Perry v. Spencer

94 F.4th 136

Opinion

United States Court of Appeals For the First Circuit

No. 16-2444 JWAINUS PERRY,

Plaintiff, Appellant,

v.

LUIS S. SPENCER, Commissioner; THOMAS DICKHAUT, Former Superintendent; ANTHONY M. MENDONSA, Former Deputy of Classification; JAMES J. SABA, Superintendent; ABBE NELLIGAN, Deputy of Classification; PATRICK TOOLIN, Correctional Program Officer; KRISTIE LADOUCEUR; CAROL MICI; THOMAS NEVILLE,

Defendants, Appellees,

JENS SWANSON, Property Officer, Defendant.

APPEAL FROM THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF MASSACHUSETTS

[Hon. M. Page Kelley, Magistrate Judge]

Before

Barron, Chief Judge,

Lynch, Howard, Thompson, Kayatta, Gelpí, Circuit Judges.

Rosalind E. Dillon, with whom Daniel Greenfield, Alexis Garmey Chardon, Roderick & Solange MacArthur Justice Center, and Terry Garmey & Associates were on brief, for appellant.

Nancy Ankers White, Special Assistant Attorney General, with whom Sheryl F. Grant, Counsel, and Mary Eiro-Bartevyan, Counsel, were on brief, for appellees.

Claudia Pare, Melissa Giangrande, Matthew Marchiori, and Hogan Lovells US LLP on brief for amici curiae Former Corrections Officials.

John P. Bueker on brief for amicus curiae Prisoners' Legal Services.

Mary B. McCord, Kelsi Brown Corkran, Amy Marshak, Seth Wayne, Shelby Calambokidis, and Institute for Constitutional Advocacy and Protection on brief for amici curiae Current and Former Prosecutors and Department of Justice Officials.

Nancy Gertner on brief for amici curiae Center for Law, Brain, and Behavior and Neuroscientists.

John J. Butts, Nina B. Garcia, Hannah E. Gelbort, and Wilmer Cutler Pickering Hale and Dorr LLP on brief for amici curiae Terry Kupers, Craig Haney, Pablo Stewart, and Stuart Grassian.

Jennifer A. Wedekind, Carol J. Garvan, Zachary L. Heiden, Gilles R. Bissonnette, Matthew R. Segal, Jessie J. Rossman, Areeba Jibril, and Lynette Labinger on brief for amici curiae American Civil Liberties Union and its state affiliates.

Jaime A. Santos, William E. Evans, and Goodwin Procter LLP on brief for amicus curiae Professor John F. Stinneford.

Clark M. Neily III and Jay R. Schweikert on brief for amicus curiae Cato Institute.

February 21, 2024

Opinion En Banc

BARRON, Chief Judge. This appeal concerns Jwainus Perry's challenge to the grant of summary judgment on his claim under the Due Process Clause of the Fourteenth Amendment to the United States Constitution. See U.S. Const. amend. XIV, § 1 ("[N]or shall any state deprive any person of . . . liberty . . . without due process of law."). Perry brought the claim in the United States District Court for the District of Massachusetts pursuant to 42 U.S.C. § 1983 against officials of the Massachusetts Department of Corrections ("DOC"). He alleges that the DOC officials violated his right to procedural due process by depriving him of "liberty . . . without due process of law" during a portion of his life sentence for first-degree murder under Massachusetts law. He alleges that the DOC officials did so by placing him in solitary confinement for a total of nearly two years without affording him either notice of the factual basis for that confinement or an opportunity for rebuttal.

The District Court granted summary judgment to the defendants based on qualified immunity after determining that, at the relevant time, clearly established law did not treat such prolonged solitary confinement as a deprivation of a liberty interest that the Due Process Clause protects. Perry v. Spencer, No. 12-CV-12070, 2016 WL 5746346, at *16 (D. Mass. Sept. 30, 2016). A panel of this Court affirmed that ruling. Perry v. Spencer, 751 F. App'x 7 (1st Cir. 2018). But the panel also ruled that, even

4

if Perry's confinement did implicate a liberty interest that the Due Process Clause protects, Perry failed to show that he had been denied the process that he was due under clearly established law at the relevant time. Id. at 11. Rehearing the case en banc, we affirm the District Court's grant of summary judgment to the defendants based on qualified immunity. We do so on the ground that, while Perry supportably has shown that his segregated confinement implicated a liberty interest and that the defendants denied him the process that was due to him before depriving him of that interest, there is no basis for concluding on this record that it would have been clear to a reasonable corrections officer that the confinement implicated a liberty interest. In so holding, we clarify both the circumstances in which the use of solitary confinement results in a deprivation of a liberty interest that the Due Process Clause protects and the process that is due in consequence of such a deprivation.

I.

A.

On December 10, 2010, while Perry was confined in the general prison population at the DOC's Souza-Baranowski Correctional Center ("SBCC"), DOC officials "received an anonymous informant letter." Perry, 2016 WL 5746346, at *8. The letter alleged that "Perry had made threats against an inmate who was a suspected gang member," would "'get anybody' from" a different

5

gang "in retaliation for [a] knife fight he had been involved in" two years prior, and had "motivated" another attack that had occurred in the prison the previous month. Id.

The same day, and following the letter's receipt, DOC officials moved Perry from the general prison population within SBCC to a Special Management Unit ("SMU") in that same facility. Id. at *9. The then-operative DOC regulations described an SMU as "[a] separate housing area from general population . . . in which inmates may be confined for reasons of administrative segregation, protective custody, or disciplinary detention." 103 MASS. CODE REGS. § 423.06 (2007).

The parties agree that prior to Perry's placement in the SMU he was seen by medical staff and cleared for SMU placement, as required by the DOC regulations. See id. § 423.08(2)(a). The parties agree that Perry was thereafter held in an SMU for a total of fifteen consecutive months -- spanning his time at both SBCC and another DOC facility, Massachusetts Correctional Institution-Cedar Junction ("MCI-Cedar Junction") -- on "awaiting action status." Perry, 2016 WL 5746346, at *1, *9.

The regulations then in place defined confinement on "awaiting action status" as "confinement of an inmate in an individual cell, which may be . . . in a[n SMU], until an investigation is completed or hearing is held relative to a disciplinary, administrative, or classification matter." 103 MASS.

6

CODE REGS. § 902.01 (2007). The regulations further provided that an inmate may be placed in an SMU in various "instances," such as when the inmate is "pending investigation for disciplinary offenses," "pending transfer," or "for the inmate's own protection."1 Id. § 423.08(1). Perry was told at various points during the fifteen months that he was confined in an SMU at either SBCC or MCI-Cedar Junction that he was so confined "pending investigation," "pending classification," or "pending an out-of- state placement." Perry, 2016 WL 5746346, at *1.

Both the District Court and the panel in this case described the conditions that Perry experienced in the SMUs during the fifteen months as "akin to solitary confinement." Perry, 751 F. App'x at 8; see also Perry, 2016 WL 5746346, at *1 n.3

We note that inmates thought to pose "a substantial threat 1

to the safety of others" or to "the operation of [the] facility" could be placed in a Departmental Segregation Unit ("DSU") instead of an SMU. 103 MASS. CODE REGS. § 421.09 (2007). Under the DSU regulations, inmates could also be placed on "awaiting action" status in "restrictive confinement" while awaiting the Commissioner's decision regarding placement in a DSU, but in such a case, the applicable regulations entitled inmates to receive notice and a hearing or be released within fifteen or thirty days. Id. § 421.08. The defendants contend that these regulatory requirements regarding confinement in a DSU did not apply to Perry's confinement in SMUs. We do not understand Perry's procedural due process claim to be premised solely on the understanding that the DSU regulations did not authorize his segregated confinement, see infra note 12, and insofar as he may be understood to be contending in part that there was a procedural due process violation simply because the DSU regulations did not authorize the confinement, that contention is without merit, see infra Section III.A.1.

7

("Notwithstanding the bureaucratic nomenclature, as Justice Kennedy has said, 'administrative segregation' is better known as 'solitary confinement.'" (quoting Davis v. Ayala, 576 U.S. 257, 286 (2015) (Kennedy, J., concurring))). Specifically, Perry spent up to twenty-three or twenty-four hours per day alone in a windowless cell,2 in which he ate his meals alone and from which he was permitted to leave only for certain limited reasons, including for outdoor exercise alone one hour per day five days per week. Perry, 2016 WL 5746346, at *4 & n.7, *5. He also had limited visitation and phone privileges. Id.

The regulations required Perry to receive "SMU Reviews"

while he was in an SMU. See 103 MASS. CODE REGS. § 423.08(2)(b) (2007). Those reviews ensured that "Perry received periodic written notifications that he was on awaiting action status . . . and that administrative reviews of his placement had been conducted." Perry, 2016 WL 5746346, at *11.

The regulations did not require that Perry receive either notice of the factual basis for his placement in the SMUs on "awaiting action status" or an opportunity to rebut his placement in them on that basis. See 103 MASS. CODE REGS. § 423.08

The parties dispute whether, as Perry asserts, he was 2

sometimes placed on "solid door status" while in the SMUs, meaning that, as opposed to his cell being closed with a grille door, it was closed with a solid door, such that he was surrounded by walls on all four sides.

8

(2007). The regulations also did not place a limit on the time that an inmate could be kept in an SMU on "awaiting action status." See id.

In March 2012, while Perry was still in the SMU at SBCC, he was transferred to an out-of-state prison facility in Connecticut. Perry, 2016 WL 5746346, at *1 & n.4. Approximately six months later, he was sent back to DOC custody, at which point DOC officials returned Perry to the SMU at MCI-Cedar Junction for another four and a half months, from September 2012 to February 2013. Id. Then, too, the parties agree, Perry was held in the SMU on "awaiting action status" and in conditions akin to those that he experienced during his initial fifteen-month period of confinement in SMUs.3 In November 2012, two months after Perry was returned to the SMU at MCI-Cedar Junction, the Supreme Judicial Court of Massachusetts ("SJC") held in LaChance v. Commissioner of Correction that "an inmate confined to administrative segregation on awaiting action status . . . is entitled, as a matter of due process, to notice of the basis on which he is so detained; a

3 Perry does not allege conditions of confinement more severe than those outlined in the regulations. The defendants allege that "from July 28, 2011 through January 12, 2012," while Perry was in the SMU at MCI-Cedar Junction for five months during his initial fifteen-month stint in SMUs, he received some benefits beyond those provided in the regulations, namely "the ability to participate in GED Distance Learning" and "to purchase and possess items" beyond those outlined in the regulations.

9

hearing at which he may contest the asserted rationale for his confinement; and a post[-]hearing written notice explaining the reviewing authority's classification decision." 978 N.E.2d 1199, 1206-07 (Mass. 2012). The SJC also held in LaChance that "in no circumstances may an inmate be held in segregated confinement on awaiting action status for longer than ninety days without a hearing." Id. at 1207. A few months after LaChance was decided, on February 19, 2013, DOC officials returned Perry to the general prison population of a medium-security facility: MCI-Shirley. Perry, 2016 WL 5746346, at *1 n.4.

B.

While Perry was still in the SMU at MCI-Cedar Junction, he filed a handwritten pro se complaint in the United States District Court for the District of Massachusetts on November 5, 2012, against various DOC officials. Later, he filed a first amended complaint on March 5, 2013. Then, over a year later, and with assistance of counsel, he filed a second amended complaint on April 30, 2014, which we refer to as the "complaint."

The complaint alleges that the defendants, collectively, violated Perry's "right to due process, as secured by the Fifth and Fourteenth Amendments to the United States Constitution and 42 U.S.C. § 1983," "his right to equal protection" under the Fourteenth Amendment, and the Americans with Disabilities Act, 42 U.S.C. §§ 12101 et seq., by placing Perry in SMUs in the conditions

10

to which he was subjected for fifteen consecutive months and then over four months thereafter without affording him the process that the Due Process Clause requires. The District Court on February 12, 2015, dismissed all the claims in Perry's complaint except for his procedural due process claim, which the District Court concluded had "set[] forth sufficient factual allegations to warrant discovery." Perry v. Spencer, No. 12-CV-12070, 2015 WL 628538, at *1, *6 (D. Mass. Feb. 12, 2015) (Sorokin, J.).4 Following discovery on Perry's procedural due process claim, the defendants filed a motion in December 2015 for summary judgment based on qualified immunity. The District Court granted that motion. See Perry, 2016 WL 5746346, at *1 (Kelley, U.S.M.J.). A panel of this court affirmed that ruling, Perry, 751 F. App'x at 8, and Perry thereafter filed a petition for rehearing en banc. We granted the petition, vacated the judgment of the panel, requested supplemental briefing, and then heard oral argument following the parties' submission of such briefing.5

The case thereafter proceeded before a magistrate judge upon 4

consent of all the parties. See 28 U.S.C. § 636(c).

5We acknowledge with appreciation the assistance of amici curiae Former Corrections Officials; Prisoners' Legal Services; Current and Former Prosecutors and Department of Justice Officials; the Center for Law, Brain, & Behavior and Neuroscientists; Terry Kupers et al.; the American Civil Liberties Union and its state affiliates; Professor John F. Stinneford; and the Cato Institute.

11

II.

We often follow "a two-step approach" to decide whether a defendant is entitled to summary judgment based on qualified immunity. Stamps v. Town of Framingham, 813 F.3d 27, 34 (1st Cir. 2016). The first step addresses whether there is a genuine issue of disputed fact that would allow a reasonable finder of fact to determine that the defendant violated the plaintiff's federal constitutional rights. See id. If there is, then we move on to the second step, which addresses whether the right that the plaintiff can supportably show was violated was clearly established at the time of the defendant's alleged violation. See id.

We have discretion to bypass the first step if we conclude that the right was not clearly established at the time of its alleged violation. See Tolan v. Cotton, 572 U.S. 650, 656 (2014) (per curiam); Pearson v. Callahan, 555 U.S. 223, 236 (2009) (overruling Saucier v. Katz, 553 U.S. 194 (2001)). We decline to bypass that first step here, however, because of the stakes involved in the use of prolonged solitary confinement and the concomitant need to provide legal clarity in this area. See Pearson, 555 U.S. at 236.6

6 Although Massachusetts prisoners do now enjoy more procedural protections than Perry did, we are not barred here from exercising our discretion to "clarify the legal standards

12

III.

With respect to the first step, we will start by addressing whether Perry has supportably shown that a reasonable juror could find that the defendants deprived him of a "liberty" interest that the Due Process Clause protects. Because we conclude that Perry has made that showing, we then will address whether he has supportably shown that a reasonable juror could find that the defendants denied him the "process" that he was due before depriving him of that liberty interest. We conclude that Perry has made this latter showing as well. Our review at this first step of the analysis is de novo. Wilber v. Curtis, 872 F.3d 15, 20 (1st Cir. 2017). We consider "the evidence in the light most favorable to" Perry and "draw all reasonable inferences in h[is] favor." Hicks v. Johnson, 755 F.3d 738, 743 (1st Cir. 2014).

governing public officials" in this Circuit. Camreta v. Greene, 563 U.S. 692, 707 (2011). Perry's claim presents a question that will not "frequently arise in cases in which a qualified immunity defense is unavailable," Pearson, 555 U.S. at 236, as we have no reason to expect that it will often arise outside actions that, like his, seek damages for at least some time spent in segregated confinement. Our step-two analysis also benefits from our clarifying what the ordinary incidents of prison life are, the kinds of segregated confinement that mark a dramatic departure from them, and what the record shows about the kind of confinement Perry endured. Id. (quoting Lyons v. Xenia, 417 F.3d 565, 581 (6th Cir. 2005) (Sutton, J., concurring)) (recognizing that it is proper to address step one when it is "difficult to decide" whether the claimed right is "clearly established without deciding precisely what the existing constitutional right happens to be")). Moreover, in clarifying the applicable federal constitutional framework, we resolve no contested point of state law. See, e.g., Tremblay v. McClellan, 350 F.3d 195, 200 (1st Cir. 2003).

13

A.

In assessing the "liberty interest" component of the inquiry, we will begin by reviewing the controlling case law that establishes the standard that we must apply to determine whether the solitary confinement at issue implicates such an interest. As we will explain, that standard requires that Perry show that the confinement constitutes an "atypical and significant hardship" relative to the "ordinary incidents of prison life." We then will review the precedent that helps define what constitute the "ordinary incidents prison life," before also addressing the precedent that helps define what constitutes an "atypical and significant hardship." With those pieces of the analysis in place, we will explain why we conclude that Perry has supportably shown that the solitary confinement that he endured constituted an "atypical and significant hardship."

1.

A "liberty" interest that the Due Process Clause protects "may arise from two sources -- the Due Process Clause itself and the laws of the States." Hewitt v. Helms, 459 U.S. 460, 466 (1983) (citing Meachum v. Fano, 427 U.S. 215, 223–27 (1976)). The Supreme Court of the United States has held that the imposition of a condition of confinement in prison implicates the first type of liberty interest -- one rooted in the "Constitution itself, by reason of guarantees implicit in the word 'liberty,'"

14

Wilkinson v. Austin, 545 U.S. 209, 221 (2005) -- only if the condition "exceed[s] the sentence" in a particularly "unexpected manner," Sandin v. Connor, 515 U.S. 472, 484 (1995).

The Court has held that the imposition of a condition of confinement implicates a liberty interest of this first kind in the case of an inmate being placed in a mental hospital pursuant to a transfer. See Vitek v. Jones, 445 U.S. 480, 493 (1980). It has also held that imposition of a condition of confinement implicates a liberty interest of this first kind in the case of an inmate being involuntarily administered psychotropic drugs. See Washington v. Harper, 494 U.S. 210, 221-22 (1990).

The Court has not held that a prison system's segregated confinement of an inmate -- whether for administrative or disciplinary reasons -- implicates this first kind of liberty interest. See Wilkinson, 545 U.S. at 221–22. But the Court has held that the use of segregated confinement in some circumstances implicates the second kind of liberty interest -- that arising "from an expectation or interest created by state laws or policies." Id. at 221 (citing Wolff v. McDonnell, 418 U.S. 539, 556–58 (1974) (finding liberty interest in avoiding loss of good-time credits earned under state-created system)). That is the kind of liberty interest that Perry asserts is implicated here.

The Court's path to this conclusion about segregated confinement begins with Sandin. There, the Court held that the

15

thirty days of segregated confinement that the Hawaii state prison system had imposed on an inmate as a disciplinary measure did not implicate a state-created liberty interest. Sandin, 515 U.S. at 484-87. The Court reached this conclusion on the ground that state-created liberty interests are implicated only by conditions of confinement that "impose[] atypical and significant hardship on the inmate in relation to the ordinary incidents of prison life," id. at 484, and that the segregated confinement at issue did not impose such an "atypical and significant hardship," id. at 486– 87.

Sandin began the analysis that resulted in that key "atypical and significant hardship" formulation by rejecting an earlier line of cases that started with Hewitt. The Hewitt line of authority held that state regulations that restricted the prison administrator's discretion to impose a specific condition of confinement in and of themselves could suffice to show -- by so limiting the administrator's discretion -- that the imposition of the condition implicated a state-created liberty interest of the inmate who was subjected to it. See id. at 479–83 (describing such cases).

Sandin explained that the Hewitt line of authority wrongly "shift[ed] the focus of the liberty interest inquiry to one based on the language of a particular regulation, and not the nature of the deprivation." Id. at 481. Sandin therefore rejected

16

any contention that the limitations that the Hawaii prison system's regulations placed on the discretionary imposition of the segregated confinement at issue in that case in and of themselves sufficed to demonstrate that the imposition of the confinement implicated a state-created liberty interest. See id. at 483.

Sandin went on to assess, however, whether the nature of the segregated confinement at issue nonetheless meant that the imposition of that kind of confinement implicated a state-created liberty interest. The Court then concluded that the imposition of the segregated confinement there at issue did not because it "did not present the type of atypical, significant deprivation in which a State might conceivably create a liberty interest." Id. at 486. Sandin explained that the imposition of such segregated confinement -- given the confinement's nature, including its limited duration, and the prevailing conditions in Hawaii's prison system -- was not a "deprivation" of that kind because it "did not work a major disruption in [the inmate's] environment," id., or a "dramatic departure from the basic conditions" of the inmate's sentence, id. at 485.

Sandin pointed in part to the "significant amounts of 'lockdown time' even for inmates in the general population." Id. at 486. It pointed as well to the "conditions imposed upon inmates in administrative segregation and protective custody," which the

17

Court described as "totally discretionary" forms of confinement.7 Id.

Thereafter, in Wilkinson, the Court applied Sandin's "atypical and significant hardship" test for assessing when the imposition of a condition of confinement implicates a state- created liberty interest. See Wilkinson, 545 U.S. at 223. In doing so, Wilkinson held that placing Ohio inmates in that state's "supermax" prison based on their assessed "threat" levels did so. See id. at 217, 223. The Court explained that such placement resulted in the imposition of "an atypical and significant hardship" under "any plausible baseline." Id. at 223. That was so, according to the Court, because the placement of the inmates in that facility subjected them to conditions at least as severe as those akin to "most solitary confinement facilities"; did so for an indefinite period; and, in doing so, rendered the inmates ineligible for parole. Id. at 224.

7 As Sandin's reference to "administrative segregation"

reveals, the terminology used to describe forms of segregation may matter. In some instances, courts will use "administrative segregation" as Sandin did to distinguish it from either disciplinary segregation or "protective custody," while in other instances courts use the term "administrative segregation" to refer to all forms of non-disciplinary segregation, see, e.g., Skinner v. Cunningham, 430 F.3d 483, 486 (1st Cir. 2005). In using the term "administrative segregation" in what follows, we refer to all non-disciplinary forms of segregation unless otherwise specified.

18

2.

It is clear from this precedent, then, that the determination of when, if ever, the imposition of solitary confinement less onerous than that at issue in Wilkinson implicates a state-created liberty interest is partly a function of the baseline that the "ordinary incidents of prison life" establish. We thus need to describe what those "incidents" are. That is not an easy task, given the available precedent. But, as we will explain, certain clear principles do exist to guide the inquiry.

a.

Wilkinson observed that "the Courts of Appeals have not reached consistent conclusions for identifying the baseline." 545 U.S. at 223. And, following Wilkinson, we similarly described there being no consensus view: "[s]ome circuits compare the confinement conditions to those of the general prison population," by which we meant those faced by inmates not in segregated confinement; "others look to the conditions of nondisciplinary administrative segregation;" and "[the Fifth Circuit] holds that disciplinary segregation never implicates a liberty interest unless it lengthens a sentence." Skinner, 430 F.3d at 486-87.

In addition, we note, the Second Circuit has adopted a fact-specific baseline that, in assessing whether the severity of the conditions makes them not ordinary, looks to the conditions in both administrative segregation and the general population within

19

the prison system at issue. See Sealey v. Giltner, 197 F.3d 578, 588-89 (2d Cir. 1999). And the D.C. Circuit, for its part, looks to the most restrictive conditions routinely imposed for administrative reasons on inmates serving similar sentences in the prison system at issue to determine what the ordinary incidents of prison life are in that system. See Aref v. Lynch, 833 F.3d 242, 254 (D.C. Cir. 2016). Finally, the Seventh Circuit, at least prior to Wilkinson, looked to the conditions of the harshest facility in the prison system at issue. See id. at 253-54 (citing Wagner v. Hanks, 128 F.3d 1173, 1175 (7th Cir. 1997) (Posner, J.)).

We have not purported in the wake of Skinner to assess which, if any, of the baselines described above is the proper one for defining the "ordinary incidents of prison life." Nor have we identified comprehensively what those "incidents" are.8

8 Perry argues that our statement in a case regarding revocation of supervised release that "Sandin's 'atypical hardship' standard remains our lodestar" when "the baseline liberty being deprived is that of the general prison population," González-Fuentes v. Molina, 607 F.3d 864, 889 (1st Cir. 2010) (emphasis added), established as the First Circuit's baseline the conditions to which the "general prison population" is subject when not held in administrative segregation. But that statement served to distinguish prison-confinement cases in which Sandin applies from cases involving "parole-like arrangement[s]" in which Sandin does not apply at all. Id. Thus, González-Fuentes's dicta about the baseline that Sandin deployed does not speak to the question of whether the conditions of "the general prison population" do or do not include periods of administrative segregation to which members of that population may be subjected.

20

We note, however, that although we attempted to describe the various approaches of other circuits in Skinner, a careful review of the state of the precedent from other circuits reveals the limits of relying on the labels that have been ascribed to baselines that the circuits have in fact identified. Indeed, even when a circuit has been identified as having adopted a certain baseline -- and thus a certain account of which incidents of prison life are ordinary -- a fuller review of that circuit's own precedents can reveal that the circuit has not always been consistent in following that baseline. See, e.g., DiMarco v. Wy. Dept. of Corr., 473 F.3d 1334, 1340 (10th Cir. 2007) ("We note[] that our circuit, in analyzing whether segregation is atypical and significant, has used inconsistent standards in applying a comparative baseline.").

We thus find it most fruitful, with respect to the baseline issue, to begin by reviewing the key Supreme Court precedents in this area: Sandin and Wilkinson. As we will next explain, that review reveals certain features of the "ordinary incidents of prison life" -- and thus the baseline -- that other circuits turn out to agree upon, no matter how they otherwise may differ in describing that baseline. And, as we will see, our own approach in Skinner is itself consistent with that out-of-circuit case law, thus affording clarity to the applicable law in that respect.

21

b.

As an initial matter, Sandin and Wilkinson lead us to conclude that the "ordinary incidents of prison life" include only the conditions of confinement for inmates within the state prison system at issue, rather than the conditions of confinement that may be imposed by other states on inmates in their custody.9 This conclusion follows from Sandin's focus on whether the state itself had "create[d]" the asserted liberty interest through the reasonable "expectations" to which the state's own actions (whether in the form of practices or policies) gave rise. Sandin, 515 U.S. at 484, 486 n.9. This conclusion follows as well from

9Although the Seventh Circuit early on suggested in dicta that Sandin's logic "implies" that an inmate's conditions of confinement might also be compared to the conditions in out-of- state prisons to which they might be transferred, the Seventh Circuit explicitly declined to "decide whether logic should be pressed so far." Wagner, 128 F.3d at 1176. Moreover, no circuit has ever so held, and the Seventh Circuit, even when mentioning this dicta, has consistently applied a state-specific standard, see, e.g., Lekas v. Briley, 405 F.3d 602, 609–10 (7th Cir. 2005), and has cast some doubt on whether Wagner survived Wilkinson, see Westefer v. Snyder, 422 F.3d 570, 585, 590 (7th Cir. 2005) (reversing district court's pre-Wilkinson finding of no liberty interest based on Wagner and holding that, even if the conditions at issue did not impose an "atypical and significant hardship under any plausible baseline" as in Wilkinson, the district court would have to "confront the issue of what does constitute the appropriate baseline for the Illinois system"). We add only that a state- specific standard for conditions of confinement makes sense, given that, even if a state creates certain expectations related to out- of-state transfers for inmates in its custody, the conditions that inmates would then face when transferred to those other states are determined solely by those other states themselves and the expectations they in turn create.

22

Sandin's comparison in that case of the conditions of confinement at issue there to only the conditions that other inmates in the same state prison system might face. Id. at 486–87. And this conclusion is further supported by Wilkinson, which, in characterizing the "ordinary incidents of prison life," referred to a "baseline from which to measure what is atypical and significant in any particular prison system." Wilkinson, 545 U.S. at 223 (emphasis added).

Next, a review of Sandin, Wilkinson, and case law from our sister circuits leads us to the conclusion that, within a given prison system, segregated confinement (even when it amounts to solitary confinement) may in some circumstances be an "ordinary incident of prison life." Indeed, Sandin itself concluded that thirty days of segregated confinement did not impose an "atypical and significant hardship" in relation to the "ordinary incidents of prison life." We note that Sandin based the determination that the thirty days of disciplinary segregation at issue there "did not work a major disruption in [the inmate's] environment" in part on a comparison of those conditions to the "conditions imposed upon inmates in administrative segregation and protective custody." Sandin, 515 U.S. at 486. In fact, we know of no circuit that holds that segregated confinement -- regardless of its nature and length -- can never itself be an "ordinary incident of prison life."

23

On the flip side, however, Sandin also indicates that segregation less severe than that involved in Wilkinson is not always an "ordinary incident of prison life," no matter its nature and duration. After all, segregation is often, in practice, a form of solitary confinement, see Davis, 576 U.S. at 286 (Kennedy, J., concurring), and solitary confinement is known to have serious adverse psychological effects on those subjected to it, even when it persists for less than thirty days, see Brief of Former Corrections Officials as Amici Curiae at 18 ("Put simply, 'there is not a single published study of solitary or supermax-like confinement . . . for longer than 10 days . . . that failed to result in negative psychological effects.'" (quoting Porter v. Clarke, 923 F.3d 348, 356 (4th Cir. 2019))); see also In re Medley, 134 U.S. 160, 168 (1890) (commenting regarding solitary confinement facilities nationwide that "[a] considerable number of the prisoners fell, after even a short confinement, into a semi- fatuous condition . . . and others became violently insane; others, still, committed suicide"). For this reason, solitary confinement, even if it takes a form that is less severe than the form it took in Wilkinson, cannot be equated with those conditions of confinement that Sandin considered to be so minor in their adverse impact on inmates as to always be, by their nature, either "ordinary incidents of prison life" or materially indistinguishable from them. See Sandin, 515 U.S. at 483

24

(criticizing cases based on Hewitt that found state-created liberty interests in "receiving a tray lunch rather than a sack lunch" or "receiving a paperback dictionary").

In fact, no circuit now holds that the bare legal possibility that a prison administrator may impose solitary confinement less severe than that at issue in Wilkinson on a member of the general prison population in and of itself makes any form of solitary confinement an "ordinary incident of prison life" in that prison system regardless of the nature and duration of that confinement. Rather, the circuits agree that even when the conditions of confinement at issue in a given case are no harsher than those that may obtain in "most solitary confinement facilities," Wilkinson, 545 U.S. at 224, and even when the regulations in a particular prison system permit officials to subject members of the general population to solitary confinement for administrative rather than only disciplinary reasons, the use of solitary confinement can still constitute an "atypical and significant hardship" rather than an "ordinary incident of prison life" in that prison system. See Aref, 833 F.3d at 254–55 ("[D]uration itself is widely regarded as a crucial element of the Sandin analysis." (citation omitted)).

For example, even those circuits that purport to look to "administrative segregation" as the baseline (rather than, say, the conditions of confinement that members of the general prison

25

population experience when not subject to such segregated confinement) have found in some circumstances that very lengthy periods of segregation imposed an atypical and significant hardship solely based on their length. See, e.g., Shoats v. Horn, 213 F.3d 140, 143-44 (3d Cir. 2000) (holding that eight years of administrative segregation imposed "atypical and significant hardship"); Harris v. Caruso, 465 F. App'x 481, 484 (6th Cir. 2012) (same). And, the Fifth Circuit, which is often cited for having held that administrative segregation can never implicate a liberty interest, has recognized that "extraordinary" durations of administrative segregation are indeed not "ordinary incidents of prison life," and that district courts must consider the combination of the length and conditions of confinement in determining whether a prisoner's administrative segregation implicates a liberty interest. Compare Pichardo v. Kinker, 73 F.3d 612, 612 (5th Cir. 1996) ("[A]dministrative segregation as such, being an incident to the ordinary life as a prisoner, will never be a ground for a constitutional claim . . . ."), with Wilkerson v. Goodwin, 774 F.3d 845, 853–55 (5th Cir. 2014) (holding that 39-year stay in administrative segregation imposed "an atypical and significant hardship"), and Carmouche v. Hooper, 77 F.4th 362, 366-67 (5th Cir. 2023) (holding that district court must "look[] at the length and conditions of confinement on a case- by-case basis to determine whether they give rise to a liberty

26

interest" and that there is no minimum period of time a prisoner must be held in administrative confinement before he can show atypicality, and reversing a district court's dismissal for frivolousness of pro se prisoner's claim for a violation of procedural due process arising from his being held in administrative confinement for over 300 days after serving a 30- day disciplinary sentence).

Moreover, our own precedent in Skinner -- though it does not purport to take a position on the baseline debate -- is to similar effect. It rejected the claim that the forty days of administrative segregation imposed on an inmate while he was being investigated for murdering another inmate was an "atypical and significant hardship." 430 F.3d at 486–87. But Skinner did not do so simply because the prison system regulations there at issue permitted administrative segregation of that duration to be imposed for that reason. Nor did Skinner suggest that forty days of administrative segregation was somehow too short in duration ever to qualify as an "atypical and significant hardship." Rather, Skinner held as it did only after assessing the nature and duration of the segregated confinement in relation to the reason for it, ultimately concluding that "six weeks is hardly an excessive time to conduct a preliminary inquiry into a murder." See id. at 487 (considering whether confinement at issue "was rational," whether "its duration was not excessive," and whether "the central

27

condition -- isolation from other prisoners -- was essential to its purpose"). Skinner therefore supports the notion that some forms of administrative segregation of a certain nature or duration may not be "ordinary incidents of prison life."

c.

The question therefore is this: Under what circumstances does a particular form of segregation -- given its nature, duration, or the combination of its nature and duration -- constitute an "ordinary incident of prison life" in a particular prison system rather than an "atypical and significant hardship" relative to those incidents? Here, too, Sandin is instructive. It indicates that our focus must be on distinguishing what is "ordinary" and "normally expected" from what is a "major disruption" or a "dramatic departure" from the "basic conditions" of the inmate's sentence. 515 U.S. at 485–87. In other words, Sandin shows that segregation will constitute an "ordinary incident of prison life" within a prison system if such confinement (accounting for its specific nature and duration) would be "normally expected" by such an inmate in the general prison population of that prison system. Sandin, 515 U.S. at 487; see also, e.g., Griffin v. Vaughn, 112 F.3d 703, 706 (3d Cir. 1997) ("[T]he baseline . . . is ascertained by what a sentenced inmate

28

may reasonably expect to encounter as a result of his or her conviction . . . ." (emphasis added)).10 The ordinary incidents of prison life that would be "normally expected," Sandin, 515 U.S. at 487, are those forms of confinement that would "likely" be faced, Hatch, 184 F.3d at 858. And, by "likely," we mean "not that the combination of events must be more probable than not, but that there must be a substantial chance of its occurrence." Id. (emphasis added); see also Kalwasinski v. Morse, 201 F.3d 103, 107 (2d Cir. 1999) ("[T]he district court must consider the periods of comparable deprivation typically endured by other prisoners in the ordinary course of prison administration, including general population prisoners and those in various forms of administrative and protective custody." (emphasis added) (internal quotations omitted))).

10The D.C. Circuit is "unique in considering the duration of confinement relative to similarly situated prisoners," Aref, 833 F.3d at 254, meaning those "serving similar sentences," Hatch v. District of Columbia, 184 F.3d 846, 856 (D.C. Cir. 1999) (articulating baseline as "the most restrictive confinement conditions that prison officials, exercising their administrative authority to ensure institutional safety and good order, routinely impose on inmates serving similar sentences"). Given that, as we hold below, Perry can supportably show that his confinement implicated a liberty interest regardless of whether the baseline is limited to similarly situated prisoners, we do not decide whether the baseline may be narrowed in this way. We also do not address circumstances in which a certain type of offender is categorically consigned to solitary confinement for reasons that are not specific to the individual prisoners.

29

3.

Based on this review of the precedent as to what constitute "the ordinary incidents of prison life," we now turn to the question of what the precedent shows with respect to when segregated confinement constitutes an "atypical and significant hardship" relative to those incidents. As is evident from our review of the precedent thus far, the determination of whether a specific form of segregation, at least when it amounts, as it often does, to "solitary confinement," Davis, 576 U.S. at 286 (Kennedy, J., concurring), constitutes an "atypical and significant hardship" usually will turn on specific showings regarding the use of such confinement in the prison system at issue, Sandin, 515 U.S. at 485. Nonetheless, some general observations are possible, even for cases involving forms of segregated confinement that are materially less severe than the form involved in Wilkinson, which, of course, was a form of segregated confinement that the Court there held constituted an "atypical and significant hardship." See Wilkinson, 545 U.S. at 223-24.

First, at least when the solitary confinement that grounds an inmate's procedural due process claim exceeds thirty days, that confinement will constitute an "atypical and significant hardship" per Skinner if, in relation to the reason for that confinement, it is "[ir]rational," "[in]essential," or "excessive" in duration. 430 F.3d at 487. This presumption is

30

warranted because solitary confinement for longer than thirty days imposes a meaningful hardship, and a member of the general prison population would not reasonably expect to be subjected to such unreasoned uses of it. This is not to say, though, that confinement beyond thirty days constitutes a per se due process violation.

Second, the length of segregated confinement that amounts to solitary confinement may make that confinement an "atypical and significant hardship," based on the length alone, when the inmate can show that "few" members of the general prison population have experienced similar durations of such confinement.11 See, e.g., Shoats, 213 F.3d at 144 (noting that eight-year stay in administrative segregation is "atypical" because "very few Pennsylvania prisoners have been confined in administrative custody for periods of eight years or more"); id. (citing prison officials' testimony that only "one percent of the inmate population at [the facility had] been confined in restricted housing for such lengthy periods of time"); see also Jones v. Baker, 155 F.3d 810, 815 (6th Cir. 1998) (Gilmin, J., concurring)

11 We do not suggest, of course, that there are no other circumstances in which such confinement could be shown to be an "atypical and significant hardship." Additionally, a showing that the prison system at issue rarely uses solitary confinement for a comparable length for administrative purposes suffices to show that confinement of such length in that system implicates a state- created "liberty" interest where the challenged confinement was of that kind.

31

(finding liberty interest implicated because segregation for two- and-a-half years "is clearly a rare occurrence" (emphasis added)); Lee v. Coughlin, 26 F. Supp. 2d 615, 635 (S.D.N.Y. 1998) (holding that 376-day stay in segregated confinement was atypical because it "was longer in duration than 99% of the general inmate population who are ever sentenced to [segregated confinement]"). Although in so concluding we do not identify a minimum length of confinement to which a plaintiff must have been subjected, we do agree that the showing can be made when the length of the confinement is as long as the lengths identified in the cases that we have cited for this proposition.

After all, the procedural due process claim in such cases seeks only to secure a procedural opportunity to challenge the basis for the continuation of the challenged confinement, not to bar the confinement's use for any length of time or purpose. And we agree with the cases cited above that solitary confinement that persists for the lengths at issue in them constitutes a "significant hardship."

Third, as a matter of procedure and burden allocation, the solitary confinement at issue also may be shown to constitute an "atypical and significant hardship" based on its prolonged nature even when the inmate makes no empirical showing as to the frequency with which the prison system at issue imposes solitary confinement of comparable length. See Colon v. Howard, 215 F.3d

32

227, 231 (2d Cir. 2000) ("Confinement . . . for 305 days is in our judgment a sufficient departure from the ordinary incidents of prison life to require procedural due process protections under Sandin. . . . As to atypicality, we are unaware of any data showing that New York frequently removes prisoners from general population for as long as [] 305 days. . . . If New York could have shown that [305 days was not atypical] . . . its opportunity to do so occurred at trial."). Indeed, even some circuits that look to administrative segregation as the baseline have proceeded on the understanding that long durations of segregated confinement are "atypical" based simply on the lengthy nature of the duration itself and thus without the inmate having made any additional showing regarding the frequency of the use of such confinement. See, e.g., Harris, 465 F. App'x at 484 (concluding without reference to empirical evidence that eight years of administrative segregation was of "atypical duration" and so implicated state- created liberty interest).

We do not identify a minimum length for grounding such a solely length-based showing that the segregated confinement constitutes an "atypical and significant hardship." But we do agree with the precedents cited above that the lengths that were involved in them were long enough supportably to show, based on length alone and without any empirical showing by the inmate that

33

such confinement is used only rarely, that the segregated confinement was an "atypical and significant hardship."

Of course, when the inmate's showing is based solely on the length of the confinement, defendants -- who have greater access to evidence concerning ordinary practices in a prison system -- are free to attempt to rebut that showing. And they may do so with evidence that the solitary confinement at issue -- prolonged though it was -- is indeed an "ordinary incident of prison life" in that state's prison system. But, in the absence of the defendants' conclusively making any such showing in rebuttal, the length of the confinement itself can be long enough to show supportably that the inmate subjected to it would not have reasonably expected to be subjected to solitary confinement for that length of time.

Fourth, and again as a matter of procedure and burden allocation, the state's own regulations, while not the source of any liberty interest themselves, can inform the inquiry into whether the solitary confinement at issue persisted long enough to ground such a solely length-based showing that the confinement constituted an "atypical and significant hardship." Indeed, Wilkinson explains that Sandin's overruling Hewitt signaled a "return to the due process principles . . . established in . . . Wolff and Meachum," and those precedents treated the language of state regulations, while not determinative, as relevant evidence

34

in determining whether the state had "create[d] liberty interests . . . protected by the Due Process Clause." Wilkinson, 545 U.S. at 222 (quoting Sandin, 515 U.S. at 483–84).

Such a regulations-informed inquiry also accords with how the SJC has applied Sandin. Specifically, as we noted up front, the SJC considered in LaChance whether the ten months of administrative segregation at issue in that case "exceeded the bounds of reasonable confinement in administrative segregation." 978 N.E.2d at 1206. In concluding that the ten months did exceed those bounds, the SJC explained that ten months far exceeded any of the ranges of time otherwise expressly contemplated by the then- operative DOC regulations for how long an inmate could be held in segregated confinement (including as discipline) before being afforded notice and an opportunity for rebuttal, as the longest such range spanned only ninety days. Id. at 1206–07, 1207 n.14. Thus, the SJC went on to explain, the ten months of such confinement constituted an "atypical and significant hardship" in that prison system because it lasted for more than ninety days. Id. at 1205, 1207.12

12 Following Sandin's approach of looking to other forms of segregated confinement if their conditions of confinement are "substantially similar" or "mirror[]" each other, Sandin, 515 U.S. at 476 n.2, 486, the SJC in LaChance also looked to the DOC regulations governing segregated confinement in a DSU, as the lower court in that case had found that the conditions of confinement in the DSU were essentially equivalent or even more severe than those

35

Using a state's own regulations to inform the durational inquiry as the SJC did in LaChance helps to focus the inquiry on the reasonable "expectations" that the state's own laws and policies have generated about what an inmate reasonably should understand to constitute the basic experience of prison life. A condition of confinement constitutes an "atypical and significant hardship" such that its imposition implicates a state-created liberty interest when that condition represents a "dramatic departure" from those state-generated expectations. And the state's own regulations may reasonably engender such expectations. See Wilkinson, 545 U.S. at 221, 223 (rooting state-created liberty interests in individual "expectation[s]" created by "state laws or policies"); Vitek, 445 U.S. at 489 (analyzing Meachum and related prison-confinement cases from the perspective of whether "state law" gave prisoners "justifiable" and "objective" expectations).

Such a regulations-informed inquiry also accords with Sandin's admonition "to afford appropriate deference and flexibility to state officials trying to manage a volatile environment." 515 U.S. at 482. It does so by grounding the determination about whether a particular duration of solitary

in the SMU. LaChance, 978 N.E.2d at 1205. We agree with the SJC that, under Sandin's logic, regulations governing "substantially similar" forms of segregated confinement can also bolster the conclusion that a particular duration of segregated confinement would be "atypical."

36

confinement is atypical due to its length alone in the state's own judgments about how long such "totally discretionary" uses of solitary confinement should be permitted to persist in related contexts. Id. at 486.

We recognize that -- as was the case in LaChance -- there may be no cap in the relevant state regulations on the length that solitary confinement may be imposed for the specific reason that the state has given for imposing solitary confinement in the case at hand. But we do not understand the absence of such a cap to demonstrate that solitary confinement when imposed for that reason is therefore an "ordinary incident of prison life" regardless of its length or nature. As we have explained, Sandin does not suggest that solitary confinement -- at least when it exceeds thirty days -- may be deemed an "ordinary incident of prison life" just because there is a bare legal possibility under the regulations that lengthy solitary confinement may be imposed for administrative reasons. What matters under the Sandin standard is whether the specific solitary confinement at issue -- given its duration and nature -- would constitute a "dramatic departure from the basic conditions" of prison life. Id. at 485.

Thus, we conclude, in accord with LaChance, that a plaintiff may rely on the state prison system's regulations supportably to show that the solitary confinement at issue is atypical due to its length alone, at least when the challenged

37

confinement is longer than thirty days. A plaintiff may do so by showing that such confinement exceeds the longest defined period of time that the state's own regulations specify as the time by which an inmate subject to comparable confinement must receive notice of the factual basis for it and an opportunity for rebuttal.13 When such a showing has been made, it is fair to proceed on the understanding (if not rebutted) that such solitary confinement is not an "ordinary incident of prison life" within the state's prison system. In addition, we agree with the SJC in LaChance that state regulations need not speak directly to the specific type of confinement or reason for confinement at issue to show it to be an atypical and significant hardship; indeed, LaChance, in considering a prisoner's confinement in an SMU while on awaiting action status -- like the segregation faced by Perry in this case -- drew on DOC regulations governing segregation in a DSU pending a disciplinary investigation or proceeding. See 978 N.E.2d at 1207 n.14; cf. Haverty v. Comm'r of Corr., 776 N.E.2d

13 We do not understand Sandin to foreclose the possibility that solitary confinement for less than thirty days could in some situations constitute "an atypical and significant hardship," whether because the conditions of the confinement at issue are particularly severe, the "ordinary incidents of prison life" in the prison system at issue are notably less severe than those that the Court found characterized Hawaii's prison system at the time of Sandin, or the use of the confinement is unreasoned. Given that the duration of Perry's confinement significantly exceeded thirty days, we need not determine in what circumstances that might be true.

38

973, 991 (Mass. 2002) ("[T]he procedural protections contained in [the DSU-specific regulations] must be afforded to all prisoners before they are housed in DSU-like conditions.").

None of this is to say that such state regulations necessarily create a liberty interest (regardless of what showing the defendants make) or that no such interest exists in the absence of such regulations. Sandin, as we have noted, rejected a purely regulations-based focus in overturning Hewitt. In that regard, we agree with the Second Circuit that, wholly apart from what a state's regulations provide, a "more fully developed record" might show that "even relatively brief confinements" of more than thirty days "under normal [segregation] conditions [are], in fact, atypical" in the prison system at issue.14 See Palmer v. Richards, 364 F.3d 60, 65-66 (2d Cir. 2004). Indeed, we note that the Second Circuit has observed that "[i]n the absence of a detailed factual record," it has affirmed dismissal of due process claims like

Perry does not allege conditions of confinement more severe 14

than those outlined in the regulations governing confinement in an SMU. However, we note that especially severe conditions of confinement can also render an otherwise typical duration of segregated confinement an "atypical and significant hardship." See Aref, 833 F.3d at 254 ("[E]specially harsh conditions endured for a brief interval and somewhat harsh conditions endured for a prolonged interval might both be atypical." (quoting Sealey, 197 F.3d at 586)); cf. Hutto v. Finney, 437 U.S. 678 (1978) ("[T]he length of confinement cannot be ignored in deciding whether the confinement meets constitutional standards. A filthy, overcrowded cell and a diet of 'grue' might be tolerable for a few days and intolerably cruel for weeks or months.").

39

Perry's "only in cases where the period of time spent in [segregated confinement] was exceedingly short -- less than the 30 days that the Sandin plaintiff spent in [segregated confinement] -- and there was no indication that the plaintiff endured unusual . . . conditions." Id. And, precisely because of the various ways in which discovery might yield a record showing that more than thirty days of confinement "under normal [segregation] conditions [are], in fact, atypical," id., we can see why that would be so.

4.

Against this legal backdrop, we now turn back to Perry's contention that his segregated confinement implicates a state- created liberty interest. He makes various arguments in this regard.

Perry first argues that the conditions of his confinement "were essentially on all fours with those in Wilkinson, and where they deviated, they were more severe." On that basis, he contends that Wilkinson compels the conclusion that his segregated confinement was "atypical and significant under any plausible baseline." Wilkinson, 545 U.S. at 223.

The defendants respond that the solitary confinement in Wilkinson differed from Perry's. They point to the fact that, among other things, that confinement was "indefinite," while

40

Perry's was dependent on a named contingent event (his transfer out of state).

We need not resolve this dispute at this first step of the qualified immunity inquiry. And that is because Perry separately contends, and we agree, that the record supportably shows that the solitary confinement to which he was subjected constituted an "atypical and significant hardship" due to the length of its duration and the absence of any showing by the defendants as to how typical such a lengthy use of solitary confinement was in the DOC.15 As to this aspect of Perry's contention, the record supportably shows that, as both the District Court and the panel found, Perry's administrative segregation is properly understood to constitute solitary confinement. Perry, 2016 WL 5746346, at *1 n.3; Perry, 751 F. App'x at 8. The record also shows that the form of the segregated confinement was at least as severe as the

15As to Perry's separate argument that the basis for his initial and continued placement in the SMU was pretextual, we agree with the District Court that Perry "has presented no facts to support this assertion," and that there are therefore "no facts or reasonable inferences sufficient to raise a genuine issue on the propriety of placing and keeping Perry in the SMUs." Perry, 2016 WL 5746346, at *13 & n.31. We also note that, for reasons we will explain in our analysis of step two of the qualified immunity inquiry, there is no merit to Perry's contention -- raised for the first time in his en banc briefing -- that the state's conduct violated clearly established law based on our pre-Sandin ruling in Stokes v. Fair, 795 F.2d 235 (1st Cir. 1986). See infra at 25.

41

segregated confinement at issue in Sandin;16 indeed, the record shows that the segregated confinement that Perry endured is comparable to conditions of confinement that the Supreme Court has otherwise described as constituting "solitary confinement." See Davis, 576 U.S. at 286 (Kennedy, J., concurring) (describing inmate's "'administrative segregation' or, as it is better known, solitary confinement," as his "be[ing] held . . . in a windowless cell no larger than a typical parking spot for 23 hours a day; and in the one hour when he leaves it, he likely is allowed little or no opportunity for conversation or interaction with anyone"); see also Wilkinson, 545 U.S. at 224.

In addition to the nature of the confinement being solitary confinement, the record shows, with respect to the solitary confinement's length, that the length was "prolonged." Perry, 2016 WL 5746346, at *15. Indeed, the segregated confinement far exceeded the thirty days involved in Sandin.

16 The confinement at issue in Sandin involved similar periods of isolation and contained similar restrictions on exercise and personal visits. See Petitioner's Brief, Sandin, 515 U.S. 472, 1994 WL 646163, at *7-*8; compare id. at *48A–*71A, with 103 MASS. CODE REGS. § 423.09 (2007). The record in Sandin also did not reflect any use of "solid door status," which Perry alleges he endured during at least part of his stay in the SMUs. See Petitioner's Brief, Sandin, 515 U.S. 472, 1994 WL 646163, at *48A– *71A. Moreover, the Sandin court pointed to the twelve to sixteen hours a day of "lockdown time" endured by members of the general prison population as comparable to the confinement at issue there, Sandin, 515 U.S. at 486 & n.8, which is significantly less than the amount of time that Perry was required to spend in his cell every day.

42

The defendants do contend in response to Perry's focus on the length that the record conclusively shows that Perry's confinement was necessary to its purpose even given its substantial length. The defendants emphasize in this regard that the segregated confinement was initially imposed, and then persisted, due to safety concerns relating to Perry's alleged gang affiliation.

But, even if a reasonable juror could not find the duration of the solitary confinement "irrational," "inessential," or "excessive" in relation to its purpose, Skinner, 430 F.3d at 487, there still is force to Perry's contention that the duration of the solitary confinement was nonetheless "excessive" in the sense that it imposed on him an "atypical and significant hardship" due to its length. True, Perry has not introduced empirical evidence about how frequently (if at all) inmates under DOC custody are held in administrative segregation for as long as he was.17 But, his fifteen months of solitary confinement -- even prior to his initial out-of-state transfer -- is of such a length that it suffices supportably to show that it imposed on him an "atypical

17After the conclusion of supplemental en banc briefing and oral arguments, Perry submitted a letter to this Court pursuant to Federal Rule of Appellate Procedure 28(j) contending that since 2015 -- which was after the SJC decided LaChance -- less than three percent of Massachusetts state prisoners endure fifteen or more consecutive days in solitary confinement, and that such confinement is therefore "rar[e]." Because this data was not presented below, we do not consider it here.

43

and significant hardship" in the absence of the defendants making any contrary showing regarding the frequency of use of such prolonged solitary confinement for administrative reasons.

This conclusion accords with those conclusions that have been reached by courts that have applied a baseline like the one that we have now adopted. See, e.g., Colon, 215 F.3d at 230-31 (holding that approximately ten months of administrative segregation constituted an "atypical and significant hardship" based on the extent of the duration itself). And, we note, nothing in the record indicates that a member of the DOC's general prison population stands a "substantial chance" of being subjected to administrative segregation for that long. See id. (noting that nothing in the record rebutted the conclusion that approximately ten months of administrative segregation was "a sufficient departure from the ordinary incidents of prison life to require procedural due process protections under Sandin").

This conclusion draws independent support from the DOC regulations that governed administrative segregation in the SMU at the time of Perry's confinement. Those regulations did not provide that the inmates in DOC who were on "awaiting action status" had to receive notice of the factual basis for their confinement and an opportunity for rebuttal. Nor did those regulations place any cap on how long such confinement could persist for inmates being held while on that status. See 103 MASS. CODE REGS. § 423.08(2)

44

(2007). Those regulations, however, did entitle inmates confined in the SMU for disciplinary investigation to be either formally charged with a disciplinary offense -- thus entitling them to notice and a hearing, id. § 430.11 -- or released from the SMU within ninety days, id. § 430.21(2). Moreover, those regulations provided that even inmates subjected to disciplinary detention in an SMU upon a formal finding of wrongdoing could not be so detained for more than fifteen days for one offense or thirty days for multiple offenses "unless specifically authorized by the commissioner" after a disciplinary proceeding. Id. § 423.06. And, finally, the DOC regulations then in place provided that inmates who were on awaiting action status while awaiting a final decision regarding their potential placement in a DSU had to be afforded a hearing within fifteen days, or within thirty days if they were under investigation for a possible disciplinary offense. Id. § 421.08(3). We thus conclude that, given the regulations then in place, a duration of solitary confinement of more than ninety days is certainly long enough to make such confinement an "atypical and significant hardship" within the prison system at issue in Perry's case, at least in the absence of a showing by the defendants regarding the frequency with which inmates in that system are subjected to such lengthy confinement for administrative reasons.

We note, too, that insofar as the defendants "could have shown" that it was not "atypical" for members of the general prison

45

population to be held in administrative segregation of the sort that Perry endured for a comparable period, they have not identified evidence in the record that conclusively does so. See Colon, 215 F.3d at 231. Therefore, their "opportunity to do so" would "occur[] at trial." Id.

B.

Having established that Perry has supportably shown that his confinement implicated a "liberty" interest, we next must determine -- still at the first step of the qualified immunity inquiry -- whether a reasonable juror also could find that Perry was denied the process that he was due in consequence under the Due Process Clause.18 As we will explain, we conclude that a reasonable juror could.

The defendants contend that Perry has waived this objection 18

because his pro se brief to the panel argued only that "he should have been afforded process outlined in" certain state regulations that did not apply to his particular placement, and that he did not otherwise argue what process was "mandated by the Constitution." But Perry did in his opening brief cite both Mathews and Wilkinson to argue that he received "no notice or opportunity to be heard," and quoted Wilkinson to contend that "the defendants provided [him] with no procedural safeguard[s]" or "a basis for objection before the next decisionmaker or in a subsequent classification review," such that the process he received "did not satisfy" the Mathews test. These arguments are more than sufficient to preserve the claim more fully developed in his counseled supplemental briefing. See Erickson v. Pardus, 551 U.S. 89, 94 (2007).

46

1.

Based on the three-factor framework of Mathews v.

Eldridge, 424 U.S. 319, 335 (1976), Wilkinson held that inmates subjected to solitary confinement in a manner that implicates a liberty interest must be afforded the "informal, nonadversary procedures" set forth in both Hewitt and Greenholtz v. Inmates of Neb. Penal & Corr. Complex, 442 U.S. 1 (1979). See Wilkinson, 545 U.S. at 229. Those procedures require at a minimum "[n]otice of the factual basis" for the confinement, id. at 225-26, and "an opportunity to present [the inmate's] views" to the official charged with the decision to confine the inmate, Hewitt, 459 U.S. at 476.19 With respect to the latter, the Court explained in Greenholtz that permitting the inmate to "appear before the Board and present letters and statements on his own behalf" gave the inmate an "effective opportunity" both to ensure that the Board's records were "in fact the records relating to his case," and to "present any special considerations" that could inform the decision. 442 U.S. at 15. And Wilkinson described the required process in this respect as one that affords the inmate a "fair opportunity for rebuttal." 545 U.S. at 226.

19Sufficient process must also provide inmates with a "short statement of reasons" for the ultimate decision following a hearing, Wilkinson, 545 U.S. at 226, and meaningful periodic reviews of ongoing confinement, Hewitt, 459 U.S. at 477 n.9.

47

2.

Perry contends that he can supportably show on this record that he did not receive the required "opportunity to be heard." If Perry is right on that score, then he can show that he was denied the process that was due, even if we were to assume that he cannot supportably show that he was deprived of the required notice of the factual basis for his confinement. We thus begin and end our analysis at the first step of the qualified immunity inquiry into the process component of Perry's procedural due process claim with his contention regarding the "opportunity for rebuttal" afforded to him. For, as we will explain, we conclude that Perry can supportably show on this record that he was denied that opportunity.

In arguing otherwise, the defendants first point to the "SMU Reviews," which were conducted approximately three times per week and which the defendants claim "reflect[ed] periodic updates" about Perry's status "as relevant changes or issues arose." But Perry asserts in a sworn affidavit that he "received no notice that [the SMU Reviews] occurred prior to their occurrence, nor any opportunity to be heard" or to "participate in writing or in person."

The defendants' sole response is that "[i]nmates were free to provide staff with any information during their time in the SMU." Perry, 2016 WL 5746346, at *7 n.19. The relevant

48

regulations, however, did not require officials to consider such information as part of their SMU Reviews. See 103 MASS. CODE REGS. § 423.08 (2007). Furthermore, the SMU Review entries that appear in the record do not acknowledge any statements made by Perry. Thus, a juror could reasonably find that, while SMU Reviews might constitute the "periodic review" that due process separately requires, Hewitt, 459 U.S. at 477 n.9, they did not give Perry an "opportunity to present his views," id. at 476, and the defendants do not point to anything in the record that suggests that Perry was in fact given an opportunity to present his views during those SMU Reviews.

The defendants separately argue that a grievance procedure was available to Perry and that it provided him with the required opportunity for rebuttal. But the state has not conclusively shown that the grievance procedure was available for this issue or that the grievance procedure provided an opportunity for such rebuttal. See Rodi v. Ventetuolo, 941 F.2d 22, 29 (1st Cir. 1991); see also Black v. Parke, 4 F.3d 442, 449 n.5 (6th Cir. 1993). And, we note, Sandin itself pointed to grievance procedures as a form of "other protection" prisoners "retain" when they are not protected by the Due Process Clause. See 515 U.S. at 487 n.11. Finally, the record supportably shows that DOC officials either declined to entertain or summarily denied every grievance in the

49

record in which Perry attempted to contest his confinement, often by stating that his concerns were "non-grievable."20 We come, then, to the defendants' contention that the classification hearing that Perry received on December 17, 2010 -- one week after his initial placement in the SMU -- indisputably afforded him the required "fair opportunity for rebuttal."21 That hearing was not governed at the relevant time by the SMU regulations; instead, the hearing was governed by separate regulations applicable to all DOC inmates. See 103 MASS. CODE REGS. §§ 420.06–420.09 (2007). That latter set of regulations required that inmates "undergo an initial classification process" upon commitment to the DOC, then receive another reclassification "at least annually." Perry, 2016 WL 5746346, at *6.

The record shows that Perry submitted at least eight 20

grievances challenging his segregated confinement in the SMU. Four of those were dismissed as "non-grievable." One was denied as duplicative. Three were denied with summary citations to the regulations governing awaiting action status, with a note that Perry was "pending a transfer," but without engaging with his substantive challenges to his designation under those regulations.

Given our conclusion that Perry has supportably shown that 21

his segregated confinement implicated a liberty interest after ninety days, we do not consider Perry's second and third classification hearings (on December 16, 2011, and October 31, 2012, respectively). Those hearings occurred over one year after Perry's placement in the SMU, and the defendants do not argue that, even if the first classification hearing did not provide Perry with a sufficient opportunity to be heard, the second and/or third hearings did, and that Perry therefore received sufficient process.

50

The classification process determined an inmate's "security level" and, consequently, the DOC facility in which the inmate should be placed. 103 MASS. CODE REGS. § 420.07 (2007). The record supportably shows, however, that the process did not determine whether an inmate would be placed in segregated confinement -- including whether such confinement was to be in an SMU -- once assigned to a facility within the system. Indeed, the regulations reveal that it was officials at the facility in which an inmate was placed pursuant to the classification process and that contained SMUs -- such as facilities like SBCC and MCI-Cedar Junction in which Perry was housed in SMUs -- who separately determined, as issues arose, whether to isolate that inmate in an SMU. Compare 103 MASS. CODE REGS. § 423.00 et seq. (2007), with 103 MASS. CODE REGS. § 420.00 et seq. (2007).

The record does clearly show that Perry's first classification hearing occurred only one week after his initial placement in the SMU. But nothing in the record indicates that, beyond this coincidence of timing,22 the classification hearing and

Perry had been transferred from a Department Disciplinary 22

Unit ("DDU") at MCI-Cedar Junction to the general population at SBCC on November 18, 2010, and DOC regulations required that inmates receive a reclassification hearing "within thirty business days of release from the DDU." 103 MASS. CODE REGS. § 420.09 (2007). Perry then spent three weeks in the general population at SBCC before being transferred to the SMU on December 10, at which time he was already "also awaiting a classification hearing," Perry, 2016 WL 574346, at *9, making it possible that this hearing had

51

the SMU placement were related in that the former could be used to contest the latter.

To the contrary, the "Classification Report" from Perry's hearing stated, "Inmate Perry is requesting placement at MCI Concord and this writer concur[s] with inmate request," and the Classification Board recommended only that Perry be transferred to the general population at MCI Concord, a medium- security facility. The record then shows that it was not until two months later that Lori Cresey, the "Commissioner's designee" responsible for making the final determination, see 103 MASS. CODE REGS. § 420.08(i)–(j) (2007), concluded that Perry's placement in any Massachusetts facility would lead to gang-related tensions, and so made "a final decision to screen Perry for an out-of-state placement due to security concerns," Perry, 2016 WL 5746346, at *9- *10. And yet, this final decision did not state that Perry would have to remain in the SMU until he could be transferred. Rather, the record supportably shows that this determination was communicated to Perry only ten days later, when the Superintendent of SBCC at the time wrote to Perry that he had decided "to hold [Perry] in the [SMU] for the safety and security of the facility as well as [Perry]," and that Perry "w[ould] not be allowed to enter General Population at [SBCC]" while he awaited out-of-state

been scheduled, and that Perry had received notice of the hearing, even before his placement in the SMU.

52

transfer. See also Supplemental Brief of Defendants-Appellees at 6 ("Ultimately, the Classification Division determined . . . that Perry required an out-of-state placement. SBCC staff deemed it necessary that Perry remain in the SMU to ensure the safety and security of the institution." (emphases added)).

In other words, the record does not indisputably show that the pre-scheduled classification hearing afforded the required opportunity for rebuttal. Rather, a reasonable juror could find on this record that this hearing was dedicated to determining which facility Perry should be housed in, not whether he should be housed in an SMU within such a facility, and so did not afford him the requisite rebuttal opportunity.

C.

Thus, despite the various forms of process that the record indisputably shows that the defendants offered Perry, the record supportably shows that he was not given the opportunity to contest the factual basis for his confinement in the SMU. Moreover, because of our previous conclusion that Perry has supportably shown that his solitary confinement implicated a state-created liberty interest by virtue of its extended duration (at least given the absence of any showing by the defendants to the contrary), Perry has also supportably shown that he was entitled to such an opportunity to contest the factual basis for his confinement as a matter of procedural due process.

53

Accordingly, a reasonable juror could find that Perry was denied the procedural due process right that he claims.23 IV.

Having specified the contours of the right at issue and the specific grounds on which Perry rests his claim that the record supportably shows that he was deprived of a state-created liberty interest without due process, we are now well-positioned to turn to step two of the qualified immunity inquiry. Here, our focus is on whether the defendants are entitled to summary judgment based on qualified immunity because the law under which a reasonable juror could supportably find that Perry was denied the right to procedural due process was not clearly established at the time of the denial. Savard v. Rhode Island, 338 F.3d 23, 27 (1st Cir. 2003).

The defendants contend, as the panel ruled, that the law as to both the liberty interest and process issues was not clearly established as of the time of the alleged violation of the constitutional right that Perry claims. Perry, 751 F. App'x at

23Of course, jurisdictions in our Circuit are free to limit solitary confinement beyond what constitutes an atypical hardship under this opinion and jurisdictions are free, also, to provide more process than today's holding would require. However, our decision today establishes, clearly, the baseline for analyzing when prisoners' state-created liberty interests are curtailed by solitary confinement and the procedural protections that must follow from such curtailments.

54

11–12. But we need not decide whether, if officials knew that Perry's confinement implicated a liberty interest, they could have reasonably believed that they provided him with constitutionally sufficient process. And that is because we conclude, reviewing de novo, Savard, 338 F.3d at 27, that the law was not clearly established that Perry's solitary confinement implicated a liberty interest.

A.

"'Clearly established' means that, at the time of the officer's conduct, the law was 'sufficiently clear' that every 'reasonable official would understand that what he is doing' is unlawful." District of Columbia v. Wesby, 138 S. Ct. 577, 589 (2018) (quoting Ashcroft v. al-Kidd, 563 U.S. 731, 741 (2011)). "This demanding standard protects 'all but the plainly incompetent or those who knowingly violate the law.'" Id. (quoting Malley v. Briggs, 475 U.S. 335, 341 (1986)).

To be clearly established, the law "must define 'the right allegedly violated . . . in a particularized sense so that the contours of the right are clear to a reasonable official.'" Eves v. LePage, 927 F.3d 575, 583 (1st Cir. 2019) (en banc) (quoting Reichle v. Howards, 566 U.S. 658, 665 (2012)). "[G]eneral statements of the law" may give "fair and clear warning, and in other instances a general constitutional rule already identified in the decisional law may apply with obvious clarity to the

55

specific conduct in question." United States v. Lanier, 520 U.S. 259, 271 (1997); see also Hope v. Pelzer, 536 U.S. 730, 741 (2002). Furthermore, "[a]mbiguity in the law cannot be manufactured by borrowing from factually and legally distinguishable cases." Marrero-Méndez v. Calixto-Rodríguez, 830 F.3d 38, 48 (1st Cir. 2016).

"A rule is clearly established either when it is 'dictated by controlling authority or a robust consensus of cases of persuasive authority.'" Irish v. Fowler, 979 F.3d 65, 76 (1st Cir. 2020) (quoting Wesby, 138 S. Ct. at 589-90), cert. denied, 142 S. Ct. 74 (2021). "A 'robust consensus' does not require the express agreement of every circuit," and can be established by "sister circuit law" that "would provide notice to every reasonable officer that his conduct was unlawful." Id.

B.

Perry certainly cannot show that it was clearly established at the time of his confinement that a reasonable juror could conclude from the nature of the DOC's regulations that solitary confinement of greater than ninety days is excessive in duration and so for that reason alone imposes "an atypical and significant hardship" on a DOC inmate. Indeed, other courts as of that time had held to the contrary. Griffin, 112 F.3d at 708–09 (finding, based on Pennsylvania prison regulations, that "one can conclude with confidence that stays of many months [in

56

administrative custody] are not uncommon," despite fact that regulations also "confer[red] upon an inmate in [the plaintiff's] position a right to be released from administrative custody to general population after 20 days in the absence of a misconduct charge"). And, we note, the SJC in LaChance, looking back, did not itself understand the legal landscape at that time to be clear in establishing such a principle. See 978 N.E.2d at 1207–08 (granting qualified immunity to defendants regarding 10-month confinement due to this conflicting case law).

Nor was it clearly established at the relevant time that the length of the solitary confinement that Perry served -- given the conditions that he endured during it -- made it, even if only when not rebutted, an "atypical and significant hardship." As we have seen, length alone can suffice to establish that solitary confinement is a hardship of that sort. Indeed, today no circuit holds otherwise. But we had not made clear prior to 2010 that administrative segregation -- even involving conditions like Perry's -- was an atypical and significant hardship if it persisted for as long as his did. See Skinner, 430 F.3d at 486-87; Judgment, Dávila v. Maloney, No. 05-2520 (1st Cir. Oct. 2, 2007). Indeed, we had explained, prior to that time, that the baseline from which to measure what is such a hardship was the subject of dispute throughout federal circuits in the wake of Sandin. And we had done so in explaining our reasons for rejecting claims that

57

prolonged periods of such confinement were hardships of that kind. See Skinner, 430 F.3d at 486-87; Judgment, Dávila v. Maloney, No. 05-2520 (1st Cir. Oct. 2, 2007). True, Perry's confinement was quite lengthy. But, even still, the state of the case law within our own circuit as of the time of Perry's solitary confinement was such that it was not clearly established from our own precedents that his confinement was an "atypical and significant hardship."

Moreover, as we have seen, those circuits that as of 2010 treated administrative segregation as the baseline for measuring whether prolonged administrative segregation constituted an atypical and significant hardship had, as of that time, held that comparable and even longer periods of solitary confinement did not implicate a liberty interest. See, e.g., Griffin, 112 F.3d at 706–08 (3d Cir. 1997) (no liberty interest for fifteen months because conditions matched those of administrative confinement in that prison system); Jones, 155 F.3d at 812–13 (6th Cir. 1998) (no liberty interest for thirty months); DiMarco, 473 F.3d at 1344 (10th Cir. 2007) (no liberty interest for fourteen months). And, as of that time, one circuit had appeared to hold that administrative segregation is essentially incapable of creating a liberty interest, see Pichardo, 73 F.3d at 612 (5th Cir. 1996) ("[A]dministrative segregation as such, being an incident to the ordinary life as a prisoner, will never be a ground for a constitutional claim . . . ."), which Skinner itself pointed

58

out, 430 F.3d at 486–87. So, there was no robust consensus of out-of-circuit precedent then in place that supported his claim to have endured an "atypical and significant hardship" insofar as that claim was based on the confinement's length alone.24 Thus, like the District Court, see Perry, 2016 WL 5746346, at *14, and the panel, Perry, 751 F. App'x at 9–11, we agree that, Wilkinson aside, it was not clearly established at the time of the alleged violation of Perry's right to procedural due process that he had been deprived of a liberty interest. He thus needs to show that there is some other basis for concluding that the confinement that he endured implicated a liberty interest in a way that would have been clearly established at the relevant time.25

24 In light of Wilkinson, we do not rely on the Seventh Circuit's baseline of "the most restrictive prison" in a state's system, Wagner, 128 F.3d at 1176, in concluding that the violation of Perry's rights was not clearly established. Indeed, although the Seventh Circuit has, at times, post-Wilkinson continued to invoke the "most restrictive prison" baseline, see, e.g., Marion v. Columbia Corr. Inst., 559 F.3d 693, 699 (7th Cir. 2009), it has also, at least once, vacated a district court's ruling in the immediate wake of Wilkinson on the ground that the district court analyzed the baseline under Wagner "without the benefit of the Supreme Court's decision in Wilkinson," Westefer v. Snyder, 422 F.3d 570, 586 (7th Cir. 2005).

25We also reject Perry's arguments, raised for the first time on appeal in his en banc briefing, that our pre-Sandin circuit precedent in Stokes or his mental illness "further tip[] the scale toward a liberty interest" being clearly established. Stokes's broadly worded holding that Massachusetts DOC "regulations[] created a liberty interest entitling [a prisoner] to procedural due process in the initiation and continuance in awaiting action

59

C.

To make that showing, Perry argues that the conditions of his confinement were "on all fours with those in Wilkinson, and where they deviated, they were more severe," such that his confinement was also "atypical" under "any plausible baseline." Perry is right in principle that if he can supportably show that his confinement was clearly comparable to the confinement at issue in Wilkinson, then he can supportably show that he was deprived of

status detention," 795 F.2d at 238, has not been cited since Sandin was issued and depended entirely on the specific language of the then-applicable regulations, see id. at 237-38. The fact that the regulations in effect when Perry was in solitary confinement, as he puts it, "generally track the regulations at issue in Stokes" cannot clearly establish that he had a protected liberty interest given that Stokes does not describe the nature or length of the confinement at issue in that case and given that Sandin "abrogated" the exact methodology on which Stokes was decided. Wilkinson, 545 U.S. at 222; see also LaChance, 978 N.E.2d at 1207–08 (concluding that DOC regulations did not clearly establish that a 10-month period of administrative confinement implicated state-created liberty interest).

Perry does also contend he can show what he must, based on the fact that he was in the challenged confinement while suffering from mental illness. But, beyond the fact that Perry may have waived this argument, as the defendants argue, we agree with the defendants that the cases Perry cites are not sufficient to clearly establish, at least as of the time of his confinement, "if, when, or how mental illness would be pertinent to the liberty interest analysis." See Hamner v. Burls, 937 F.3d 1171, 1178–80 (8th Cir. 2019) (granting qualified immunity because "[w]hile it is possible . . . that a combination of circumstances involving solitary confinement could curtail a liberty interest, it is not beyond debate that the defendant officials did so by segregating a prisoner with [borderline personality disorder, post-traumatic stress disorder, antisocial personality disorder, anxiety, and depression] for 203 days under the conditions alleged" (citations omitted)).

60

a liberty interest under clearly established law. After all, his prolonged solitary confinement post-dated Wilkinson.

We did not address at step one of the qualified immunity inquiry whether Perry can supportably show on this record that his confinement is "on all fours with Wilkinson." But we must address that contention here, given that it is the sole remaining basis on which he may show that he was deprived of a liberty interest under clearly established law. As we will explain, though, Perry's contention falters, notwithstanding that we have no reason to doubt that the conditions in the general prison population in Ohio during the time of the confinement at issue in Wilkinson are comparable to those of the general prison population in the DOC during the time of Perry's confinement.

In Wilkinson, the Supreme Court identified five conditions of confinement that supported its conclusion that confinement in the Ohio "supermax" facility imposed an "atypical and significant hardship under any plausible baseline." 545 U.S. at 223. The Court explained that the first three conditions --(1) "almost all human contact [was] prohibited"; (2) "the light, though it [could] be dimmed," was on for twenty-four hours; and (3) inmates were permitted to exercise for only "1 hour per day" and "only in a small indoor room," id. at 223–24 -- "likely would apply to most solitary confinement facilities," id. at 224. But the Court then identified "two added components" that supported its

61

conclusion: fourth, that the "duration" of confinement was "indefinite"; and fifth, that the "placement disqualifie[d] an otherwise eligible inmate for parole." Id.

The Court explained that the confinement at issue in Wilkinson was "indefinite" because it was "limited only by an inmate's sentence" and because it was "reviewed just annually." Id. at 214–15, 224 (emphasis added). Perry's segregated confinement, by contrast, was contingent on an identifiable event -- his transfer to an out-of-state facility -- that in its nature would precede the end of his sentence. Thus, even assuming that Perry's confinement is otherwise materially indistinguishable from that in Wilkinson,26 the question is whether it was clearly

26The record does not supportably show that the light in Perry's cell was left on throughout his entire confinement, even if there is some evidence in the record showing that it may have been left on for at least some period of it. Some courts, however, have commented that the lack of this factor does not necessarily "prove[] that conditions of confinement at the [Ohio supermax facility in Wilkinson] are significantly more restrictive than the conditions of confinement" at issue in a given case. See Westefer v. Snyder, 725 F. Supp. 2d 735, 760 (S.D. Ill. 2010), vacated on other grounds by Westefer v. Neal, 682 F.3d 679 (7th Cir. 2012).

The record also does not supportably show that Perry's confinement rendered him ineligible for parole, given that he was already ineligible. Perry contends, however, that this does not weaken the similarity between his case and Wilkinson because in Wilkinson, too, several of the plaintiffs were ineligible for parole, but the Court still found that all the plaintiffs had a liberty interest in avoiding confinement. See Am. Compl. at ¶ 62(d), Austin v. Wilkinson, No. 4:01-CV-071, 2001 WL 34903823 ("Plaintiffs Benge and Robb are among more than a dozen death row inmates who are housed at OSP in violation of AR 5120-9-12."); Austin v. Wilkinson, 372 F.3d 346, 348 n.1 (6th Cir. 2004) (naming

62

established as of the time of that confinement that it was "indefinite" within the meaning of Wilkinson, notwithstanding that an endpoint -- contingent though that endpoint was -- had been identified for it. We agree with the defendants that it was not clearly established that such confinement was indefinite.

There is no doubt that, in this case, Perry's confinement was at least "prolonged." Perry, 2016 WL 5746346, at *15. And Perry maintains that he did not know "when, or even whether, his solitary confinement would end," given that his confinement's termination depended on a contingency not within his control: the availability of an out-of-state placement. Moreover, the uncertainty as to duration persisted for over a year until his out-of-state transfer occurred in March 2012. We also recognize that the exercise of a prison administrator's wholly discretionary choice to cut short solitary confinement prior to the sentence's end cannot count as a contingency that makes such confinement definite rather than indefinite: that same contingency existed in Wilkinson itself. See Wilkinson, 545 U.S. at 217.

Here, however, there is no supportable basis for finding that the end of Perry's solitary confinement was dependent only on the prison administrators' discretionary choice to bring that

these death row plaintiffs).

We need not resolve either of these issues, however, because we hold that it was nonetheless unclear whether Perry's confinement was "indefinite" within the meaning of Wilkinson.

63

confinement to an end. There is nothing in the record to indicate that Perry's designation as an inmate awaiting an out-of-state transfer was clearly a pretext for holding him for however long prison administrators wished. Perry, 2016 WL 5746346, at *13. Moreover, far from containing evidence that indicates that prison officials did not take action to pursue the transfer, the record shows without dispute that one defendant, Thomas Neville, reached out to at least five states regarding Perry's potential transfer, beginning with a request to Virginia on April 20, 2011 (two months after the initial February 4, 2011 decision to screen Perry for out-of-state transfer), and requests to New Jersey and Idaho on July 12, 2011.

The record also contains undisputed evidence purporting to explain the timeframes both for sending initial requests to other states as well as receiving responses from those states. For example, Neville explained in a sworn affidavit that the out- of-state transfer process in Perry's case "included trying to determine which states might be more appropriate than others, considering various factors, including whether the other states have issues with the same [gangs] and/or whether other [DOC] inmates with [gang] ties have already been transferred to other states." His affidavit further attests that he "made phone calls to other states inquiring as to referrals and to ascertain which inmates those states would refer to the [DOC] in return -- because

64

it is typically a one-to-one swap, meaning [Neville] would then have to review information on any inmate the other state wishes to send to Massachusetts." Finally, Neville's affidavit states that referring an inmate to another state "includes providing a lot of information," and that after making the referrals for Perry he "periodically made phone calls to officials in the other states to inquire as to the status of their review."

We recognize that Neville's affidavit does indicate that the length of Perry's solitary confinement depended in part on the willingness of other states to facilitate his transfer. But the record also shows without dispute that in January 2012, the Virginia Department of Corrections indicated to Neville that it was considering accepting Perry and requested more information, and that Perry, after learning of this, sent Neville a letter in which he threatened to kill himself if transferred outside New England. The record also shows, again without dispute, that from then on DOC officials only pursued transfer within New England, which the defendants argue without dispute likely delayed Perry's transfer and further prolonged his confinement.

That Perry, upon learning that his confinement in an SMU might soon be coming to an end, took action that may have further prolonged his confinement supports the conclusion that a reasonable officer could have understood that Perry's segregated confinement was not indefinite within the meaning of Wilkinson.

65

See Eves, 927 F.3d at 584. As we have explained, the confinement there by design could last as long as prison administrators chose to permit it to extend.

Indeed, as of the time of Perry's confinement, one other circuit had found that a period of segregated confinement that was comparable to Perry's but was limited only by the inmate's sentence was itself not "indefinite" within the meaning of Wilkinson, given the limited duration of the sentence (which was itself two to four years) and that the inmate's placement in segregated confinement was reviewed more than annually. See DiMarco, 473 F.3d at 1343– 44. Nor had this Circuit or any other -- or for that matter any other court -- held that solitary confinement pending an out-of- state transfer was indefinite within the meaning of Wilkinson. In fact, while LaChance was decided in 2012 -- and thus after Perry's initial fifteen months of solitary confinement had ended -- it did have the benefit of the case law as it had developed as of that time. And LaChance held that ten months of solitary confinement pending out-of-state transfer did not implicate a liberty interest under clearly established law as it stood at that time (even though the SJC held that solitary confinement pending out-of-state transfer would implicate such a liberty interest going forward if it lasted for more than ninety days). LaChance, 978 N.E.2d at 1202 & n.8, 1207–08. The available precedent, therefore, did not

66

clearly establish that Perry's segregated confinement was "indefinite" under Wilkinson.27 D.

Accordingly, we agree with the defendants that it would not have been clear to a reasonable corrections officer at the time of Perry's confinement that, under Sandin and Wilkinson, such confinement implicated a state-created liberty interest under the Due Process Clause. The defendants are thus entitled on this ground to summary judgment based on qualified immunity.

V.

We therefore affirm the District Court's grant of summary judgment for the defendants.

-Concurring Opinions Follow-

27 Perry does argue, for the first time on the final page of his en banc brief, that the defendants are not entitled to qualified immunity for the less-than-ninety-day period that he remained in the SMU after LaChance was decided on November 27, 2012. But Perry develops no argument as to why it was clearly established following LaChance that the defendants were required to act faster in returning him to the general prison population or in providing him with the hearing required by LaChance. We also note that the defendants have represented that after LaChance was decided the DOC expeditiously began conducting hearings in accordance with that decision, that they prioritized hearings for prisoners who had been continuously held for the longest periods of time (which did not include Perry who had only returned to Massachusetts custody a couple months before LaChance was decided), that a hearing had been scheduled for Perry, and that he was returned to the general prison population at MCI-Shirley before his hearing was scheduled to occur.

67

LYNCH J., concurring in part. Appellant Jwainus Perry, convicted in 2004 of first-degree murder, and so confined in a maximum-security prison in Massachusetts, was initially placed in administrative confinement based on information that he posed a lethal threat to other inmates were he not segregated from the general population. For various other reasons, Perry's administrative confinement was continued, including in different prisons. In this lawsuit, he asserts his due process rights under the United States Constitution were violated.

I concur only in the court's holding under the second prong that the defendant state prison officials are entitled to qualified immunity. I do not concur in the remainder of the majority opinion.

In Pearson v. Callahan, 555 U.S. 223 (2009), the Supreme Court rejected the "two-step protocol" set forth in Saucier v. Katz, 533 U.S. 194 (2001), under which courts had to first decide whether alleged facts "ma[de] out a violation of a constitutional right" when resolving the issue of qualified immunity. Pearson, 555 U.S. at 232. The previous requirement of resolving the federal constitutional question first was rejected because, among other reasons, it violated "the general rule of constitutional avoidance" and "result[ed] in a substantial expenditure of scarce judicial resources on difficult questions" unnecessary to resolving the case. Id. at 236-41. The Pearson Court held instead

68

that judges should "exercise their sound discretion in deciding which of the two prongs of the qualified immunity analysis should be addressed first in light of the circumstances in the particular case at hand." Id. at 236.

The majority has chosen to exercise this discretion to address the first prong of the qualified immunity analysis before addressing the second prong. Under the second prong, in my view, the grant of qualified immunity is plainly compelled. I do not join the majority's choice to exercise that discretion here, not only because I have doubts about the correctness of its analysis and its proposed guidelines, but because I think that choice is both mistaken and inappropriate under the circumstances of this case.

This court should "think hard, and think hard again,"

before using its discretion to address the first prong of qualified immunity analysis when its holding depends solely on the second prong of the test. Camreta v. Greene, 563 U.S. 692, 707 (2011). This is in accordance with the "longstanding principle of judicial restraint [which] requires that courts avoid reaching constitutional questions in advance of the necessity of deciding them." Id. at 705 (quoting Lyng v. Nw. Indian Cemetery Protective Ass'n, 485 U.S. 439, 445 (1988)).

There are many reasons in this case why the court "should address only the immunity question." Id. First, because the court

69

finds that the appellees are entitled to qualified immunity protections on the basis that the law was not clearly established, the first prong analysis "ha[s] no effect on the outcome of the case." Pearson, 555 U.S. at 237. This is not a case where it is "difficult to decide whether a right is clearly established without deciding precisely what the existing constitutional right happens to be." Id. at 236 (quoting Lyons v. City of Xenia, 417 F.3d 565, 581 (6th Cir. 2005) (Sutton, J., concurring)). Indeed, the court's ultimate holding that the law was not clearly established in no way depends upon the majority's determination as to the first prong. The court should therefore decline to "engag[e] in the 'essentially academic exercise' of determining whether that right exists at all." Walker v. Prince George's County, 575 F.3d 426, 429 (4th Cir. 2009) (quoting Pearson, 555 U.S. at 237).

The majority's first prong analysis moreover turns what would be a "small case[] into [a] large one[]." Camreta, 563 U.S. at 707. Where an immunity question is "one that we can 'rather quickly and easily decide,'" Rasul v. Myers, 563 F.3d 527, 530 (D.C. Cir. 2009) (quoting Pearson, 555 U.S. at 239), principles of judicial restraint counsel against wading into "the more difficult question whether the facts make out a constitutional violation at all," Pearson, 555 U.S. at 239. The majority opinion easily disposes of the clearly established question, noting that, on multiple occasions prior to the appellant's confinement, this

70

court had suggested that the issue at hand "was the subject of dispute." Maj. Op. at 56.

Indeed, it is "far from obvious" that Perry's confinement constituted a constitutional violation. Pearson, 555 U.S. at 237. As the majority concedes, it "is not an easy task, given the available precedent," to determine whether Perry's confinement implicated a state-created liberty interest. Maj. Op. at 18; see Aref v. Lynch, 833 F.3d 242, 253 (D.C. Cir. 2016) (holding that the Supreme Court has "not define[d] the baseline from which to measure what is 'atypical and significant' in a particular prison system," and characterizing the circuit split on the issue). Because the appellees are "plainly entitled to prevail at the second step" of qualified immunity analysis, we should exercise our "discretion to skip the first step." Orn v. City of Tacoma, 949 F.3d 1167, 1174 (9th Cir. 2020).

Further, this court should refrain from "explor[ing] the first prong of the qualified-immunity test" where it involves "a task that the [Massachusetts] legislature and courts are better equipped to handle than this court." Waeschle v. Dragovic, 576 F.3d 539, 550 (6th Cir. 2009). The majority opinion explains that the liberty interest that Perry asserts is implicated here arises "from an expectation or interest created by state laws or policies." Maj. Op. at 14 (quoting Wilkinson v. Austin, 545 U.S. 209, 221 (2005)). Accordingly, the majority's first prong analysis

71

involves an inquiry into the practices of the Massachusetts prison system, the Massachusetts prison system's regulations, and the expectations the Massachusetts "laws and policies have generated about what an inmate reasonably should understand to constitute the basic experience of prison life." Maj. Op. at 35. These are matters within the expertise of the Massachusetts legislature and courts.

In fact, Massachusetts state law provides protection to inmates in various forms of administrative segregation which goes beyond any federal due process requirements. Perry himself has benefited from the 2012 decision of the Massachusetts Supreme Judicial Court (SJC) in LaChance v. Comm'r of Corr., 978 N.E.2d 1199 (Mass. 2012). The SJC held in that case that

an inmate confined to administrative segregation on awaiting action status, whether such confinement occurs in an area designated as an SMU [special management unit], a DSU [departmental segregation unit], or otherwise, is entitled, as a matter of due process, to notice of the basis on which he is so detained; a hearing at which he may contest the asserted rationale for his confinement;

and a posthearing written notice explaining the reviewing authority's classification decision.

Id. at 1206-07.

Since LaChance, Massachusetts inmates in various types of administrative segregation have been given even greater protections than those afforded by the federal Constitution, both

72

as a matter of Massachusetts Department of Correction (DOC) regulations and as a matter of state court decisions. Both these regulations and those judicial rules were established before this court granted en banc rehearing. In Cantell v. Comm'r of Corr., 60 N.E.3d 1149 (Mass. 2016), the SJC reaffirmed that LaChance was decided under federal constitutional principles, but that, as a matter of state law, inmates had additional rights derived from state law under longstanding judicial precedent in Haverty v. Comm'r of Corr., 776 N.E.2d 973 (Mass. 2002). Cantell, 60 N.E.3d at 1156-57. The SJC stressed that these protections arise as a matter of state law independent of any federal constitutional protections. See id.

In addition to these judicial decisions, inmates in all types of correctional facilities have had the added protection of Massachusetts DOC regulations governing any form of "non- disciplinary Restrictive Housing." 103 Mass. Code Regs. §§ 423.00, 423.04 (2019). The regulations set forth notice, hearing, and durational requirements for such non-disciplinary segregation. They also significantly limit when such segregation may take place. For example, the regulations provide that "[e]very inmate in Restrictive Housing for 30 days or more shall be provided with" a "review" hearing "within 30 days of his or her Restrictive Housing placement." Id. § 423.09(3)(a). The inmate must receive "48 hours written notice prior to the review . . . stat[ing] the

73

basis upon which the inmate is housed in Restrictive Housing," "the opportunity to participate in the review in person," and "a written statement as to the evidence relied on and the reasons for the placement decision if no placement change is ordered." Id. At the hearing, "the inmate may offer a verbal and/or written statement and/or submit documentation." Id. "Within two calendar days of the review," the officer who conducted it must issue a written "recommendation," "served on the inmate," which must include a "description of the underlying basis that led to" placement and a "determination whether the inmate's return to general population would pose an unacceptable risk to life, property, staff or other inmates, or to the security or orderly running of the institution." Id. The recommendation must also "generally describe [its] factual basis or bases . . . , including a brief description of any evidence relied upon," and "document whether the inmate made or submitted any statements or documents." Id. The inmate may appeal the recommendation. Id. For longer placements, additional review hearings with the same protections are provided "[w]ithin 90 days of . . . initial placement . . . and within every 90 days thereafter." Id. § 423.09(3)(c). For reviews "conducted 180 days after initial placement, and every 180 days thereafter, the inmate may request that the [review hearing] be recorded." Id.

74

These regulations provide additional protections. They provide that inmates in such administrative segregation must receive daily visits "by a member of the medical staff (unless medical attention is needed more frequently)." Id. § 423.12. Also, "[a]n inmate diagnosed with [a serious mental illness]" cannot be confined longer than 30 days unless it is determined that "the inmate poses an immediate and present danger to others or to the safety of the institution." Id. § 423.09(3)(b). Those inmates' status is reviewed three times a week, and the inmates may "participate in [those] review[s] in writing." Id. Similar additional procedures apply to inmates being segregated from general population "to protect [them] from harm by others," and, in addition, those inmates may not be placed in Restrictive Housing longer "than 72 hours unless the Commissioner or a designee" makes certain certifications. Id.

For inmates who are "awaiting the adjudication of disciplinary charges," there are additional placement reviews "held every 15 days," for which notice must be provided and in which the inmate may "participate . . . in writing."28 Id. Inmates in Restrictive Housing "for [o]ther [r]easons," like "pending

28 Inmates who are segregated in a Department Disciplinary Unit as the result of a disciplinary sanction are subject to the different procedures in 103 Mass. Code Regs. § 430 (2019). All uses of Restrictive Housing appear to be covered by the combination of 103 Mass. Code Regs. § 430 (2019) and 103 Mass. Code Regs. § 423 (2019), both promulgated in 2019.

75

investigation, pending classification, pending transfer, or refusing housing placement," have their status reviewed three times a week (and are, of course, entitled to participate in person at the 30-day and 90-day hearings that are afforded to all inmates regardless of the reason for their placement). Id.

The regulations also provide that Restrictive Housing may not be imposed in certain instances. Before placement, inmates must be screened by mental health professionals for serious mental illness "or to determine if Restrictive Housing is otherwise clinically contraindicated," id. § 423.08, in which case they cannot be segregated unless the Commissioner makes certain certifications, id. § 423.09(2)(a). Inmates with a "permanent physical disability that precludes placement in Restrictive Housing" and pregnant inmates also cannot be placed in Restrictive Housing. Id. §§ 423.08-09.

Because Massachusetts inmates in administrative segregation receive greater protection under state law than is required by federal due process, any future case on administrative segregation in Massachusetts is likely to "resolve[] into whether [Massachusetts] law authorize[s] the officer's action," and so likely will not benefit from our attempts to "clarify the law for the future." Tremblay v. McClellan, 350 F.3d 195, 200 (1st Cir. 2003).

76

For an additional reason, recognized in Pearson, the majority's analysis of the first prong constitutional question poses the risk of being both "premature[] and incorrect[]" due to inadequate briefing by the parties to this case.29 Pearson, 555 U.S. at 239 (quoting Lyons, 417 F.3d at 582 (Sutton, J., concurring)). The majority's determination that Perry's confinement constituted an "atypical and significant hardship" is based in part on what it calls the defendants' failure to make "any contrary showing regarding the frequency of use of such prolonged solitary confinement for administrative reasons." Maj. Op. at 43. The majority's decision not to "confine [its] inquiry to the clearly established prong of the qualified-immunity analysis" therefore "pos[es] a risk that [it has] decide[d] the [first prong] issue incorrectly." Barber v. Miller, 809 F.3d 840, 845 (6th Cir. 2015); see Evans v. Skolnik, 997 F.3d 1060, 1070 (9th Cir. 2021) ("As Pearson makes clear, we should not address an avoidable constitutional issue when the briefing is inadequate.").

As Chief Justice Roberts recently wrote: "out of adherence to a simple yet fundamental principle of judicial

29The briefing from both parties, in my view, is inadequate. For example, Perry's briefs to the en banc court never even mentioned the 2019 DOC regulations described above. Further, the defendants-appellees mentioned the regulations only in one line and without further argument in their reply brief to the en banc court. And their briefing was focused on the second prong.

77

restraint[,] [i]f it is not necessary to decide more to dispose of a case, then it is necessary not to decide more."30 Dobbs v. Jackson Women's Health Org., 597 U.S. 215, 348 (2022) (Roberts, C.J., concurring).

30 Indeed, the panel opinion, withdrawn when en banc review was granted, was correct in its analysis holding that the defendants were protected by qualified immunity because the law was not clearly established at the time. See Perry v. Spencer, 751 Fed. Appx. 7 (2018).

78

HOWARD, Circuit Judge, concurring. I agree with the unanimous judgment of the court that qualified immunity applies in this case. I write separately to note that, while I have reservations about the necessity of going further than the qualified immunity holding, I agree in large part with the majority's analysis of the liberty interest issue, and I agree that an inmate's solitary separation from the general prison population lasting longer than 30 days ordinarily implicates a liberty interest sufficient to require that the deprivation of that liberty must be attended by due process.