before us. The inmates’ suit sought declaratory and injunctive relief. On the eve of trial Ohio promulgated its New Policy and represented that it contained the procedures to be followed in the future. The District Court and Court of Appeals evaluated the adequacy of the New Policy, and it therefore forms the basis for our determination here.
After an 8-day trial with extensive evidence, including testimony from expert witnesses, the District Court made findings and conclusions and issued a detailed remedial order. First, relying on this Court’s decision in Sandin v. Conner, 515 U. S. 472 (1995), the District Court found that the inmates have a liberty interest in avoiding assignment to OSP. Austin I, 189 F. Supp. 2d, at 738–740. Second, the District Court found Ohio had denied the inmates due process by failing to afford a large number of them notice and an adequate opportunity to be heard before transfer; failing to give inmates sufficient notice of the grounds serving as the basis for their retention at OSP; and failing to give the inmates sufficient opportunity to understand the reasoning and evidence used to retain them at OSP. Id., at 749. Third, the District Court held that, although Ohio’s New Policy provided more procedural safeguards than its Old Policy, it was nonetheless inadequate to meet procedural due process requirements. Id., at 736, 750–754. In a separate order it directed extensive modifications to that policy. 204 F. Supp. 2d 1024 (ND Ohio 2002) (Austin II).
The modifications the District Court ordered to Ohio’s New Policy included both substantive and procedural reforms. The former narrowed the grounds that Ohio could consider in recommending assignment to OSP. For instance, possession of drugs in small amounts, according to the District Court, could not serve as the basis for an OSP assignment. Id., at 1028. The following are some of the procedural modifications the District Court ordered:
(1)Finding that the notice provisions of Ohio’s New Policy were inadequate, the District Court ordered Ohio to provide the inmates with an exhaustive list of grounds believed to justify placement at OSP and a summary of all evidence upon which the Classification Committee would rely. Matters not so identified, the District Court ordered, could not be considered by the Committee. Id., at 1026.
(2)The District Court supplemented the inmate’s opportunity to appear before the Committee and to make an oral or written statement by ordering Ohio to allow inmates to present documentary evidence and call witnesses before the Committee, provided that doing so would not be unduly hazardous or burdensome. The District Court further ordered that Ohio must attempt to secure the participation of any witness housed within the prison system. Id., at 1026–1027.
(3)Finding the New Policy’s provision of a brief statement of reasons for a recommendation of OSP placement inadequate, the District Court ordered the Classification Committee to summarize all evidence supporting its recommendation. Id., at 1027. Likewise, the District Court ordered the Bureau of Classification to prepare a “detailed and specific” statement “set[ting] out all grounds” justifying OSP placement including “facts relied upon and reasoning used.” Id., at 1027. The statement shall “not use conclusory,” “vague,” or “boilerplate language,” and must be delivered to the inmate within five days. Id., at 1027–1028.
(4)The District Court supplemented the New Policy’s 30-day and annual review processes, ordering Ohio to notify the inmate twice per year both in writing and orally of his progress toward a security level reduction. Specifically, that notice must “advise the inmate what specific conduct is necessary for that prisoner to be reduced from Level 5 and the amount of time it will take before [Ohio] reduces the inmate’s security level classification.” Ibid.
Ohio appealed. First, it maintained that the inmates lacked a constitutionally protected liberty interest in avoiding placement at OSP. Second, it argued that, even assuming a liberty interest, its New Policy provides constitutionally adequate procedures and thus the District Court’s modifications were unnecessary. The Court of Appeals for the Sixth Circuit affirmed the District Court’s conclusion that the inmates had a liberty interest in avoiding placement at OSP. 372 F. 3d 346, 356 (2004). The Court of Appeals also affirmed the District Court’s procedural modifications in their entirety. Id., at 359–360. Finally, it set aside the District Court’s far-reaching substantive modifications, concluding they exceeded the scope of the District Court’s authority. This last aspect of the Court of Appeals’ ruling is not the subject of review in this Court.
We granted certiorari to consider what process an inmate must be afforded under the Due Process Clause when he is considered for placement at OSP. 543 U. S. ___ (2004). For reasons discussed below, we conclude that the inmates have a protected liberty interest in avoiding assignment at OSP. We further hold that the procedures set forth in the New Policy are sufficient to satisfy the Constitution’s requirements; it follows, then, that the procedural modifications ordered by the District Court and affirmed by the Court of Appeals were in error.
Withdrawing from the position taken in the Court of Appeals, Ohio in its briefs to this Court conceded that the inmates have a liberty interest in avoiding assignment at OSP. See Pet. for Cert. i; Brief for Petitioners i. The United States, supporting Ohio as amicus curiae, disagrees with Ohio’s concession and argues that the inmates have no liberty interest in avoiding assignment to a prison facility with more restrictive conditions of confinement. See Brief for United States 10. At oral argument Ohio initially adhered to its earlier concession, see Tr. of Oral Arg. 5, but when pressed, the State backtracked. See id., at 6–7. We need reach the question of what process is due only if the inmates establish a constitutionally protected liberty interest, so it is appropriate to address this threshold question at the outset.
The Fourteenth Amendment’s Due Process Clause protects persons against deprivations of life, liberty, or property; and those who seek to invoke its procedural protection must establish that one of these interests is at stake. A liberty interest may arise from the Constitution itself, by reason of guarantees implicit in the word “liberty,” see, e.g., Vitek v. Jones, 445 U. S. 480, 493–494 (1980) (liberty interest in avoiding involuntary psychiatric treatment and transfer to mental institution), or it may arise from an expectation or interest created by state laws or policies, see, e.g., Wolff v. McDonnell, 418 U. S. 539, 556–558 (1974) (liberty interest in avoiding withdrawal of state-created system of good-time credits).
We have held that the Constitution itself does not give rise to a liberty interest in avoiding transfer to more adverse conditions of confinement. Meachum v. Fano, 427 U. S. 215, 225 (1976) (no liberty interest arising from Due Process Clause itself in transfer from low- to maximum-security prison because “[c]onfinement in any of the State’s institutions is within the normal limits or range of custody which the conviction has authorized the State to impose”). We have also held, however, that a liberty interest in avoiding particular conditions of confinement may arise from state policies or regulations, subject to the important limitations set forth in Sandin v. Conner, 515 U. S. 472 (1995).
Sandin involved prisoners’ claims to procedural due process protection before placement in segregated confinement for 30 days, imposed as discipline for disruptive behavior. Sandin observed that some of our earlier cases, Hewitt v. Helms, 459 U. S. 460 (1983), in particular, had employed a methodology for identifying state-created liberty interests that emphasized “the language of a particular [prison] regulation” instead of “the nature of the deprivation.” Sandin, 515 U. S., at 481. In Sandin, we criticized this methodology as creating a disincentive for States to promulgate procedures for prison management, and as involving the federal courts in the day-to-day management of prisons. Id., at 482–483. For these reasons, we abrogated the methodology of parsing the language of particular regulations.
“[T]he search for a negative implication from mandatory language in prisoner regulations has strayed from the real concerns undergirding the liberty protected by the Due Process Clause. The time has come to return to the due process principles we believe were correctly established in and applied in Wolff and Meachum. Following Wolff, we recognize that States may under certain circumstances create liberty interests which are protected by the Due Process Clause. But these interests will generally be limited to freedom from restraint which, while not exceeding the sentence in such an unexpected manner as to give rise to protection by the Due Process Clause of its own force, nonetheless imposes atypical and significant hardship on the inmate in relation to the ordinary incidents of prison life.” Id., at 483–484 (citations and footnote omitted).
After Sandin, it is clear that the touchstone of the inquiry into the existence of a protected, state-created liberty interest in avoiding restrictive conditions of confinement is not the language of regulations regarding those conditions but the nature of those conditions themselves “in relation to the ordinary incidents of prison life.” Id., at 484.
Applying this refined inquiry, Sandin found no liberty interest protecting against a 30-day assignment to segregated confinement because it did not “present a dramatic departure from the basic conditions of [the inmate’s] sentence.” Id., at 485. We noted, for example, that inmates in the general population experienced “significant amounts of ‘lockdown time’ ” and that the degree of confinement in disciplinary segregation was not excessive. Id., at 486. We did not find, moreover, the short duration of segregation to work a major disruption in the inmate’s environment. Ibid.
The Sandin standard requires us to determine if assignment to OSP “imposes atypical and significant hardship on the inmate in relation to the ordinary incidents of prison life.” Id., at 484. In Sandin’s wake the Courts of Appeals have not reached consistent conclusions for identifying the baseline from which to measure what is atypical and significant in any particular prison system. Compare e.g., Beverati v. Smith, 120 F. 3d 500, 504 (CA4 1997), and Keenan v. Hall, 83 F. 3d 1083, 1089 (CA9 1996), with Hatch v. District of Columbia, 184 F. 3d 846, 847 (CADC 1999). See also Wagner v. Hanks, 128 F. 3d 1173, 1177 (CA7 1997). This divergence indicates the difficulty of locating the appropriate baseline, an issue that was not explored at length in the briefs. We need not resolve the issue here, however, for we are satisfied that assignment to OSP imposes an atypical and significant hardship under any plausible baseline.
For an inmate placed in OSP, almost all human contact is prohibited, even to the point that conversation is not permitted from cell to cell; the light, though it may be dimmed, is on for 24 hours; exercise is for 1 hour per day, but only in a small indoor room. Save perhaps for the especially severe limitations on all human contact, these conditions likely would apply to most solitary confinement facilities, but here there are two added components. First is the duration. Unlike the 30-day placement in Sandin, placement at OSP is indefinite and, after an initial 30-day review, is reviewed just annually. Second is that placement disqualifies an otherwise eligible inmate for parole consideration. Austin I, 189 F. Supp. 2d, at 728. While any of these conditions standing alone might not be sufficient to create a liberty interest, taken together they impose an atypical and significant hardship within the correctional context. It follows that respondents have a liberty interest in avoiding assignment to OSP. Sandin, supra, at 483.
OSP’s harsh conditions may well be necessary and appropriate in light of the danger that high-risk inmates pose both to prison officials and to other prisoners. See infra, at 15–16. That necessity, however, does not diminish our conclusion that the conditions give rise to a liberty interest in their avoidance.
A liberty interest having been established, we turn to the question of what process is due an inmate whom Ohio seeks to place in OSP. Because the requirements of due process are “flexible and cal[l] for such procedural protections as the particular situation demands,” Morrissey v. Brewer, 408 U. S. 471, 481 (1972), we generally have declined to establish rigid rules and instead have embraced a framework to evaluate the sufficiency of particular procedures. The framework, established in Mathews v. Eldridge, 424 U. S. 319 (1976), requires consideration of three distinct factors:
“First, the private interest that will be affected by the official action; second, the risk of an erroneous deprivation of such interest through the procedures used, and the probable value, if any, of additional or substitute procedural safeguards; and finally, the Government’s interest, including the function involved and the fiscal and administrative burdens that the additional or substitute procedural requirement would entail.” Id., at 335.
The Court of Appeals upheld the District Court’s procedural modifications under the assumption that Sandin altered the first Mathews factor. It reasoned that, “[i]n this first factor, Sandin affects the due process balance: because only those conditions that constitute ‘atypical and significant hardships’ give rise to liberty interests, those interests will necessarily be of a weight requiring greater due process protection.” 372 F. 3d, at 358–359. This proposition does not follow from Sandin. Sandin concerned only whether a state-created liberty interest existed so as to trigger Mathews balancing at all. Having found no liberty interest to be at stake, Sandin had no occasion to consider whether the private interest was weighty vis-ŕ-vis the remaining Mathews factors.
Applying the three factors set forth in Mathews, we find Ohio’s New Policy provides a sufficient level of process. We first consider the significance of the inmate’s interest in avoiding erroneous placement at OSP. Prisoners held in lawful confinement have their liberty curtailed by definition, so the procedural protections to which they are entitled are more limited than in cases where the right at stake is the right to be free from confinement at all. See, e.g., Gerstein v. Pugh, 420 U. S. 103 (1975); Wolff, 418 U. S. 539. The private interest at stake here, while more than minimal, must be evaluated, nonetheless, within the context of the prison system and its attendant curtailment of liberties.
The second factor addresses the risk of an erroneous placement under the procedures in place, and the probable value, if any, of additional or alternative procedural safeguards. The New Policy provides that an inmate must receive notice of the factual basis leading to consideration for OSP placement and a fair opportunity for rebuttal. Our procedural due process cases have consistently observed that these are among the most important procedural mechanisms for purposes of avoiding erroneous deprivations. See Greenholtz v. Inmates of Neb. Penal and Correctional Complex, 442 U. S. 1, 15 (1979); Cleveland Bd. of Ed. v. Loudermill, 470 U. S. 532, 543 (1985); Fuentes v. Shevin, 407 U. S. 67, 80 (1972) (“For more than a century the central meaning of procedural due process has been clear: ‘Parties whose rights are to be affected are entitled to be heard; and in order that they may enjoy that right they must first be notified’ ” (quoting Baldwin v. Hale, 1 Wall. 223, 233 (1864))). Requiring officials to provide a brief summary of the factual basis for the classification review and allowing the inmate a rebuttal opportunity safeguards against the inmate’s being mistaken for another or singled out for insufficient reason. In addition to having the opportunity to be heard at the Classification Committee stage, Ohio also invites the inmate to submit objections prior to the final level of review. This second opportunity further reduces the possibility of an erroneous deprivation.
Although a subsequent reviewer may overturn an affirmative recommendation for OSP placement, the reverse is not true; if one reviewer declines to recommend OSP placement, the process terminates. This avoids one of problems apparently present under the Old Policy, where, even if two levels of reviewers recommended against placement, a later reviewer could overturn their recommendation without explanation.
If the recommendation is OSP placement, Ohio requires that the decisionmaker provide a short statement of reasons. This requirement guards against arbitrary decisionmaking while also providing the inmate a basis for objection before the next decisionmaker or in a subsequent classification review. The statement also serves as a guide for future behavior. See Greenholtz, supra, at 16.
As we have noted, Ohio provides multiple levels of review for any decision recommending OSP placement, with power to overturn the recommendation at each level. In addition to these safeguards, Ohio further reduces the risk of erroneous placement by providing for a placement review within 30 days of an inmate’s initial assignment to OSP.
The third Mathews factor addresses the State’s interest. In the context of prison management, and in the specific circumstances of this case, this interest is a dominant consideration. Ohio has responsibility for imprisoning nearly 44,000 inmates. Austin I, 189 F. Supp. 2d, at 727. The State’s first obligation must be to ensure the safety of guards and prison personnel, the public, and the prisoners themselves. See Hewitt, 459 U. S., at 473.
Prison security, imperiled by the brutal reality of prison gangs, provides the backdrop of the State’s interest. Clandestine, organized, fueled by race-based hostility, and committed to fear and violence as a means of disciplining their own members and their rivals, gangs seek nothing less than to control prison life and to extend their power outside prison walls. See Brief for State of California et al. as Amici Curiae 6. Murder of an inmate, a guard, or one of their family members on the outside is a common form of gang discipline and control, as well as a condition for membership in some gangs. See, e.g., United States v. Santiago, 46 F. 3d 885, 888 (CA9 1995); United States v. Silverstein, 732 F. 2d 1338, 1341 (CA7 1984). Testifying against, or otherwise informing on, gang activities can invite one’s own death sentence. It is worth noting in this regard that for prison gang members serving life sentences, some without the possibility of parole, the deterrent effects of ordinary criminal punishment may be substantially diminished. See id., at 1343 (“[T]o many inmates of Marion’s Control Unit [a federal Supermax facility,] the price of murder must not be high and to some it must be close to zero”).
The problem of scarce resources is another component of the State’s interest. The cost of keeping a single prisoner in one of Ohio’s ordinary maximum-security prisons is $34,167 per year, and the cost to maintain each inmate at OSP is $49,007 per year. See Austin I, supra, at 734, n. 17. We can assume that Ohio, or any other penal system, faced with costs like these will find it difficult to fund more effective education and vocational assistance programs to improve the lives of the prisoners. It follows that courts must give substantial deference to prison management decisions before mandating additional expenditures for elaborate procedural safeguards when correctional officials conclude that a prisoner has engaged in disruptive behavior.
The State’s interest must be understood against this background. Were Ohio to allow an inmate to call witnesses or provide other attributes of an adversary hearing before ordering transfer to OSP, both the State’s immediate objective of controlling the prisoner and its greater objective of controlling the prison could be defeated. This problem, moreover, is not alleviated by providing an exemption for witnesses who pose a hazard, for nothing in the record indicates simple mechanisms exist to determine when witnesses may be called without fear of reprisal. The danger to witnesses, and the difficulty in obtaining their cooperation, make the probable value of an adversary-type hearing doubtful in comparison to its obvious costs.
A balance of the Mathews factors yields the conclusion that Ohio’s New Policy is adequate to safeguard an inmate’s liberty interest in not being assigned to OSP. Ohio is not, for example, attempting to remove an inmate from free society for a specific parole violation, see, e.g., Morrissey, 408 U. S., at 481, or to revoke good time credits for specific, serious misbehavior, see, e.g., Wolff, 418 U. S., at 539, where more formal, adversary-type procedures might be useful. Where the inquiry draws more on the experience of prison administrators, and where the State’s interest implicates the safety of other inmates and prison personnel, the informal, nonadversary procedures set forth in Greenholtz v. Inmates of Neb. Penal and Correctional Complex, 442 U. S. 1 (1979), and Hewitt v. Helms, 459 U. S. 460 (1983), provide the appropriate model. Greenholtz, supra, at 16 (level of process due for inmates being considered for release on parole includes opportunity to be heard and notice of any adverse decision); Hewitt, supra, at 473–476 (level of process due for inmates being considered for transfer to administrative segregation includes some notice of charges and an opportunity to be heard). Although Sandin abrogated Greenholtz’s and Hewitt’s methodology for establishing the liberty interest, these cases remain instructive for their discussion of the appropriate level of procedural safeguards. Ohio’s New Policy provides informal, nonadversary procedures comparable to those we upheld in Greenholtz and Hewitt, and no further procedural modifications are necessary in order to satisfy due process under the Mathews test. Neither the District Court nor the Court of Appeals should have ordered the New Policy altered.
The effect of the Prison Litigation Reform Act of 1995 (PLRA), in particular 18 U. S. C. §3626(a)(1)(A), in this case has not been discussed at any length in the briefs. In view of our disposition it is unnecessary to address its application here.
Prolonged confinement in Supermax may be the State’s only option for the control of some inmates, and claims alleging violation of the Eighth Amendment’s prohibition of cruel and unusual punishments were resolved, or withdrawn, by settlement in an early phase of this case. Here, any claim of excessive punishment in individual circumstances is not before us.
The complaint challenged OSP assignments under the Old Policy, and the unwritten policies that preceded it, and alleged injuries resulting from those systems. Ohio conceded that assignments made under the Old Policy were, to say the least, imprecise. The District Court found constitutional violations had arisen under those earlier versions, and held that the New Policy would produce many of the same constitutional problems. Austin I, supra, at 749–754. We now hold that the New Policy as described in this opinion strikes a constitutionally permissible balance between the factors of the Mathews framework. If an inmate were to demonstrate that the New Policy did not in practice operate in this fashion, resulting in a cognizable injury, that could be the subject of an appropriate future challenge. On remand, the Court of Appeals, or the District Court, may consider in the first instance what, if any, prospective relief is still a necessary and appropriate remedy for due process violations under Ohio’s previous policies. Any such relief must, of course, satisfy the conditions set forth in 18 U. S. C. §3626(a)(1)(A).
The Court of Appeals was correct to find the inmates possess a liberty interest in avoiding assignment at OSP. The Court of Appeals was incorrect, however, to sustain the procedural modifications ordered by the District Court. The portion of the Court of Appeals’ opinion reversing the District Court’s substantive modifications was not the subject of review upon certiorari and is unaltered by our decision.
The judgment of the Court of Appeals is affirmed in part and reversed in part, and the case is remanded for further proceedings consistent with this opinion.
It is so ordered.