Brathwaite v. Phelps

• Court: District Court, D. Delaware
• Decided: March 30, 2023
• Docket: 1:10-cv-00646
• Status: Unknown

Opinion

IN THE UNITED STATES DISTRICT COURT FOR THE DISTRICT OF DELAWARE

KEVIN C. BRATHWAITE,

Plaintiff,

v. No. 10-cv-646-SB

WARDEN PERRY PHELPS, et al.,

Defendants.

Timothy Devlin, Andrew DeMarco, James Lennon & Veronica Shad, DEVLIN LAW FIRM LLC, Wilmington, Delaware

Counsel for Plaintiff

Nicholas Picollelli, Jr., DELAWARE DEPARTMENT OF JUSTICE, Wilmington, Delaware

Counsel for Defendants

MEMORANDUM OPINION March 30, 2022 BIBAS, Circuit Judge, sitting by designation. Process must precede punishment. Kevin Brathwaite has raised a genuine dispute about whether he got the process he was due before he was kept in solitary confinement year after year. And defendants’ obligation to give him that process was clearly established. So I grant his summary-judgment motion in part. I. BACKGROUND A. Factual background Unless otherwise noted, the following facts are undisputed. 1. Brathwaite’s transfer to solitary confinement. Brathwaite is serving six

consecutive life sentences plus 110 years in prison. D.I. 273-3, at 27; State v. Brathwaite, 2003 WL 1410155, at *1 (Del. Super. Ct. Mar. 17, 2003). In 1998, he was convicted of (among other offenses) sixteen counts of sex crimes after assaulting three different women. Brathwaite v. Phelps, 2009 WL 3345595, at *1 (D. Del. Oct. 16, 2009). At first, he was part of the general prison population. D.I. 273-2, at 14. But after marijuana was found in his cell in 2004, he was transferred to the secure-

housing unit—maximum-security solitary confinement. Id. at 26. (Technically, not everyone in secure housing was in solitary. See D.I. 273-9, at 15:11–22. But Brathwaite was. So I refer to them interchangeably.) The marijuana was neither Brathwaite’s first nor last problem. In prison-speak, defendants say he suffered from “[p]oor institutional adjustment.” See, e.g., D.I. 273- 3, at 16. Before his transfer to solitary, he had also been booked for bribery and “failure to obey.” See id. at 46; D.I. 273-2, at 28–29. After his transfer, he became

“very dangerous.” D.I. 273-6, at 96. According to defendants, he threw his urine and feces at staff, smeared his feces on his cell’s walls, refused to shower, once assaulted staff with a homemade weapon, and more. Id. at 96, 114, 126–27. Naturally, they considered him a safety risk and kept him in solitary. Id. at 45–46. 2. Brathwaite’s conditions in solitary. While in solitary, Brathwaite had almost no human contact. He did not have a cellmate. D.I. 273-1 ¶ 2. He ate meals alone in his cell. Id. ¶ 9. Defendants say he might have spoken with other prisoners through the vents or during recreation, but Brathwaite denies having any contact with other prisoners. Compare D.I. 273-1 ¶ 4, with D.I. 273-2, at 18:6–9. He was allowed one

fifteen-minute phone call per week, legal or otherwise. D.I. 273-1 ¶ 7. He could host two or three visits per month, but all were noncontact. Id. ¶ 6; D.I. 278-1 ¶ 6. “Noncontact” means that Brathwaite and his visitor were separated by a glass wall and spoke through a phone. D.I. 273-13, at 39:20–24; D.I. 273-7, at 17:9–21. Brathwaite did have some contact with prison officials. He talked to guards when they walked by his cell. D.I. 273-1 ¶ 5; D.I. 278-1 ¶ 5. And every ninety days, he spoke with a counselor about his quality-of-life level. D.I. 278-1 ¶ 5. Within the secure-

housing unit, there were multiple quality-of-life levels, with varying privileges. D.I. 273-13, at 42:8–23. Brathwaite was at the highest quality-of-life level. D.I. 278-1 ¶ 6; D.I. 273-2, at 16:2–23. So along with the telephone and visitation privileges described above, he got $45 in commissary credit every two weeks, a radio and television, access to some educational programs, and access to the law (but not general) library. D.I. 273-2, at 17:2–6, 21:21–24:10.

Despite his high quality-of-life level, his freedom of movement was still constrained. For forty-five minutes every other day, he was allowed out of his cell for recreation. D.I. 273-1 ¶ 3. Recreation took place both inside and outside, but just in empty cages set up for the purpose. D.I. 273-2, at 19:1–12. He could also take five fifteen-minute showers per week. D.I. 278-1 ¶ 3. Whenever he left his cell, he was handcuffed. D.I. 273-1 ¶ 3. 3. Brathwaite’s duration in solitary. By the time he filed this lawsuit in 2010, Brathwaite had been in solitary for nearly six and a half years. D.I. 273-2, at 14; D.I. 278-1 ¶ 1 (transferred to solitary Feb. 19, 2004); Compl. D.I. 2 (complaint filed

Aug. 2, 2010). His stay was continuous apart from a few brief interruptions: For three days in late 2004, he was downgraded to the medium-high-security housing unit, one step below secure housing. D.I. 278-1 ¶ 1. But he “expressed his contempt at the con- ditions … and made threats that he was going to do certain things,” so he was trans- ferred back to solitary. D.I. 273-6, at 101:11–12. After that, he spent two periods in the prison infirmary in late 2004 and early 2005, totaling about three months. D.I. 278- 1 ¶ 1. After February 24, 2005, his stay in the secure-housing unit was unbroken. Id.

His multiyear solitary confinement was, to say the least, unusual. One defendant testified that a “reasonable time frame” in solitary was “a matter of months.” D.I. 273- 6, at 42:14–17. Though an extended stay was not unheard of, another defendant tes- tified that a period of several years was “uncommon” and “rare.” D.I. 273-1 ¶ 28. 4. Review of Brathwaite’s secure-housing classification. Brathwaite’s secure- housing classification was reviewed annually. See generally D.I. 273-3. See also

D.I. 277, at 13; D.I. 273-8, at 23. These reviews had two parts. First, a two-person multidisciplinary team (consisting of a counselor and a guard) would recommend a classification. D.I. 273-8, at 16–17. Then, that team’s decision would be reviewed by the prison’s four-person classification committee. Id. at 17–18. According to prison policy, “[e]ach offender” whose classification was under review “shall be interviewed” and “must be present and … encouraged to participate.” Id. at 16, 21; D.I. 273-1 ¶¶ 10–11, D.I. 278-1 ¶¶ 10–11. The policy also required that the prisoner receive the final decision in person or in writing. D.I. 273-8, at 18, 21. No matter what the com- mittees decided, Delaware law gave the warden a veto. Id. at 21.

These reviews used a risk-assessment score. To find that score, a prison official would fill out a form. The form assigned points for the severity of the offense of incar- ceration, escape and misconduct history, age, and so on. See, e.g., D.I. 273-3, at 20– 21. The official would then tally up the score and match it to the risk-assessment scale. If a prisoner had a score from zero to eight, the default recommendation was minimum security; if nine to sixteen, medium security; if seventeen and up, maximum security. Id. But officials had discretion to override these defaults. Id.

At several classification reviews, Brathwaite’s risk-assessment score was low enough to qualify him for a security downgrade. D.I. 273-1 ¶ 14; D.I. 278-1 ¶ 14. But someone always overrode that score. Id.; see also D.I. 273-3, at 2, 5, 10, 13–14, 18, 21, 26, 34, 37. In 2009, his risk-assessment score qualified him for minimum security. D.I. 278-1 ¶ 14. But the classification committee kept him in maximum security and gave this explanation: “Brathwaite has a long history of not following the institution[’]s

rules. He continues to receive minor write-ups. [He] is on the Remain in SHU list and has been for some time. His behavior shows that he is still not ready to come off this list.” D.I. 273-3, at 4. His 2007 and 2008 reviews were much the same story. Though he qualified for medium security, he was kept in the secure-housing unit because of his 2004 staff assault and because he was on the remain-in-SHU list. Id. at 10–11, 13–14. This remain-in-SHU list is a point of contention. According to Brathwaite, the list guaranteed that he would be kept in the secure-housing unit, rendering the commit- tees’ reviews meaningless. Plus, he says it was impossible to get off the list. Defend-

ants counter that the list was merely a consolidated place to keep track of trouble- some prisoners. They say the list kept safety risks from slipping through the cracks, especially when staff turned over. See D.I. 278-1 ¶¶ 15–22. They admit that there was no official process for getting off the list, and one defendant testified that she had never known a prisoner to be removed from it. D.I. 273-1 ¶¶ 23, 25. No one knows exactly when the list was created, or when Brathwaite was added to it. But Brathwaite’s 2007, 2008, and 2009 review forms all mentioned the list to justify his

continued solitary confinement. See D.I. 273-3, at 4, 10, 14. As for the reviews themselves, Brathwaite testified that (before suing) he was never present for them. D.I. 273-1 ¶ 12; D.I. 278-1 ¶ 12. He also testified that, at least several times, he did not see the committees’ written decisions. D.I. 273-2, at 34:16– 22, 36:4–10, 59:10–60:12, 67:2–24, 76:4–22, 79:3–81:8. Over the years, he wrote many letters to the warden to get “taken off that remain-in-SHU list” and “get scheduled

for reclassification.” Id. at 38, 67–68, 100:14–18. Brathwaite testified that he never got a response. Id. at 37:7, 100:19–21. 5. This suit. So he sued. Under 42 U.S.C. § 1983, Brathwaite seeks damages, claim- ing that prison officials violated his Fourteenth Amendment due-process rights. See D.I. 2; D.I. 9; D.I. 235, at 3. At the motion-to-dismiss stage, I held that each classifi- cation review was a separate alleged violation. D.I. 235, at 6–9. Working backward from when he filed his complaint, his four most recent reviews were held on December 9, 2009; October 30, 2008; November 6, 2007; and October 2, 2006. D.I. 277, at 14; D.I. 273-3, at 2–18. By those points, he had been in solitary confinement for about

five years and ten months, four years and eight months, three years and nine months, and two years and eight months, respectively. 6. Defendants’ involvement. I previously dismissed two named defendants. D.I. 235, at 4–5. Five remain: Perry Phelps, David Pierce, James Scarborough, Linda Kemp, and Larry Savage. From 2008 to 2010, Phelps was the warden and Pierce was deputy warden. D.I. 278-1 ¶ 30. From 2006 to 2010, Scarborough was security super- intendent, Kemp was a counselor, and Savage was a classification officer. D.I. 273-6,

at 10; D.I. 273-9, at 10; D.I. 273-10, at 10. Defendants Kemp and Savage admit that they participated in Brathwaite’s classification reviews in 2008 and 2009. D.I. 278-1 ¶ 29. They also admit that Kemp participated in the 2007 review. Id. Scarborough, Phelps, and Pierce oversaw the classification process but did not always directly in- fluence classification decisions. Id. ¶ 30. Brathwaite also claims that each of these three had a hand in the remain-in-SHU list. D.I. 273-1 ¶ 30.

B. Legal background 1. Summary judgment. The parties now cross-move for summary judgment. D.I. 269, D.I. 271. “The court shall grant summary judgment if the movant shows that there is no genuine dispute as to any material fact and the movant is entitled to judgment as a matter of law.” Fed. R. Civ. P. 56(a). In general, a dispute is “genuine” if a reasonable jury could find for either side. Anderson v. Liberty Lobby, Inc., 477 U.S. 242, 251–52 (1986). (The analysis for this case is more complex, but we will cross that bridge when we come to it.) And a fact is “material” if it “could affect the out- come.” Lamont v. New Jersey, 637 F.3d 177, 181 (3d Cir. 2011). Defendants claim qualified immunity. D.I. 270, at 14–15, 19–20. In qualified-

immunity cases, there are two more summary-judgment requirements. First, I must “specify those material facts that are and are not subject to genuine dispute and ex- plain their materiality.” Forbes v. Township of Lower Merion, 313 F.3d 144, 146 (3d Cir. 2002). Second, I must “analyze separately, and state findings with respect to, the specific conduct of each [defendant].” Grant v. City of Pittsburgh, 98 F.3d 116, 126 (3d Cir. 1996). 2. Qualified immunity and clearly established law. Even if officials violate the

Constitution, they are entitled to qualified immunity unless the law was “clearly es- tablished.” Ziglar v. Abbasi, 582 U.S. 120, 151 (2017). In defining the relevant law, I must be careful “not to define [it] at a high level of generality.” Ashcroft v. al-Kidd, 563 U.S. 731, 742 (2011). Instead, I look to the “specific context of the case.” Mullenix v. Luna, 577 U.S. 7, 12 (2015) (per curiam). The “salient question” is whether “the state of the law” gave reasonable officials

“fair warning.” Hope v. Pelzer, 536 U.S. 730, 741 (2002). That warning comes through case law. A prior case need not address “the very action in question,” but it must make the action’s unlawfulness “apparent.” Abassi, 582 U.S. at 151 (internal quotation marks omitted). Cases with “fundamentally similar” facts are “strong support” that the violation would be apparent. Hope, 536 U.S. at 741. Other times, though, “general statements of the law … may apply with obvious clarity to the specific conduct in question.” Id. (internal quotation marks omitted). Either way, the alleged violation must have been clear to a reasonable officer when the action was taken. The hierarchy of relevant case law reflects “the Judiciary’s structure.” Id. at 747.

That is, I “look first for applicable Supreme Court precedent.” Porter v. Pa. Dep’t of Corr., 974 F.3d 431, 449 (3d Cir. 2020) (internal quotation marks omitted). If there is no such precedent or it leaves ambiguity, I consider “controlling authority in [the Third Circuit] or a robust consensus of cases of persuasive authority in the Courts of Appeals.” Id. Two more notes about qualified immunity. First, because clearly established law must set a question “beyond debate,” a circuit split is strong evidence that the law

was not clearly established. al-Kidd, 563 U.S. at 741. Defendants do not refer to any splits. But this area of law is riddled with them, so I will flag some along the way. Ultimately, none affects this case’s outcome. Second, the Third Circuit has not settled whether plaintiffs (or defendants) have the burden to show that the law was (or was not) clearly established. Compare Lozano v. New Jersey, 9 F.4th 239, 245 (3d Cir. 2021) (burden on plaintiff), with Jefferson v.

Lias, 21 F.4th 74, 80 (3d Cir. 2021) (burden on defendant). So I evaluate this question from a neutral starting point. With all that in mind, I begin with whether Brathwaite had a liberty interest pro- tected by the Fourteenth Amendment. II. BRATHWAITE HAD A CLEARLY ESTABLISHED LIBERTY INTEREST To show a due-process violation, Brathwaite must prove two things: first, that he had a liberty interest that could not be taken without due process; and second, that he did not receive the process he was due. And to overcome qualified immunity, he must show that his liberty interest and the process he was due were clearly established. I begin with the first question and conclude that Brathwaite did have a clearly

established liberty interest by the time of the 2007 violation. To reach that conclusion, I look at case law only before then (plus a few 2008 cases, which make the 2008 and 2009 violations even clearer). Finally, I also canvass more recent cases to ensure that no intervening law would change the liberty-interest finding. A. The liberty-interest standard 1. The test. Usually, a duly sentenced criminal has no right to complain about the roughness of prison life. But even an inmate has an interest in relative liberty. Under

the Fourteenth Amendment’s Due Process Clause, a prisoner’s liberty interest can arise in two ways. First, the Due Process Clause, “of its own force,” protects against punishments that exceed a prisoner’s sentence. Sandin v. Conner, 515 U.S. 472, 484 (1995). Second, states can create liberty interests. Like most prisoner-liberty cases, this one deals with the latter type of interest. In this category, prisoners have a due- process-protected interest in “freedom from restraint which … imposes atypical and sig-

nificant hardship on the inmate in relation to the ordinary incidents of prison life.” Id. This hardship standard was clearly established in Sandin v. Conner. But it is too abstract. Though a prison official likely knows what is “atypical,” “significant hard- ship” is malleable. So I must look to other cases to clarify this standard. Those other cases explain that hardship is a function of the confinement’s (1) conditions and (2) duration. Shoats v. Horn, 213 F.3d 140, 144 (3d Cir. 2000). And “the ordinary in- cidents of prison life” are measured by “‘routine’ prison conditions in … state institutions.” Id. Whether an inmate faced atypical and significant hardship is a ques- tion of law. Colon v. Howard, 215 F.3d 227, 230 (2d Cir. 2000). 2. Historical residue. For context, one should know that Sandin’s hardship stand-

ard represented a shift. Previously, in cases like Hewitt v. Helms, the Court had held that prison regulations’ “mandatory language” could create a liberty interest. 459 U.S. 460, 472 (1983). Sandin departed from that framework for finding a liberty in- terest. But as we will see later, Hewitt’s analysis about the process needed to vindicate a liberty interest survived. So pre-1995 cases remain relevant, but only for some purposes. This shift created a circuit split. By 2007, some circuits thought (and think) that Sandin created an additional requirement without disposing of the previous one.

That is, prisoners now had to point to both a mandatory policy and a substantial hardship. See Prieto v. Clarke, 780 F.3d 245, 249 & n.3 (4th Cir. 2015) (summarizing the split). This split does not affect Brathwaite’s claim for three reasons. First, de- fendants have not invoked it. Second, the Third Circuit’s stance has been clear. Its cases do not require any mandatory prison policy. Shoats, 213 F.3d at 143. Third, Brathwaite can point to a policy. Prison policy here required annual classification

reviews, with the prisoner’s participation, and a notice of a decision. D.I. 273-8, at 21. With Sandin’s standard in hand, I evaluate whether the conditions and duration of Brathwaite’s solitary confinement clearly imposed an atypical and significant hard- ship on him in relation to routine prison life. B. Conditions 1. The clearly established law. As we will see, most of the case law focuses on du- ration. As for conditions, the cases have not clearly held that certain conditions are necessary or sufficient to create a liberty interest. Instead, the cases find a liberty in- terest when a prisoner spends long enough in typical solitary-confinement conditions. Start with Supreme Court precedent. In Sandin, the Court held that the prisoner

in solitary confinement there did not have a liberty interest. But Sandin said rela- tively little about confinement conditions. Instead, it focused on the short duration, just 30 days. Sandin, 515 U.S. at 486. By contrast, in Wilkinson v. Austin, the Court held that the prisoners there had a liberty interest. 545 U.S. 209, 223 (2005). The Court emphasized the prison’s ban on “almost all human contact,” the twenty-four-hour artificial light in the cells, and the limit to one hour of indoor recreation each day. Id. at 223–24. On top of this “extreme

isolation” (and its “indefinite” duration, which I turn to later), the Court noted that the prison made inmates ineligible for parole. Id. at 214, 224. “While any of these conditions standing alone might not be sufficient to create a liberty interest, taken together they impose an atypical and significant hardship within the correctional context.” Id. at 224. So Wilkinson reveals what is sufficient, but not what is necessary. To clarify that, I turn to Third Circuit precedent.

By 2007, the seminal Third Circuit case on this subject was Shoats v. Horn. There, the court found that the prisoner had a liberty interest. It noted these conditions of confinement: “virtual isolation”; confinement to his cell for “23 hours a day, five days a week”; eating meals alone; no contact with his family; and no educational or other activities. Shoats, 213 F.3d at 144. Yet in an earlier Third Circuit case, Griffin v. Vaughn, the court found that a prisoner held in similar conditions did not have a liberty interest. 112 F.3d 703, 708 (3d Cir. 1997). These different outcomes are explained by the prisoners’ vastly different durations

in solitary. While Griffin had been in solitary for fifteen months, Shoats had done eight years. Again, I will explore duration in more detail below. But these cases reveal two things. First, solitary confinement imposes an atypical and significant hardship sufficient for a liberty interest—if paired with a substantial duration. Second, losing parole eligibility is not necessary: even though Shoats predates Wilkinson, Shoats remains good law and does not mention parole. By 2007, other circuits also viewed “standard SHU conditions” as sufficiently re-

strictive. Colon, 215 F.3d at 231–32; see also Magluta v. Samples, 375 F.3d 1269 (11th Cir. 2004) (finding a liberty interest for similar conditions). True, there was (and is) a circuit split relating to this question. The circuits disa- gree about the relevant baseline when determining “ordinary” prison conditions. Some circuits hold that administrative segregation is itself an ordinary incident of prison life, so it forms the baseline. Hatch v. District of Columbia, 184 F.3d 846, 851

(D.C. Cir. 1999) (summarizing the split). (Administrative segregation is yet another way to refer to solitary confinement. In contrast to disciplinary segregation, admin- istrative segregation is typically longer and not necessarily triggered by a specific violation of prison rules.) Yet this split does not affect Brathwaite’s claim for three reasons. First, defend- ants have not invoked it. Second, by 2007, the Third Circuit’s baseline was clearly established. True, some other circuits have cited Griffin for the proposition that the Third Circuit considers administrative segregation the relevant baseline. See, e.g., id. at 851. But Shoats came three years after Griffin, and it looked to “‘routine’ prison

conditions in Pennsylvania state institutions.” Shoats, 213 F.3d at 144. And when discussing Shoats’s confinement, the court compared it to the entire prisoner popula- tion. Id. Third, all circuits consider the duration of solitary confinement. See Harden- Bey v. Rutter, 524 F.3d 789, 793 (6th Cir. 2008) (Sutton, J.) (collecting cases). So even in circuits that consider administrative segregation the baseline, that just shifts the focus to duration. See, e.g., Hatch, 184 F.3d at 858. As I turn to Brathwaite’s conditions, keep in mind that they must have imposed

an atypical and significant hardship on him and that the conditions in Wilkinson and Shoats are good examples of such hardship. 2. Brathwaite’s conditions. Brathwaite’s conditions in solitary confinement were substantially harsher than those in the general population. Brathwaite was isolated: He was alone in his cell about twenty-three hours per day. Some days, he was not allowed to leave. Outside his cell, he was handcuffed. His recreation was limited to

bodyweight exercises in a cage. He had little to no contact with fellow prisoners. And though he was allowed some phone calls, educational programs, and visitors, those privileges were restricted. All these restraints distinguish his treatment from that of the general population, making his confinement clearly atypical. These limits are akin to those in Wilkinson and Shoats, clearly making his con- finement a significant hardship too. The table below compares Brathwaite’s conditions to those of the general population at Brathwaite’s prison and those of the inmates in Wilkinson and Shoats:

. General “778 Brathwaite . Wilkinson Shoats Population Cell size 8x12 ft Varies 7x14 ft Not mentioned

Meals Alone in cell Communal Alone in cell Alone in cell

Visits Non-contact Contact Non-contact Non-contact 45 min/every . other day, in- . 1 hr/day, 1 hr/day, Recreation door and out- Unrestricted only indoor 5 days/wk door, in cages Sieg One 15-min Unrestricted Not mentioned Only legal privileges call/wk or emergency Contact . with staff Yes Yes Not mentioned Yes Television . . : Yes Not in record Not mentioned No and radio Library Only law Law and Not mentioned Only law general ee Unrestricted Not mentioned None programs access Ineligible for Other ued parole; 24-hour artificial light See D.I. 273-2, at 14-25; D.I. 273-8, at 14; 273-14, at 48:3-6. Wilkinson, 545 U.S. at 214-15; Shoats, 213 F.3d at 142, 144. Of course, the conditions are not identical. For instance, the Wilkinson inmates’ cells had solid metal doors, so they could not interact with guards as they walked by. 545 U.S. at 214. Brathwaite could, and defendants say he might have spoken to other

prisoners through the vents too. Even so, his secure housing was “synonymous with extreme isolation.” Wilkinson, 545 U.S. at 214. True, there is one more significant difference. As mentioned, Wilkinson empha-

sized that, there, “otherwise eligible” prisoners lost parole eligibility. Id. at 224. That consequence, “taken together” with isolation and the duration of confinement, gave rise to a liberty interest. Id. Here, otherwise eligible prisoners in secure housing lost furlough eligibility. D.I. 273-8, at 12. But Brathwaite was not eligible for furlough anyway. See 11 Del. C. § 6538(e). In some cases, this distinction might make a difference. But Shoats shows that a collateral consequence, like losing parole eligibility, is not strictly necessary. Parole

revocation or not, Brathwaite’s conditions “differ[ed] significantly from ‘routine’ prison conditions.” Shoats, 213 F.3d at 144. His conditions were isolating and strik- ingly similar to other cases that found a protected liberty interest. And recall that the cases need not be carbon copies. “Fundamentally” or “materially” similar facts are “strong” support that the law was clearly established. Hope, 536 U.S. at 741. So Brathwaite’s conditions of solitary confinement were sufficient to create a liberty in-

terest, so long as they were paired with a substantial duration. C. Duration Now turn to the second hardship factor: duration. To support a liberty interest, Brathwaite’s duration in solitary confinement must have been sufficiently “prolonged and indefinite.” Porter, 974 F.3d at 450. (Though the cited case using this helpful phrase was published in 2020, these two considerations clearly framed the duration analysis by 2007—as the discussion below shows.) 1. About three years in solitary confinement was clearly established as prolonged. Again start with Supreme Court precedent. Sandin held that 30 days in solitary con- finement was insufficient. And Wilkinson dealt with a class action, so it did not ana-

lyze individual periods of confinement. Instead, it emphasized the confinement’s in- definiteness. We will turn to indefiniteness next. But the Third Circuit and other courts of appeals have also considered the prisoner’s actual duration in solitary con- finement, and whether that duration was prolonged. (As an aside, I note that this backward-looking approach is not how we typically think about due process. See, e.g., Wilkinson, 545 U.S. at 223 (describing the “state- created liberty interest in avoiding restrictive conditions of confinement” (emphasis

added)). But precedent requires it. And it makes some practical sense in the less- than-formal prison context, where periods of solitary confinement might not be clear at the outset. Plus, looking to actual outcomes to determine whether the process was sufficient is not unheard of. See, e.g., Scott v. Illinois, 440 U.S. 367, 373 (1979) (estab- lishing actual-imprisonment standard for right to counsel).) Turning to Third Circuit precedent, again the key case is Shoats. There, the court

held that “Shoats’ eight-year confinement subjects him to conditions that differ sig- nificantly from ‘routine’ prison conditions.” 213 F.3d at 144. Later Third Circuit opin- ions have consistently measured a given inmate’s period of solitary confinement against Shoats’s eight-year yardstick. See, e.g., Mitchell v. Horn, 318 F.3d 523, 532 (3d Cir. 2003) (string citation comparing durations); see also Fraise v. Terhune, 283 F.3d 506, 523 (3d Cir. 2002) (same). At the other end of the Third Circuit’s spectrum is Griffin. There, the court held that fifteen months in solitary confinement did not create a liberty interest. 112 F.3d at 708. (“Exposure to the conditions of administrative custody for periods as long as

15 months falls within the expected parameters of the sentence imposed [on him] by a court of law.” (emphasis added, alteration in original, internal quotation marks omitted)). In another case, Mitchell v. Horn, the court suggested that eighteen months would not create a liberty interest because the “marginal difference [of three months] d[id] not appear to cross the constitutional line.” Mitchell, 318 F.3d at 532. Nonethe- less, it remanded to develop the record. Id. So by 2007, the Third Circuit had set liberty-interest guideposts at fifteen (or eighteen) months and eight years. To fill in

the vast middle, I turn to other courts of appeals. By 2007, the circuits were divided over claims in the one-to-two-year range; yet as defendants themselves quote in their brief, “[t]he duration in segregated confinement that courts have found does not give rise to a liberty interest ranges up to two and one-half years.” D.I. 277, at 16 (emphasis added) (quoting Wilkerson v. Goodwin, 774 F.3d 845, 855 (5th Cir. 2014) (citing only pre-2009 cases)). Compare Jones v. Baker,

155 F.3d 810, 812 (6th Cir. 1998) (no liberty interest for two and a half years), Estate of DiMarco v. Wyo. Dep’t of Corr., 473 F.3d 1334, 1343–44 (10th Cir. 2007) (fourteen months), Hernandez v. Velasquez, 522 F.3d 556, 563–64 (5th Cir. 2008) (per curiam) (one year in a shared cell), Beverati v. Smith, 120 F.3d 500, 503–04 (4th Cir. 1997) (six months), Sealey v. Giltner, 197 F.3d 578, 589 (2d Cir. 1999) (101 days), Townsend v. Fuchs, 522 F.3d 765, 767, 772 (7th Cir. 2008) (fifty-nine days), and Skinner v. Cunningham, 430 F.3d 483, 486–87 (1st Cir. 2005) (forty days), with Fogle v. Pierson, 435 F.3d 1252, 1259 (10th Cir. 2006) (prisoner could state a due-process claim for three years), Kelly v. Brewer, 525 F.2d 394, 400–02 (8th Cir. 1975) (same), Harden-

Bey, 524 F.3d at 793 (same), and Trujillo v. Williams, 465 F.3d 1210, 1225 (10th Cir. 2006) (750 days states a claim), and with Tellier v. Fields, 280 F.3d 69, 80 (2d Cir. 2000) (514 days establishes a liberty interest), Magluta, 375 F.3d at 1282 (about 500 days), and Colon, 215 F.3d at 231–32 (305 days). By 2007, a “robust consensus of cases of persuasive authority in the Courts of Appeals” set the benchmark around three years. Porter, 974 F.3d at 449 (internal quotation marks omitted). So it was clearly established that solitary-confinement pe-

riods beyond three years gave rise to a liberty interest, especially when bolstered by indefiniteness. 2. Indefiniteness was also clearly established. By 2007, indefiniteness was also clearly established as a key part of the liberty-interest inquiry. In Wilkinson, the Su- preme Court found that the prisoners’ liberty interest was especially strong because the solitary confinement was “indefinite and … reviewed just annually”; it was “lim-

ited only by an inmate’s sentence.” Id. at 214–15, 224. The Court contrasted this in- definite administrative segregation with the defined thirty-day period of disciplinary segregation in Sandin. Id. at 224. Without a specified end date, the solitary confine- ment was literally indefinite. The Third Circuit has also emphasized indefiniteness. But the division between “prolonged” and “indefinite” is not clean. The former can signal the latter. As I noted earlier, in the fast-moving prison-discipline context, lacking a clear end date does not always mean that solitary confinement will continue indefinitely. For example, nearly all the no-liberty-interest cases involved segregation pending investigation.

See, e.g., Skinner, 430 F.3d at 486; Townsend, 522 F.3d at 771; Griffin, 112 F.3d at 705; Mitchell, 318 F.3d at 527; Jones, 155 F.3d at 812. Although the end date was not explicit, it was tied to an external force with its own timeline. So the period of con- finement was more definite. But when a prisoner is kept in solitary for less well-defined reasons, the Third Circuit has looked to prolongation to inform indefiniteness. In Shoats, the court de- scribed “eight years in administrative custody, with no prospect of immediate release

in the near future” as “indefinite segregation.” 213 F.3d at 144. Mims v. Shapp provides another important data point. 744 F.2d 946 (3d Cir. 1984). This pre-Sandin case’s liberty-interest discussion is still good law as far as it does not rely on regulations’ mandatory language. In that surviving part of the opin- ion, the court noted that five years in solitary plus “the prison administrator ha[ving] contemplated no change” showed that the “administrative segregation was of poten-

tially limitless duration.” 744 F.2d at 951. Other circuits have also viewed prolonged periods as indefinite. See, e.g., Harden- Bey, 524 F.3d at 793 (three years in solitary could state a due-process claim because it “not improbably” shows “that [the prisoner’s] placement remains ‘indefinite.’”). Thus, it was clearly established by 2007 that (1) a multiyear period in solitary confinement with no end date was considered indefinite and (2) an indefinite period in solitary confinement created a liberty interest.

3. Measuring the duration of Brathwaite’s confinement. Before considering the length of Brathwaite’s detention, I must first resolve a dispute about how to measure that time. Brathwaite argues that I should look at the full six and a half years be- tween his initial move to solitary and when he filed this lawsuit. D.I. 279, at 5–14. Defendants, on the other hand, point out that the statute of limitations is two years. D.I. 277, at 12–18. So they contend that I should look at only the two years of con- finement preceding the complaint. Id.

Both miss the mark. The relevant durations are measured on the dates that the violations occurred. In my motion-to-dismiss opinion, I held that “a discrete claim arose every time [Braithwaite] did not receive a review.” D.I. 235, at 7. So the dates on which the allegedly deficient classification reviews occurred (not, as Brathwaite argues, the date that the complaint was filed) are the relevant endpoints. But the starting point is not, as defendants claim, limited by the two-year statute

of limitations. The duration of Brathwaite’s solitary confinement is simply one of the facts supporting his claim. The limitations period bars untimely legal claims; it says nothing about facts or evidence that support a timely claim. Thus, the relevant period runs from when Brathwaite was initially placed in solitary to the time of each review. So the two-year statute of limitations began to run, at the earliest, on the date of each alleged violation. Recall that Brathwaite filed his complaint in 2010. As defendants concede, the 2008 and 2009 claims are timely. As for the 2007 claim, I will address Brathwaite’s tolling arguments at the end of this opinion. And at the time of the alleged 2007, 2008, and 2009 violations, Brathwaite had been in solitary confine-

ment for three years and nine months, four years and eight months, and five years and ten months. Nor do Brathwaite’s brief stints outside solitary change this analysis. Recall that after about seven months in solitary, Brathwaite was transferred to medium-high- security housing for just three days before being sent back. And shortly thereafter, he spent two periods in the infirmary, totaling about three months. Defendants point out these transfers but do not argue that they reset the solitary-confinement clock—

nor could they. Otherwise, a prison could transfer an inmate from solitary to the gen- eral population for one day each year and evade classification reviews entirely. Plus, during his times in the infirmary, he was still technically housed in solitary confine- ment. Even if I subtracted these few months from his total time in solitary, he still would have spent about three and a half years in solitary by his 2007 review. That duration would still clearly establish a liberty interest.

4. Brathwaite’s confinement was clearly prolonged and indefinite. By 2007, Brathwaite’s confinement was prolonged. Even defendants admit that his duration in solitary was atypical. D.I. 273-1 ¶ 28. And his forty-five-month period was already three times as long as that in Griffin and more than a year longer than the top end of the no-liberty-interest range described above. These differences are far from the “marginal” three months in Mitchell. 318 F.3d at 532. Brathwaite’s confinement was also indefinite. True, his initial transfer to solitary might have been for a defined disciplinary period (though the record reveals no pre- determined end date). D.I. 273-3, at 47. But it became indefinite administrative seg-

regation due to his “problematic behavior.” Id. at 37; see also D.I. 273-6, at 18–19; D.I. 273-13, at 68 (describing inmates’ ability to “flow down through the classification process” as not being on “a set time frame”). So his solitary confinement was literally indefinite. Two other facts also show that his detention was indefinite. First is prolongation. In Mims, the prisoner had been in solitary for five years. Likewise, by 2008, Brathwaite had been in solitary for around five years without a plan to move him. By

2007, it had been nearly four years, still more than enough to suggest indefiniteness. See Harden-Bey, 524 F.3d at 793. Second is the remain-in-SHU list. Recall that the parties dispute exactly how this list was used. But they do not dispute that the list influenced Brathwaite’s 2007, 2008, and 2009 reviews—the review forms themselves say so. And defendants do not dispute that it is at least rare, if not impossible, for an inmate on the list to return to the general population. So Brathwaite’s being on this

list also suggests that his solitary confinement was indefinite. Defendants cite some other cases, but they are unpersuasive. Many of those opin- ions are unpublished, from district courts, or both. See D.I. 277, at 16; D.I. 278, at 17; Hope, 536 U.S. at 747. And even then, they muster just one case holding that a solitary-confinement period as long or longer than Brathwaite’s did not create a lib- erty interest. See Jordan v. Fed. Bureau of Prisons, 191 Fed. Appx. 639, 642 (10th Cir. 2006). That case is unpublished, involves segregation pending investigation, and improperly mixes the state’s interest into the liberty-interest analysis. See id. at 642, 650–53. Defendants are too far out on a limb for it to support the weight of their

argument. * * * * * By 2007, Brathwaite had a clearly established liberty interest in avoiding a harsh, prolonged, and indefinite period of solitary confinement. His case is “materially”—if not “fundamentally”—similar to many other cases finding the same. Hope, 536 U.S. at 741. Though any one of Brathwaite’s restrictive conditions “standing alone might not be sufficient to create a liberty interest,” the conditions’ severity, duration, and

indefiniteness, “taken together[,] … impose[d] an atypical and significant hardship.” Wilkinson, 545 U.S. at 224. The same cannot be said for the alleged 2006 violation. By then, he had been in solitary for two years and eight months. That period is barely longer than the two and a half years at the upper end of the no-liberty-interest range. A reasonable officer could think that the two-month difference was “marginal,” like the three-month dif-

ference in Mitchell. So that claim misses the clearly established cutoff. See Brown v. Or. Dep’t of Corr., 751 F.3d 983, 988–90 (9th Cir. 2014) (finding right not clearly es- tablished because circuit had not yet held that twenty-seven months was sufficient). D. Post-2009 cases confirm Brathwaite’s liberty interest Cases after 2009 cannot clearly establish the law to officers at the time of the violations. But they could change my present evaluation of whether Brathwaite had a liberty interest. For example, if Wilkinson or Shoats were overturned after 2009, I would need to apply the new law. But nothing like that has happened here. To the contrary, post-2009 case law only confirms Brathwaite’s liberty interest. The Supreme Court has not elaborated on Wilkinson, but the Third Circuit has

issued some informative decisions since 2009. In Williams v. Pennsylvania Secretary of Department of Corrections, the court held that a prisoner in solitary confinement for six years had a liberty interest. 848 F.3d 549, 554–55 (3d Cir. 2017). It saw “no mean- ingful distinction” between that six-year period and the eight-year period in Shoats. Id. at 561. And it described his confinement as “indefinite.” Id. at 562. (The court went on to hold that the prisoner’s right was not clearly established, but that was because the prisoner was in the unusual situation of being housed on death row with a vacated

death sentence. Id. at 570–71. The case law had not yet addressed that situation. Id.) Similarly, in Porter v. Pennsylvania Department of Corrections, the Third Circuit held that a prisoner who had spent decades in solitary confinement had a liberty interest. 974 F.3d at 438. Though that prisoner’s solitary-confinement period was longer than Brathwaite’s, the court described the “wide consensus that prolonged and indefinite solitary confinement gives rise to a due process liberty interest.” Id. at 450.

It cited a string of cases for that proposition, many of which found liberty interests for solitary-confinement periods shorter than Brathwaite’s. Id. at 449–50. Plus, Porter confirmed that a claim can still be clearly established despite minor variations in conditions. Though Porter was “mindful that there [we]re some distinc- tions,” they were not enough to “distinguish [plaintiff’s case] for the purposes of his procedural due process rights.” Id. at 438–39. Other circuits agree. See Wilkerson, 774 F.3d at 855 (“Though there are some distinctions … notably that no parole ram- ifications appear to attach … there are material and substantial similarities.”); In- cumaa v. Stirling, 791 F.3d 517, 532 (4th Cir. 2015) (noting that Wilkinson “did not

engage in a point-by-point comparison of the conditions that inmates experienced in a supermax facility with the ordinary incidents of prison life.”). One recent Fourth Circuit case is particularly on point. In Smith v. Collins, the court found that a roughly four-year period in solitary could support a liberty interest. 964 F.3d 266, 281 (4th Cir. 2020). There, too, defendants said that the prisoner spoke with counselors and guards. But the court found that speaking with “[c]orrectional staff do[es] not obviate the need for social interaction.” Id. at 277.

As for duration, the court acknowledged that the plaintiff’s “period of segregated confinement is quite shy of” the decades-long periods in some other cases. Id. at 278. “But four years and three months is far longer than the thirty-day period at issue in Sandin,” and it “exceeds the length of various periods that other courts have found insufficient to trigger a liberty interest, which range[] up to two and one-half years.” Id. at 278–79 (internal quotation marks omitted, alteration in original). And it noted

that a four-year “duration of … confinement in administrative segregation … strengthen[ed] [plaintiff’s] evidentiary showing of indefiniteness.” Id. at 278. So too here. “[P]risoners need not languish in solitary confinement for decades on end in order to possess a cognizable liberty interest … . The four-plus years that [Brathwaite] spent in administrative segregation is significant enough to tip the scales in his favor, particularly in light of the other evidence of indefiniteness.” Id. at 269. III. BRATHWAITE HAD A CLEARLY ESTABLISHED DUE-PROCESS RIGHT Because Brathwaite had a liberty interest, some process was due. But due process is flexible. It involves case-specific balancing and, usually, is the kind of “extremely

abstract” right unlikely to be clearly established. Abassi, 582 U.S. at 151 (internal quotation marks omitted); Mathews v. Eldridge, 424 U.S. 319, 335 (1976). Yet in the solitary-confinement context, the Supreme Court and Third Circuit here boiled down the Constitution’s abstract standard to a series of concrete rules. Well before 2007, courts had clearly established the required minimum procedures: notice, an oppor- tunity to be heard, and an explanation of the decision. 1. Formal and informal procedures. Formal procedures are required “where the

right at stake is to be free from confinement.” Wilkinson, 545 U.S. at 225 (emphasis added). But when the liberty interest is in the nature of confinement, “the inquiry draws more on the experience of prison administrators, and … the State’s interest implicates the safety of other inmates and prison personnel.” Id. at 228–29. So “infor- mal, nonadversary procedures” suffice. Id. at 229. For instance, in Greenholtz v. In- mates of Nebraska Penal & Correctional Complex, the Court addressed the level of

process due for inmates seeking parole. 442 U.S. 1, 16 (1979); Wilkinson, 545 U.S. at 229. It held that only an opportunity to be heard and notice of “the reason for [pa- role’s] denial” were required, not a formal hearing and a summary of the evidence. Greenholtz, 442 U.S. at 15–16. 2. Initial transfers to solitary. The same is true in the most analogous cases here, Hewitt and Wilkinson. Hewitt dealt with initial transfers to administrative segrega- tion. 459 U.S. at 473. Because the “decision that an inmate … represents a threat to the institution’s security” involves “intuitive judgments,” the Court required “only an informal nonadversary review of evidence.” Id. at 474. That is, “[a]n inmate must merely receive some notice of the charges against him and an opportunity to present

his views to the prison official charged with deciding whether to transfer him to ad- ministrative segregation.” Id. at 476. At bottom, Hewitt requires the two pillars of due process: “notice … and an opportunity to be heard.” Wilkinson, 545 U.S. at 229 (summarizing Hewitt). Wilkinson required the same. There, the Court said that Greenholtz and Hewitt “provide[d] the appropriate model,” requiring notice, an opportunity to be heard, and an explanation of the decision. Id. at 228–29. In Wilkinson, the prisoners got “notice

of the factual basis leading to consideration for [supermax] placement and a fair op- portunity for rebuttal.” Id. at 226. The Court emphasized that “these are among the most important procedural mechanisms for purposes of avoiding erroneous depriva- tions.” Id. at 226. Those mechanisms were enough to “safeguard an inmate’s liberty interest.” Id. at 228. So Hewitt and Wilkinson both held that initial transfers to ad- ministrative segregation required notice, an opportunity to be heard, and some ex-

planation of the decision. 3. Periodic review. On top of that shared holding, Hewitt went a step further. It noted that “administrative segregation may not be used as a pretext for indefinite confinement,” so “officials must engage in some sort of periodic review.” Hewitt, 459 U.S. at 477 n.9. It added that this periodic “review will not necessarily require that prison officials permit the submission of any additional evidence or statements.” Id. But periodic review should ensure that the inmate stays in solitary only if he “re- mains a security risk.” Id. So Hewitt clearly established that meaningful periodic re- view was necessary.

Other circuits have reached the same conclusion, holding that Hewitt clearly es- tablished the process due for prisoners in extended solitary confinement. One case was published before 2007. Magluta, 375 F.3d at 1283–84 (“[A]mple federal law ex- isted … to give fair warning … that it was unconstitutional to hold Magluta in solitary confinement for 500 days for the purpose of punishment and with virtually no proce- dural protection in the form of periodic reviews.”). Others were published after 2007, but they confirm the scope of what Hewitt

clearly established well before then. See Selby v. Caruso, 734 F.3d 554, 560 (6th Cir. 2013) (“[S]ince Hewitt v. Helms, prison officials have been on notice that ‘administra- tive segregation may not be used as a pretext for indefinite confinement of an in- mate.’” (quoting Hewitt, 459 U.S. at 477 n.9)); Isby v. Brown, 856 F.3d 508, 529–30 (2017) (“After all, prison officials have been on notice since Hewitt that periodic re- views of administrative segregation are constitutionally required, and it is self-evi-

dent that they cannot be a sham.”); Toevs v. Reid, 685 F.3d 903, 916 (10th Cir. 2012) (“Since Hewitt, it has been clearly established that prisoners cannot be placed indef- initely in administrative segregation without receiving meaningful periodic re- views.”). Some meaningful periodic review, then, was clearly established by 2007. 4. Procedures to renew the term in solitary. By 2007, the Third Circuit had clarified Hewitt’s periodic-review requirement, requiring the same level of process as initial reviews. In Sourbeer v. Robinson, it rejected the distinction between “initial place- ments in restrictive housing [and] subsequent renewals of such placements.” 791 F.2d 1094, 1104 n.8 (3d Cir. 1986). It “s[aw] no basis … to conclude that less process is due

to [a] prisoner facing a second or third ‘term’ in restrictive housing than is due upon initial placement.” Id. These renewals still require “the most fundamental right of due process: a meaningful opportunity to be heard.” Id. at 1101 (emphasis in original). (And this pre-Sandin case remains good law on this point, clarifying Hewitt’s surviv- ing model of process.) So by 2007, it was clearly established that, before renewing a term in solitary, pris- oners were due notice, an opportunity to be heard, and some explanation of the decision.

5. Frequency of periodic reviews. Of course, what counts as another “term” of soli- tary is not always clear. And the law gives officials some flexibility. But the case law reflects clear norms. Shoats, Sourbeer, and Hewitt all involved monthly reviews. Shoats, 213 F.3d at 144–45; Sourbeer, 791 F.2d at 1101; Hewitt, 459 U.S. at 477 n.9; see also Mims, 744 F.2d at 949; Clark v. Brewer, 776 F.2d 226, 234 (8th Cir. 1985); Sheley v. Dugger, 833 F.2d 1420, 1427–28 (11th Cir. 1987).

Wilkinson involved annual reviews. But language in that opinion suggests that about one year between reviews is the longest permissible gap. Wilkinson, 545 U.S. at 224 (noting that the segregation was “indefinite and … reviewed just annually” (emphasis added)); see also Kelly, 525 F.2d at 400 (“[The] reasons for the segregation must not only be valid at the outset but must continue to subsist during the period of the segregation. … [A] reason for administrative segregation of an inmate that is valid today may not necessarily be valid six months or a year in the future.”) This range of discretion is reflected in Ninth Circuit precedent, which, by 2007,

had held that the top end of the range was somewhere between 120 and 365 days. Compare Toussaint v. McCarthy (Toussaint III), 801 F.2d 1080, 1101 (9th Cir. 1986), with Toussaint v. McCarthy (Toussaint V), 926 F.2d 800, 803 (9th Cir. 1990). So it was clearly established that periodic reviews had to happen at least once a year or so. * * * * * Since Hewitt and Sourbeer, it has been clearly established that, before renewing a prisoner’s term in solitary for another year, prison officials must meaningfully pro-

vide notice of the charges, an opportunity to be heard, and some explanation of the decision. (And as with the liberty-interest analysis, no intervening law changes this conclusion.) These requirements “are not elaborate, but they are real, and must be strictly complied with.” Hatch, 184 F.3d at 852 (internal quotation marks omitted). IV. THERE IS A GENUINE DISPUTE OVER WHETHER BRATHWAITE GOT DUE PROCESS By 2007, case law had clearly established both Brathwaite’s liberty interest and the process he was due. And as noted earlier, the facts supporting Brathwaite’s lib- erty interest are undisputed. So if the record (viewed in the light most favorable to Brathwaite) raises a genuine dispute about whether defendants did not provide the

required process, they are not entitled to summary judgment. The record does that. So I deny their summary-judgment motion. But Brathwaite wants me to go a step further and grant his motion. As he reads the record, he sees no genuine dispute over whether he got due process. Defendants have not done much to fight this point. But because the record is so foggy, Brathwaite

is not entitled to summary judgment either. A. Defendants are not entitled to summary judgment Brathwaite has pointed to enough in the record to create a genuine dispute over whether defendants fell short of Hewitt’s and Sourbeer’s requirements. First, he says he never got notice or an opportunity to be heard. Though his classification reviews did happen annually, he says he was never present for or given a chance to participate in those reviews. D.I. 273-1 ¶ 12; D.I. 278-1 ¶ 12; D.I. 273-2, at 69:5–10. Second, he

says officials did not give him any explanation of at least some of their decisions. See, e.g., D.I. 273-2, at 34:8–22, 76:11–22. And even if Brathwaite did receive his classification forms, their explanations were barebones. In 2009, his risk-assessment score qualified him for minimum secu- rity. But officials checked the box labeled “Other” under “Discretionary Overrides.” D.I. 273-3, at 5. As explanation, officials referred to the “Remain in SHU list,”

“Brathwaite[’s] … long history of not following the rules of this institution,” and his recent “minor write-ups.” Id. at 4–5. His reviews in 2007 and 2008 were much the same story. Though he qualified for medium security, the remain-in-SHU list and his 2004 assault on a staff member justified keeping him in solitary. Id. at 10–11. Again, it is unclear whether Brathwaite actually got these or any other explana- tions. But even if he did, “checking a preprinted box” accompanied by a “perfunctory” justification is unlikely to constitute “meaningful” review. Williams v. Hobbs, 662 F.3d 994, 1001–06 (8th Cir. 2011); Incumaa, 791 F.3d at 523. And if the remain-in- SHU list was used to shortcut or override other procedures, it also might have ren- dered the process meaningless.

Beyond the reviews, defendants mention Brathwaite’s letters to the warden and his consultations with his counselor, but neither of these afforded much protection. His letters merely sought process. Without notice, he could not hope to meaningfully rebut the charges. Plus, “nothing in the record shows that prison officials even con- sidered the claims [Brathwaite] raised in his letters.” Hatch, 184 F.3d at 852. In fact, Brathwaite says he stopped writing at one point because he never got a response. D.I. 273-2, at 47:16–23, 100:14–21.

And he spoke with his counselor about only his quality-of-life level, not his classi- fication. D.I. 273-2, at 67:16–69:7; D.I. 273-9, at 24:7–25:7; D.I. 278-1 ¶ 5. Even if they did discuss classification, Hewitt requires “an opportunity to present his views to the prison official charged with deciding whether to transfer him.” 459 U.S. at 476. Here, the warden and the two committees made that decision. Though Brathwaite’s coun- selor was a member of one of the committees, relaying his views through an interme-

diary is unlikely to be an adequate substitute for presenting his views himself. So on at least one version of events, defendants violated Brathwaite’s clearly es- tablished right. B. Neither is Brathwaite 1. The genuine-dispute burden-shifting framework. But that version of events re- lies almost entirely on Brathwaite’s own testimony. That is a thin reed on which to grant summary judgment. Strikingly, however, defendants have failed to refute that testimony. Instead, their summary-judgment briefs focus exclusively on the statute of limitations, Brathwaite’s liberty interest, and qualified immunity. They do not ad- dress whether Brathwaite got any process. Yet that omission does not automatically

mean that there is no genuine dispute over the process that Brathwaite got. To un- derstand why, I now explain the more complex meaning of “genuine dispute.” Whether a dispute is “genuine” depends on burden-shifting. The moving party must first, by “citing to particular … materials in the record,” show that “a fact cannot be … genuinely disputed.” Fed. R. Civ. P. 56(c)(1)(A); Celotex Corp. v. Catrett, 477 U.S. 317, 325 (1986). If the movant shows the absence of a genuine dispute, the bur- den shifts to the nonmovant to cite materials reestablishing it. Fed. R. Civ.

P. 56(c)(1)(A); Celotex, 477 U.S. at 325. Though I “need consider only the cited mate- rials,” I “may consider other materials in the record.” Fed. R. Civ. P. 56(c)(3). Sometimes, the burden never shifts. For instance, the movant’s evidence, when “viewed in the light most favorable to the opposing party,” might itself reveal a dis- pute or “fail[] to foreclose the possibility” of one. Adickes v. S. H. Kress & Co., 398 U.S. 144, 157 (1970); see also 10A Charles Alan Wright, Arthur R. Miller & Mary Kay

Kane, Federal Practice and Procedure § 2727.2 (4th ed. 2022). The movant is less likely to carry his initial burden when he relies on his own testimony or exclusive knowledge of the events. See Adickes, 398 U.S. at 157–61; Wright, Miller & Kane § 2727.2. When the movant fails to carry his burden, summary judgment must be denied, “even if the opposing party has not introduced contradictory evidence in re- sponse.” Wright, Miller & Kane § 2727.1. In essence, summary judgment is appropriate only when the truth is clear. See Poller v. Columbia Broad. Sys., Inc., 368 U.S. 464, 467 (1962). Particularly in constitutional cases, even the slightest doubt can “tip the balance”

against granting summary judgment. See Hunt v. Cromartie, 526 U.S. 541, 553 (1999); Hayes v. Cape Henlopen Sch. Dist., 341 F. Supp. 823 (D. Del. 1972); Wright, Miller & Kane § 2732.2. In general, trial courts should proceed with “caution” in as- sessing summary judgment and “may … deny summary judgment in a case where there is reason to believe that the better course would be to proceed to a full trial.” Anderson, 477 U.S. at 255; see also Wright, Miller & Kane § 2728. Here, “a fuller record is necessary … to decide properly the issues involved.” Wright, Miller & Kane § 2732.2.

2. The record is underdeveloped. Brathwaite’s evidence does not show the absence of a genuine dispute. Start with the Statements of Undisputed Facts. Brathwaite’s version says he “testified that he was not present for his classification hearings during his time in the SHU.” D.I. 273-1 ¶ 12 (emphasis added). Defendants denied that ac- count “as stated,” adding that “Brathwaite also testified that he could have spoken to Defendant Kemp about his classification on or around September 27, 2010.” D.I. 278-1

¶ 12. So what Brathwaite said might be undisputed, but not necessarily what happened. Brathwaite’s deposition adds to the uncertainty. For one thing, Brathwaite did not claim outright that he was never present for reclassification hearings. Rather, he said that he did not “recall being present for any of these classification hearings.” D.I. 273-2, at 67:4–9 (emphasis added). Similarly, a question during his deposition referred to a document that said, “Saw inmate on tier to perform classification MDT with Lieutenant Savage.” Id. at 92:17–21. When asked whether he recalled that meet- ing, Brathwaite said, “No, not at all.” Id. at 92:22–23. In another exchange, a question involved a document “claiming that [Brathwaite]

[was] always present at [his] classification.” Id. at 99:3–15. Brathwaite said, “No, that’s not true.” Id. at 99:16. But when pressed about whether Brathwaite’s counselor “came up to [his] cell … and talked to [him] about classification,” he relented: “It could be.” Id. at 99:17–20. True, this latter example was about Brathwaite’s 2010 review, which took place after he filed his complaint. But it at least creates uncertainty about whether previous reviews were also conducted in person at Brathwaite’s cell. Making matters even more uncertain, neither document was included as an exhibit to the

parties’ summary-judgment motions. So I am left in doubt. Thus, Brathwaite’s evidence, on its own, did not shift the burden to defendants. And even if it did, other parts of the record also create uncertainty. Defendants’ testimony contends that officials generally followed procedure. See, e.g., D.I. 273-6, at 15:9–12 (“[T]he warden made sure that the counselors would have face-to-face meetings with the inmates regarding their classification … when deci-

sions were being made.”); D.I. 273-9, at 13:10–16 (“Q. Were inmates ever present dur- ing the classification process? A. When we did the classification, yes. Q. Where did those classifications take place? … Outside their cell? A. Outside the cell.”); D.I. 273- 10, at 19:1–14 (similar, referring to reviews “on the tier”); D.I. 273-9, at 25:21–23 (“Q. Okay. So you would talk to an inmate before their classification hearing. Is that correct? A. Yes.”); D.I. 273-10, at 17:13–22 (“Q. … Were inmates present for their housing classification hearings? A. Yes. Q. Were they permitted to participate in the classification hearing? A. Yes.”). And one isolated statement even suggests that they followed some of the proce-

dures in Brathwaite’s case specifically. D.I. 273-9, at 24:20–24 (Defendant Kemp “s[aw] [Brathwaite] several times prior to classification to talk to him about his up- coming classification”). Plus, there are suggestions of a sort of informal appeals process, perhaps offering a partial opportunity to be heard. See, e.g., D.I. 273-6, at 14–19; id. at 53:14–19 (“They could write a letter to the warden or … some sort of a document where they could indicate that they wanted to appeal.”); id. at 55:12–15 (“So although an inmate may

not be permitted to appeal officially, they’re already making a sound, we would look into it, and sometimes changes were made.”). In looking at the whole record, the truth is less than clear. “One comes away from these depositions with nagging skepticism about whether there is anything [Brathwaite] could ever do to be released from Ad[ministrative] Seg[regation].” Proc- tor v. LeClaire, 846 F.3d 597, 613 (2d Cir. 2017). But that sense does not put the issue

beyond dispute. So a question remains for trial: What process did Brathwaite get? V. INDIVIDUAL LIABILITY Brathwaite alleges that his clearly established right was violated. But that pas- sive sentence must become an active one: Brathwaite must show defendants’ personal involvement in the violations. And to recover for the alleged 2007 violation, he must show that he is entitled to tolling. A. Active involvement As noted many pages ago now, I must “analyze separately” each defendant’s con- duct. Grant, 98 F.3d at 126. That requirement flows from § 1983’s tort-law roots. An

“essential element” of a “cause of action for … tort[] is that there be some reasonable connection between the act or omission of the defendant and the damage which the plaintiff has suffered.” Jutrowski v. Township of Riverdale, 904 F.3d 280, 289–90 (3d Cir. 2018) (internal quotation marks omitted). Building on that common-law founda- tion, the Supreme Court has required a showing of “direct responsibility.” Rizzo v. Goode, 423 U.S. 362, 376–77 (1976). In other words, the defendant’s individual con- duct must have played an “affirmative part” in the violation. Id.

A question about “the extent of each officer’s participation … is a classic factual dispute to be resolved by the fact finder.” Jutrowski, 904 F.3d at 291 (internal quota- tion marks omitted, alterations adopted). But that extent inquiry is distinct from a “dispute about the possibility of each officer’s participation.” Id. So to survive sum- mary judgment, there must be a genuine dispute over the officer’s “reasonable con- nection” to the violation. Id. at 289.

Few solitary-confinement due-process cases reach this stage. No Third Circuit case addresses what “personal involvement” means in this context. But other circuits have held liable those defendants who conducted deficient classification reviews or recommended that a prisoner stay in solitary. See Hobbs, 662 F.3d at 1006; Williamson v. Stirling, 912 F.3d 154, 171–72 (4th Cir. 2018). The circuits have also held that mul- tiple layers of review do not shield decisionmakers along the chain. See Hanrahan v. Doling, 331 F.3d 93, 97 n.4 (2d Cir. 2003); cf. Williamson, 912 F.3d at 172 n.14. B. Individual assessments Now consider the remaining defendants: Kemp, Savage, Scarborough, Phelps, and Pierce. All five had at least a reasonable connection to Brathwaite’s confinement.

Kemp’s and Savage’s roles in 2008 and 2009 are simple: they constituted the two- member multidisciplinary team for those reviews, and they both recommended con- tinued confinement. See 273-3, at 4, 10. Kemp was also part of the multidisciplinary team in 2007. Id. at 14. Savage’s involvement in 2007 is less clear. But he was a classifications officer by that time, making housing decisions and working with Kemp. See D.I. 273-10, at 10:24–11:3, 12:6–18, 16:2–24. So there is at least a triable issue on the extent of his involvement in Brathwaite’s 2007 review.

Direct involvement of Scarborough, Phelps, and Pierce is less straightforward. But Brathwaite has at least raised a genuine dispute about it. He says that all three had a hand in creating, maintaining, or using the remain-in-SHU list. Scarborough testi- fied that he put inmates’ names on the list. D.I. 273-6, at 68:5–70:6. Pierce testified that those in positions of authority, like himself, reviewed the list and could make recommendations about who to add or remove. D.I. 273-13, at 62:14–63:10. And

though Phelps denies having any memory of the list, other defendants’ testimony suggests that the “warden made the decision as to names going on and coming off the list.” Compare D.I. 273-13, at 54:21–23, with D.I. 273-14, at 51:21–53:3. See also D.I. 273-10, at 62:7–9. And the 2007, 2008, and 2009 review forms all identify the remain-in-SHU list as a justification. So this theory is sufficient. All three were involved in other ways too. Scarborough oversaw classifications from 2004 to 2010. D.I. 273-6, at 10:9–16:1. Pierce testified that part of his job was to “[e]nsur[e] that the [classification-review] process got done.” D.I. 273-13, at 45:22– 46:17. He also dealt with prisoner complaints and, at times, stood in for the warden. D.I. 273-13, at 48:9–49:12. Phelps, as warden, had a veto over every classification de-

cision and reviewed them extensively. D.I. 273-14, at 40:19–41:24, 85:1–87:5, 93:4–10. But Phelps and Pierce were not involved with the remain-in-SHU list or classifi- cations generally until they became warden and deputy warden in 2008. And Brathwaite has not connected their pre-2008 roles to his classification reviews. So a genuine dispute remains about the extent of involvement in the alleged 2008 and 2009 violations for all five defendants. As for the 2007 violation, only the claims against Kemp, Savage, and Scarborough survive—if the 2007 claim overcomes the

statute of limitations. C. There is a genuine dispute over whether to toll the limitations period for the 2007 claim The parties agree that Delaware’s two-year statute of limitations for personal- injury torts applies. D.I. 279, at 2. Brathwaite filed his complaint in 2010. So if Brathwaite’s claim accrued in 2007, it would be untimely unless I toll the statute of limitations. Brathwaite argues both that his claim accrued later and that the limita- tions period should be tolled. I previously held that each of Brathwaite’s claims accrued “when [he] knew or should have known of the injury upon which the action is based.” D.I. 235, at 5–6 (quot-

ing Sameric Corp. of Del., Inc. v. City of Philadelphia, 142 F.3d 582, 599 (3d Cir. 1998)). As for tolling, I apply Delaware law. See Wallace v. Kato, 549 U.S. 384, 394 (2007). Brathwaite claims two grounds for tolling: equitable tolling and fraudulent concealment. D.I. 279, at 2–3; see AmerisourceBergen Corp. v. Lebanon Cnty. Emps.’ Ret. Fund, 243 A.3d 417, 430 (Del. 2020). Equitable tolling is warranted if Brathwaite “was prevented in some extraordinary manner from timely asserting his rights.” Ow-

ens v. Carman Ford, Inc., 2013 WL 5496821, at *2 (Del. Super. Ct. Sept. 20, 2013) (internal quotation marks omitted). And fraudulent concealment involves defendants’ (1) preventing Brathwaite from learning material facts or (2) misleading him. In re Tyson Foods, Inc. Consol. S’holders Litig., 919 A.2d 563, 585 (Del. Ch. 2007). Brathwaite has not justified equitable tolling. For this argument, he points to his solitary confinement and that he was reasonably diligent in writing letters. But he fails to explain how his confinement prevented him from asserting his rights “in some

extraordinary manner.” Rather, as defendants point out, Brathwaite was a repeat litigant between 2004 and 2010. D.I. 281, at 7 (collecting cases). So he has not justi- fied equitable tolling. But a genuine dispute underlies his other arguments. Brathwaite makes both a delayed-accrual and a fraudulent-concealment argument. Here, they boil down to the same thing: he says he should not have known of his injury by 2007 (so I should delay

accrual) because defendants hid material facts from him (so I should toll for fraudu- lent concealment). He says prison officials told him that he would be moved out of solitary once his risk-assessment score was low enough. D.I. 279, at 6. Not until 2008, he says, did he have a reason to believe that was false. Id. Only then did he find out about the remain-in-SHU list. Id. Given that he allegedly “did not receive notices” or “get to see anything” about his reviews, defendants “kept [him] in the dark regarding the timing and occurrence of required annual reviews.” Id. at 6–7. Brathwaite has a point, but his framing is off. He says I should toll the limitations

period because he did not know of any “systematic procedural irregularity” in his classification reviews until 2008. D.I. 279, at 7 (quoting Bowen v. New York, 476 U.S. 467, 480–81 (1986)). But this is not a pattern-or-practice case; Brathwaite did not need to know about systematic irregularity. Instead, the upshot of his allegations is that he might have had good reason not to know (or defendants might have misrep- resented) that his confinement was indefinite. As we have seen, indefiniteness is a material fact for his claim. So he has raised a genuine dispute about whether he

should have known (or defendants concealed or misled him about) this material fact. * * * * * Brathwaite was held in solitary confinement for years on end. Prison officials can do that. But the Constitution requires that they give the prisoner a chance to per- suade them otherwise. Brathwaite says they did not give him that chance. Because it was clearly established that officials had to do so, we will go to trial.