Dewalt v. Hooks

IN THE SUPREME COURT OF NORTH CAROLINA 2022-NCSC-105 No. 165A21 Filed 4 November 2022 ROCKY DEWALT, ROBERT PARHAM, ANTHONY MCGEE, and SHAWN BONNETT, individually and on behalf of a class of similarly situated persons v. ERIK A. HOOKS, in his official capacity as Secretary of the North Carolina Department of Public Safety, and the NORTH CAROLINA DEPARTMENT OF PUBLIC SAFETY Appeal pursuant to N.C.G.S. § 7A-27(a)(4) from an order denying plaintiffs’ motion for class certification entered on 22 February 2021 by Judge James E. Hardin Jr. in Superior Court, Wake County. Heard in the Supreme Court on 30 August 2022. ACLU of North Carolina Legal Foundation by Daniel K. Siegel and Kristi Graunke, for plaintiff-appellants. Joshua H. Stein, Attorney General, by Orlando L. Rodriguez, Special Deputy Attorney General, Mary Carla Babb, Special Deputy Attorney General, and James B. Trachtman, Special Deputy Attorney General, for defendant-appellees. Aviance Brown, Irving Joyner, Daryl Atkinson, Whitley Carpenter, and Ashley Mitchell, for North Carolina Conference of the NAACP, amicus curiae. Lockamy Law Firm, by Scott Holmes; and Roderick & Solange MacArthur Justice Center, Northwestern Pritzker School of Law, by Daniel Greenfield, Bradford Zukerman, and Kathrina Szymborski, for Professors Sharon Dolovich, Alexander A. Reinert, Margo Schlanger, and John F. Stinneford, amici curiae. Nichad Davis and Benjamin I. Friedman, Professors and Practitioners of Psychiatry, Psychology, and Medicine, for amici curiae. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court NEWBY, Chief Justice. ¶1 In this case we consider whether the trial court erred by denying plaintiffs’ motion for class certification. Plaintiffs are inmates in North Carolina Department of Public Safety (DPS) custody. Plaintiffs brought a class action lawsuit against defendants seeking to represent certain individuals in DPS custody who are being or will be subjected to solitary confinement. Plaintiffs do not challenge the use of solitary confinement in every housing setting or allege that solitary confinement is per se unconstitutional. Rather, plaintiffs allege that defendants’ policies and practices concerning specific restrictive housing assignments violate the state constitution. The trial court denied plaintiffs’ motion for class certification. The trial court concluded plaintiffs failed to establish a common predominating issue, plaintiffs did not demonstrate that the named representatives would fairly and adequately represent the class, and that litigating as a class was not the superior method of adjudication. Plaintiffs appealed directly to this Court. Because the trial court did not abuse its discretion, we affirm the trial court’s order. ¶2 On 16 October 2019, plaintiffs1 filed a class action lawsuit seeking to certify a class of current and future inmates assigned to one of five restrictive housing classifications. Plaintiffs alleged the conditions of confinement across the five 1 Plaintiffs are Rocky Dewalt, Robert Parham, Anthony McGee, and Shawn Bonnett. Plaintiffs sought to appoint Robert Parham, Anthony McGee, and Shawn Bonnett as class representatives and requested that Rocky Dewalt remain a named plaintiff. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court restrictive housing assignments presented the same substantial risk of harm to all individuals and constituted cruel or unusual punishment. ¶3 The five challenged restrictive housing settings are: Restrictive Housing for Disciplinary Purposes (RHDP), Restrictive Housing for Control Purposes (RHCP), High Security Maximum Control (HCON), Restrictive Housing for Administrative Purposes (RHAP), and the first two phases of the Rehabilitative Diversion Unit (RDU). ¶4 RHDP is a short-term placement and “presumptive sanction” for disciplinary infractions, such as disobeying an order, possessing a cell phone, refusing a drug test, or using disrespectful or defamatory language. Individuals assigned to RHDP may have personal property in their cells, are allowed limited telephone privileges, receive visitation rights, and have access to cell study materials, such as educational programs and college coursework. Prison staff may impose up to twenty or thirty days of confinement in RHDP. Between October 2018 and October 2019, the average length of a placement in RHDP was eleven days. ¶5 RHCP “is a long-term restrictive housing assignment for the removal of [an incarcerated person] from the general offender population to confinement in a secure area.” RHCP is reserved for offenders who have displayed “disruptive behavior, assaultive actions, threats to the safety of staff or other offenders, or threats to the security and operational integrity of the facility.” People in RHCP receive one hour of DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court recreation time five days a week and have access to a shower three times a week. They eat all meals in their cell, may not attend religious, educational, or vocational programs outside of their cell, and have no guaranteed telephone or canteen access. People in RHCP are entitled to two noncontact visits every thirty days, but visitation privileges are suspended for at least twelve months if an individual is found guilty of assault on a staff member resulting in physical injury. RHCP classifications are reviewed every six months. If placement in RHCP is due to assault on a staff member resulting in physical injury, however, assignments are reviewed at twelve months. Between October 2018 and October 2019, the average length of stay in RHCP was 131 days. ¶6 HCON is the most restrictive housing assignment and is for “offenders who pose the most serious threat to the safety of staff and other offenders or who . . . require more security than can be afforded in [other housing settings].” Review of an HCON classification occurs every six months, or it occurs every twelve months if placement is due to assault on a staff member resulting in physical injury. People who are removed from HCON are automatically placed in RHCP, RDU, or the Therapeutic Diversion Unit. Between October 2018 and October 2019, the average length of stay in HCON was 154 days. ¶7 RHAP is a temporary placement for administrative, rather than disciplinary, purposes. Individuals may be placed in RHAP to protect staff members and other DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court offenders from threats of harm, to minimize the risk of escape, to preserve order, to provide control while completing an investigation, or to serve as a “cooling off measure[,]” as referred to in the policy. While assigned to RHAP, individuals have access to medical and mental health services, receive daily visits from a health care staff member, may have personal property in their cells, and are allowed telephone privileges. In addition, individuals in RHAP may receive an unlimited number of one-hour, noncontact visits. Between October 2018 and October 2019, the average length of stay in RHAP was eight days. ¶8 RDU is a placement program “designed as a safe alternative to segregation, providing positive reinforcements to increase desired behaviors, and decrease unwanted behaviors through . . . appropriate consequences . . . [and] positive reinforcement.” Individuals in RDU housing are allowed certain authorized personal property in their units, such as pencils, pens, books, a radio, a deck of cards, and hygiene items. From October 2018 to October 2019, the average length of stay in RDU was between twelve and fourteen months. ¶9 Defendants filed their answer on 21 January 2020. On 4 February 2020, the matter was designated as exceptional pursuant to Rule 2.1 of the General Rules of Practice for the Superior and District Courts and assigned to Judge James E. Hardin Jr. Plaintiffs filed their motion for class certification on 24 April 2020 pursuant to Rule 23(a) of the North Carolina Rules of Civil Procedure. See N.C.G.S. § 1A-1, Rule DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court 23(a) (2021). Plaintiffs thereafter took discovery and submitted evidence in support of their motion. Defendants filed their response in opposition with supporting evidence on 12 August 2020. The trial court held a Webex hearing on 1 December 2020 and heard oral argument from both parties. ¶ 10 On 22 February 2021, the trial court denied plaintiffs’ motion for class certification and found that a certifiable class did not exist for three independent reasons: (1) plaintiffs failed to demonstrate a common predominating issue among the group of potential class members, (2) plaintiffs did not establish that the named representatives would fairly and adequately represent the interests of all class members, and (3) litigating this case as a class action was not the superior method of adjudication. Plaintiffs appealed directly to this Court under N.C.G.S. § 7A-27(a)(4). ¶ 11 This Court reviews a trial court’s class certification order for abuse of discretion. Fisher v. Flue-Cured Tobacco Coop. Stabilization Corp., 369 N.C. 202, 209, 794 S.E.2d 699, 706 (2016). “[T]he test for abuse of discretion is whether a decision ‘is manifestly unsupported by reason[ ]’ or ‘so arbitrary that it could not have been the result of a reasoned decision . . . .’ ” Frost v. Mazda Motor of Am., Inc., 353 N.C. 188, 199, 540 S.E.2d 324, 331 (2000) (quoting Little v. Penn Ventilator Co., 317 N.C. 206, 218, 345 S.E.2d 204, 212 (1986)). “Within this general standard, when addressing a class certification order, this Court has recognized that conclusions of law are reviewed de novo, and findings of fact are considered binding if supported by DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court competent evidence.” McMillan v. Blue Ridge Cos., 379 N.C. 488, 2021-NCSC-160, ¶ 7 (citing Fisher, 369 N.C. at 209, 794 S.E.2d at 706). ¶ 12 Rule 23 of the North Carolina Rules of Civil Procedure authorizes class action lawsuits. Rule 23 provides that “[i]f persons constituting a class are so numerous as to make it impracticable to bring them all before the court, such of them, one or more, as will fairly insure the adequate representation of all may, on behalf of all, sue or be sued.” N.C.G.S. § 1A-1, Rule 23(a).2 “The party seeking to bring a class action under Rule 23(a) has the burden of showing that the prerequisites to utilizing the class action procedure are present.” Crow v. Citicorp Acceptance Co., 319 N.C. 274, 282, 354 S.E.2d 459, 465 (1987) (footnote omitted). First, the class representatives must demonstrate the existence of a class. Id. at 277, 354 S.E.2d at 462. “A proper class exists ‘when the named and unnamed members each have an interest in either the same issue of law or of fact, and that issue predominates over issues affecting only individual class members.’ ” Fisher, 369 N.C. at 209, 794 S.E.2d at 705 (quoting Crow, 319 N.C. at 280, 354 S.E.2d at 464). A common issue predominates when plaintiffs demonstrate that the potential class members’ claims share a common issue capable of resolution “in one stroke.” Wal-Mart Stores, Inc. v. Dukes, 564 U.S. 338, 350, 1131 2 There are notable differences between Rule 23 of the North Carolina Rules of Civil Procedure and Rule 23 of the Federal Rules of Civil Procedure governing class action lawsuits. Nonetheless, the federal cases which address the provision of the federal rule that is similar to the state provision are instructive to our analysis. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court S.Ct. 2541, 180 L. Ed. 2d 374 (2011) (providing that plaintiffs’ “claims must depend upon a common contention . . . capable of class[-]wide resolution—which means that determination of its truth or falsity will resolve an issue that is central to the validity of each one of the claims in one stroke”). ¶ 13 In addition to this initial requirement, the class representatives must show: (1) that they will fairly and adequately represent the interests of all members of the class; (2) that they have no conflict of interest with the class members; (3) that they have a genuine personal interest, not a mere technical interest, in the outcome of the case; (4) that they will adequately represent members outside the state; (5) that class members are so numerous that it is impractical to bring them all before the court; and (6) that adequate notice is given to all class members. Fisher, 369 N.C. at 209, 794 S.E.2d at 705–06 (internal quotations omitted) (quoting Faulkenbury v. Teachers’ & State Emps.’ Ret. Sys., 345 N.C. 683, 697, 483 S.E.2d 422, 431 (1997)). ¶ 14 When a party seeking class certification meets these prerequisites, “it is left to the trial court’s discretion ‘whether a class action is superior to other available methods for the adjudication of th[e] controversy.’ ” Id. at 209, 794 S.E.2d at 706 (alteration in original) (quoting Beroth Oil Co. v. N.C. Dep’t of Transp., 367 N.C. 333, 337, 757 S.E.2d 466, 470 (2014)). Class actions should be permitted where they are likely to serve useful purposes such as preventing a multiplicity of suits or inconsistent results. The usefulness of the class action device must be balanced, however, against DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court inefficiency or other drawbacks. . . . [T]he trial court has broad discretion in this regard and is not limited to consideration of matters expressly set forth in Rule 23 or in [existing case law]. Crow, 319 N.C. at 284, 354 S.E.2d at 466. As such, “the touchstone for appellate review of a Rule 23 order . . . is to honor the ‘broad discretion’ allowed the trial court in all matters pertaining to class certification.” Frost, 353 N.C. at 198, 540 S.E.2d at 331 (citing Crow, 319 N.C. at 284, 345 S.E.2d at 466). ¶ 15 Here the trial court identified three distinct bases for denying plaintiffs’ motion for class certification: (1) no common predominating issue; (2) inadequacy of plaintiffs as class representatives; and (3) a class action is not a superior method of adjudication. Any of the three independent bases would have been adequate to support the denial of class certification. However, because we conclude the trial court did not abuse its discretion in determining there is no common predominating issue, we limit our review to that basis. ¶ 16 The question here is whether the trial court abused its discretion in concluding plaintiffs failed to demonstrate a common predominating issue among the proposed class members. The trial court determined plaintiffs presented insufficient evidence to connect DPS’s practices and policies to an alleged risk of harm. In an attempt to support their claim that DPS’s practices caused all class members to face risks of similar harm, plaintiffs relied on four studies. The trial court found, however, that only two studies concerned DPS and only one addressed its restrictive housing DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court practices. One report, “suggest[ing] that exposure to restrictive housing is associated with an increased risk of death during community reentry[,]” provided insufficient evidence to support plaintiffs’ claim because it was a correlational analysis that, by the authors’ admission, could not support conclusions of causation. Additionally, the study was based on observational data that failed to consider confounding factors which could have affected the study’s ultimate outcome. Accordingly, this study could not provide concrete support to plaintiffs’ claim that restrictive housing causes an increase in the risk of post-release mortality. ¶ 17 Likewise, the trial court concluded the second relevant report (the Vera Report), which was prepared by the Vera Institute of Justice, was insufficient to connect DPS’s practices to the alleged risk of harm.3 The Vera Report commended DPS on its previous reform efforts, suggested that DPS “continue[ ] implementation of [both] current and future reforms,” and noted that DPS’s restrictive housing population decreased by 10% in the year following the study. Furthermore, as the trial court concluded, all but one of DPS’s policies discussed in the Vera Report has since been revised. 3 In 2016 DPS partnered with the Vera Institute of Justice to evaluate DPS’s restrictive housing policies and practices. The Vera Report “outline[d] the findings of th[e] assessment and provide[d] recommendations to [DPS] on how to safely reduce its use of restrictive housing.” The experiences of the named plaintiffs and other affiants, who have collectively experienced each of the five restrictive housing settings, generally align with the Vera Report’s findings. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court ¶ 18 Aside from these two reports, plaintiffs failed to present additional evidence, such as specific studies and expert witness reports to support their claim that DPS’s policies and practices create a uniform risk of harm to individuals assigned to each of the challenged restrictive housing settings. This lack of evidentiary support is distinguishable from the evidence presented by the claimants in many of the federal cases upon which plaintiffs rely. See Parsons v. Ryan, 754 F.3d 657, 669, 678 (9th Cir. 2014) (presenting numerous expert reports and ten specifically defined policies to which all class members were subjected); see, e.g., Braggs v. Dunn, 257 F. Supp. 3d 1171, 1236 (M.D. Ala. 2017); Davis v. Baldwin, No. 3:16-CV-600-MAB, 2021 WL 2414640 (S.D. Ill. June 14, 2021). Based upon the minimal evidence specific to DPS’s restrictive housing practices, the trial court did not abuse its discretion in concluding plaintiffs failed to establish that the potential class members’ claims share a common issue capable of resolution “in one stroke.” See Wal-Mart Stores, Inc., 564 U.S. at 350. ¶ 19 The trial court also concluded that the variety of penological purposes across the challenged housing classifications are fundamental distinctions that prevent a finding that a common issue predominates across such a broad class. Further, the circumstances which necessitate placement in restrictive housing and the length of each assignment require an individualized assessment that preclude finding a common predominating issue. ¶ 20 The lack of a “legitimate penological justification” is relevant in analyzing a DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court conditions-of-confinement claim. See Porter v. Clarke, 923 F.3d 348, 362–63 (4th Cir. 2019). The record evidence supports the trial court’s conclusion and demonstrates that each challenged housing setting serves a distinct purpose. The record shows that RHDP is used exclusively for disciplinary purposes and is reserved for incarcerated individuals who have committed a disciplinary infraction, while RHAP serves administrative purposes, such as to protect staff, minimize the risk of escape, and preserve order. Unlike both RHDP and RHAP, the purpose of RHCP is to manage incarcerated individuals who have demonstrated a risk to the operations of a facility. Alternatively, HCON is reserved for individuals who pose the most serious threat and require an increased level of security over that offered by the other settings. Finally, the purpose of RDU is to discourage unwanted behaviors through appropriate consequences and positive reinforcement. The penological purposes served by each housing setting thus inform the placement of an individual into the appropriate classification, which necessarily requires an individualized assessment. As such, the trial court did not abuse its discretion in concluding that the varying penological purposes precluded a finding that plaintiffs established a common predominating issue. ¶ 21 The trial court next concluded that the wide variation in the duration of confinement in a challenged setting precluded a finding that plaintiffs established a common predominant issue. The duration of confinement in a challenged setting is DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court highly relevant to a conditions-of-confinement claim. See Hutto v. Finney, 437 U.S. 678, 686, 98 S.Ct. 2565, 57 L. Ed. 2d. 522 (1978) (“[T]he length of confinement cannot be ignored in deciding whether the confinement meets constitutional standards.”); see also Rice v. Corr. Med. Servs., 675 F.3d 650, 666 (7th Cir. 2012) (finding duration of confinement as one factor in determining whether a stay in administrative segregation constituted cruel and unusual punishment). ¶ 22 Plaintiffs contend that once individuals are placed in restrictive housing, they are subject to the same substantial risk of harm that can manifest within fifteen days of placement, and as such, they have established a common predominating issue. The trial court concluded, in its discretion, that the length of time individuals spend in restrictive housing varies across each challenged setting and impacts the nature of each plaintiff’s claim. This conclusion is supported by the record. Placement in RHAP, for instance, is initially limited to seventy-two hours and may be extended for up to fifteen days with further extension requiring approval by the Facility Classification Committee. Between October 2018 and October 2019, the average length of stay in RHAP was eight days. RHDP, alternatively, sets a maximum assignment of thirty days, and the average placement in RHDP between October 2018 and October 2019 was eleven days. In contrast, the average length of stay in RHCP and HCON between October 2018 and October 2019 was 131 days and 154 days, respectively, and placements in RHCP and HCON are reviewed less frequently. Assignments to RHCP DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court and HCON are reviewed every six months in most instances and every twelve months for individuals who assaulted and injured a staff member. Therefore, despite plaintiffs’ claim, the differences between the challenged housing classifications are relevant given that the duration of placement varies. Because duration of confinement is relevant to a conditions-of-confinement claim and because the record evidence clearly indicates significant variations in the length of stay across each challenged restrictive housing setting, the trial court did not abuse its discretion in concluding that this factor precluded a finding that plaintiffs established a common predominating issue. ¶ 23 Next, the trial court concluded that each challenged housing setting has different procedural safeguards which affect plaintiffs’ ability to establish a common predominating issue. An assessment of procedural safeguards is relevant to a conditions-of-confinement claim. See Porter, 923 F.3d at 359–63 (holding that plaintiffs were placed in solitary confinement based upon being sentenced to death but were afforded no mechanisms for removal). ¶ 24 Here the trial court’s conclusion that different procedural safeguards accompany the challenged housing settings is supported by sufficient record evidence. As the record reveals, initial placement in RHAP may be made by an officer-in-charge without conducting a prior hearing or providing an opportunity to challenge the assignment. Review by a full committee is not required unless the placement is DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court extended beyond fifteen days. Alternatively, placement in RHDP requires an investigation resulting in compilation of a disciplinary package, a prior hearing, and an opportunity to appeal. Unlike both RHAP and RHDP, assignment to RHCP is preceded by a six-step review process by two separate committees. Moreover, an HCON placement requires a hearing, multiple reviews, and approval by specifically defined staff members, while assignment to RDU is based on recent disciplinary history and eligibility factors such as age, reading level, IQ score, and close custody designation, rather than a hearing. ¶ 25 Plaintiffs fail to account for the variations in procedural safeguards, which are relevant to a conditions-of-confinement claim. Such material variations hinder plaintiffs’ ability to establish a common predominating factor. Accordingly, the trial court did not abuse its discretion in concluding that the different procedural safeguards for each restrictive housing classification precluded a finding that plaintiffs established a common predominating issue. ¶ 26 Finally, the trial court concluded that the attendant conditions of each restrictive housing setting vary significantly, are relevant to a conditions-of-confinement claim, and prevent a finding that plaintiffs established a common predominating issue. Plaintiffs argue, though, that class-wide issues predominate when a class seeks injunctive relief from shared conditions that expose all class members to the same harm, irrespective of the specific conditions of a DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court particular housing assignment and individual experiences in restrictive housing. Given several general conditions common to all forms of restrictive housing, namely the amount of time individuals spend in their cells each day and the minimal opportunity for human interaction they receive, plaintiffs contend the trial court erred by considering conditions specific to the challenged restrictive housing settings. ¶ 27 Here the trial court determined that the most significant differences among the attendant conditions occur in the frequency of visitation, the nature of recreation, and the quantity and quality of interactions with other incarcerated people. This finding is supported by the record. ¶ 28 The record shows that visitation rights vary across the challenged settings. Offenders assigned to RHAP and RHDP may receive an unlimited number of one-hour noncontact visits, while individuals in RHCP and HCON are limited to two visits every thirty days. The record also highlights differences in which individuals in restrictive housing settings can interact with other inmates, including by location, whether restrained or unrestrained, frequency, and duration. Offenders placed in RDU, for instance, may recreate in an open yard with other inmates and access the gym. Individuals placed in the other restrictive housing settings, however, are limited to outdoor recreation, and the classifications differ on whether individual or group recreation is permitted. Further, the trial court found the availability of in-cell activities to be a relevant attendant condition. The degree to which individuals in DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court restrictive housing can participate in cell study programs and other types of stimulating activities varies by housing assignment. Placement in RDU affords individuals the opportunity to complete educational courses, receive high school and college credit, and participate in short-term work assignments similar to those offered in general population. Alternatively, offenders assigned to RHAP have access to a portable library, pastoral counseling, and cell-study materials. ¶ 29 Moreover, a journal article relied upon by plaintiffs echoes the relevance of varying attendant circumstances. The article explains that variables among housing conditions, including the availability of reading material and frequency of visitation, “might explain differing outcomes.” ¶ 30 Whether there is a substantial risk of harm depends significantly on the penological purposes served, the procedural safeguards, the duration and length of stay, and the relevant attendant circumstances to each restrictive housing assignment. Thus, the fundamental distinctions and individual issues identified by the trial court are material and far from collateral. Compare Faulkenbury, 345 N.C. at 698, 483 S.E.2d at 431–32 (holding the trial court did not abuse its discretion in certifying a class, where plaintiffs’ claim for the underpayment of benefits predominated over individual, “collateral issues”), with Fisher, 369 N.C. at 215, 794 S.E.2d at 709 (concluding the trial court did not abuse its discretion in certifying a class because “the same basic questions of fact and law will determine whether” DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court plaintiffs can recover damages from defendant). Accordingly, the trial court did not abuse its discretion in concluding that no common issue predominates over issues affecting only individual class members because of the fundamental differences across the housing classifications. ¶ 31 Plaintiffs alternatively contend the trial court erred because it “failed to acknowledge that institutionalized plaintiffs may seek broad systemic relief when faced with systemic risks of harm.” To support their contention, plaintiffs claim that when “a class seeks an indivisible injunction benefitting all its members at once, there is no reason to undertake a case-specific inquiry into whether class issues predominate,” quoting Wal-Mart Stores, Inc., 564 U.S. at 362–63. As the trial court correctly concluded, however, the Supreme Court of the United States was analyzing a subsection of Rule 23 of the Federal Rules of Civil Procedure, Rule 23(b)(2), which is not included in North Carolina’s Rule 23. Further, it is well established that this Court has interpreted North Carolina’s Rule 23 to require plaintiffs seeking class certification to establish the existence of a class, which requires plaintiffs to demonstrate that each member has “an interest in either the same issue of law or of fact, and that issue predominates over issues affecting only individual class members.” Crow, 319 N.C. at 277, 354 S.E.2d at 462. ¶ 32 Plaintiffs have failed to establish that the claims of all potential class members share a common issue capable of resolution with one stroke. Beroth Oil Co., 367 N.C. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court at 346, 757 S.E.2d at 476 (holding there was no error in the trial court’s denial of class certification because although defendant’s “generalized actions may [have been] common to all [potential class members’ properties], . . .” “liability [could] be established only after extensive examination of the circumstances surrounding each of the affected properties” (internal quotations and citation omitted)). We therefore hold the trial court did not abuse its discretion in concluding plaintiffs failed to demonstrate a common predominating issue among the purported class members. ¶ 33 While the trial court identified two additional bases for denying plaintiffs’ motion for class certification—inadequacy of plaintiffs as class representatives and that litigation as a class is not a superior method of adjudication—we do not need to reach those bases here. The record evidence firmly supports the trial court’s conclusion that plaintiffs failed to establish a common predominating issue among the purported class members. Since the trial court did not abuse its discretion in determining plaintiffs failed to meet this initial requirement to class certification, review of the additional bases is not needed. ¶ 34 A trial court possesses broad discretion in class certification. Honoring that discretion is the “touchstone for appellate review” of class certification orders. See Frost, 353 N.C. at 198, 540 S.E.2d at 331. We hold that the trial court did not abuse its discretion in denying plaintiffs’ motion for class certification and affirm the trial court’s order. DEWALT V. HOOKS 2022-NCSC-105 Opinion of the Court AFFIRMED. Justice EARLS, dissenting. ¶ 35 While a trial court has discretion to determine whether to certify a class, that discretion is not completely unfettered. When the trial court erroneously requires plaintiffs to prove their case on the merits in the guise of determining a common legal issue, and where the trial court mischaracterizes the nature of the plaintiffs claims, those legal errors cannot be endorsed in the name of fidelity to the trial court’s discretion. See Beroth Oil Co. v. N.C. DOT, 367 N.C. 333, 342 (2014) (“In determining the propriety of a class action, the question is not whether the plaintiff or plaintiffs have stated a cause of action or will prevail on the merits, but rather whether the requirements of Rule 23 [class certification] are met.” (quoting Eisen v. Carlisle & Jacquelin, 417 U.S. 156, 178 (1974))); see also Blitz v. Agean, Inc., 197 N.C. App. 296, 312 (2009) (vacating a denial of class certification based on the trial court’s “misapprehension of applicable law”) (cleaned up). ¶ 36 In 2015, Justice Kennedy echoed words Dostoyevsky wrote over 150 years ago: “The degree of civilization in a society can be judged by entering its prisons.” Davis v. Ayala, 576 U.S. 257, 290 (2015) (Kennedy, J., concurring) (quoting The Yale Book of Quotations 210 (Fred R. Shapiro ed. 2006)). “There is truth to this in our own time.” Ayala, 576 U.S. at 290. “Prisoners are shut away— out of sight, out of mind.” Id. at 288. For many people in prison, this detention includes the use of solitary confinement. Plaintiffs in this case allege that in North Carolina, people in solitary confinement are forced to live for twenty-two to twenty-four hours a day in cells no bigger than a typical parking space, with little to no opportunity for meaningful DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting human contact or environmental stimulation. And it is this policy, as a whole, that Rocky Dewalt, Robert Parham, Anthony McGee, and Shawn Bonnett (plaintiffs) challenge, not only for themselves but for anyone who is or will be subjected to solitary confinement. ¶ 37 Since at least 1890, the United States Supreme Court has noted “serious objections” regarding the use of solitary confinement. In re Medley, 134 U.S. 160, 168 (1890). In In re Medley the Court noted that the adverse effects of solitary confinement occurred “after even a short confinement.” Id; see Ruiz v. Texas, 137 S. Ct. 1246 (2017) (mem.) (Breyer, J., dissenting) (citing In re Medley, 134 U.S. at 172). More recently Justice Kennedy acknowledged that “[y]ears on end of near total isolation exact a terrible price.” Ayala, 576 U.S. at 289 (Kennedy, J., concurring). Social isolation and lack of environmental stimulation are the hallmarks of solitary confinement. These practices can exacerbate pre-existing mental illnesses and cause the “appearance of an acute mental illness in individuals who had previously been free of any such illness.”1 See Stuart Grassian, Psychiatric Effects of Solitary Confinement, 22 Wash. U. J.L. & Pol’y 325, 333 (2006) (stating common side effects of solitary confinement include anxiety, panic, withdrawal, hallucinations, self- mutilation, and suicidal thoughts and behaviors). Even more significantly, the effects 1This is especially concerning given people with mental illness are more likely to be subjected to solitary confinement than those without a mental illness. DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting of solitary confinement in many cases can be permanent. In North Carolina the effects of solitary confinement are especially harrowing, with at least one study finding that people who spent any time in solitary confinement in our state prisons “were significantly more likely to die of all causes in the first year after release than those who did not.” Statistics also demonstrate that African Americans and other people of color are disproportionately represented among persons subjected to solitary confinement.2 ¶ 38 North Carolina still allows people to be placed in solitary confinement indefinitely. Plaintiffs challenge this State’s solitary confinement policy, arguing that the policy “viewed as a whole, impose[s] cruel or usual punishment forbidden by Article I, Section 27 of the state Constitution.” They seek declaratory and injunctive relief limiting the use of solitary confinement, such that it could only be used “as a last resort, and for the shortest time possible.” Because thousands of people are subjected to solitary confinement each day under the same statewide policy, there are 2 The Vera Institute reported that “while 35 percent of the white incarcerated population had spent at least one night in restrictive housing during the [year prior to the study],” the same was true for 47 percent of African American individuals. Jessa Wilcox, Léon Digard, & Elena Vanko, Vera Inst. Of Just., The Safe Alternatives to Segregation Initiative: Findings and Recommendations for the North Carolina Department of Public Safety, 22-23 (Dec. 2016). Further, people identifying as African American were overrepresented in all but one type of restrictive housing. Id. Latino men are also disproportionately impacted by solitary confinement, as they make up 16.9% of the male restrictive housing population across all evaluated jurisdictions, despite being only 15.4% of the total male custodial population. The Corr. Leaders Ass’n & The Arthur Liman Ctr. for Pub. Int. L. at Yale L. Sch. Time-In-Cell 2019: A Snapshot of Restrictive Housing based on a Nationwide Survey of U.S. Prison Systems, 26 (Sept. 2020). DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting thousands of potential class members, all of whom face nearly identical conditions. Class members challenge the same statewide practices, rely on the same legal theory, and seek uniform relief through changes to statewide policy. As plaintiffs’ brief makes clear “no one is asking for an individually tailored remedy based on unique personal circumstances.” ¶ 39 The trial court mischaracterized plaintiffs’ argument as “depend[ing] greatly on the individual class member’s experiences in the various restrictive housing settings.” In affirming the trial court’s order, the majority goes to great lengths to find irrelevant differences that do not have any legal significance. Instead of addressing plaintiffs’ argument, which requires that this State’s solitary confinement policy be “taken as a whole,” the majority engages in an analysis of the policy’s administrative classifications for solitary confinement, the varied average lengths of time each person is kept in solitary confinement, and the varied reasons a person may be subjected to such confinement, among other things. But none of these factors are relevant to a class certification motion in a case that challenges a statewide policy “as a whole.” See Brown v. Plata, 563 U.S. 493, 505 n.3 (2011) (addressing a class action challenge to a policy “taken as a whole”); see also Parsons v. Ryan, 754 F.3d 657, 678 (9th Cir. 2014) (“That inquiry does not require us to determine the effect of those policies and practices upon any individual class member (or class members) or to undertake any other kind of individualized determination.”). Because North DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting Carolina’s solitary confinement policy allows for indefinite use of solitary confinement across all classifications, these distinctions cannot, as a matter of law, weigh against plaintiffs. See Pride v. Correa, 719 F. 3d 1130, 1137 (9th Cir. 2013) (stating that individual claims for relief “are discrete from the claims for systemic reform addressed in Plata.”). ¶ 40 It matters not how well supported by the evidence the trial court’s factual findings about the various classifications of confinement may be. “What all members of the putative class and subclass have in common is their alleged exposure, as a result of specified statewide . . . policies and practices that govern the overall conditions of . . . confinement, to a substantial risk of serious future harm . . .” Parsons v. Ryan, 754 F. 3d. 657, 678 (2014). Thus, the legal significance of this detention policy for plaintiffs’ class certification motion is that plaintiffs must show that a large number of individuals are subject to the same treatment, namely, twenty-two to twenty-four hours of isolation inside a cell for an indefinite amount of time; accordingly, as a legal matter, those individuals can request the same type of relief.3 3 The majority’s analysis is like saying that in a suit challenging the constitutionality of a reduction in public employees’ disability benefits, a class action cannot be maintained because different class members receive differing payments and thus would recover different amounts. It may be true that disability benefits and recovery amounts vary, but that’s not the point. In this example, as a class, this group challenges the constitutionality of their reduction in disability benefits, and thus class certification is appropriate for class-wide relief. See Faulkenbury v. Tchrs’ & State Emps’ Ret. Sys., 345 N.C. 683, 698 (1997) (“The predominate issue is how much the parties’ retirement benefits were reduced by an unconstitutional change in the law. This issue defines the class.”). . DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting See Braggs v. Dunn, 317 F.R.D. 634, 667 (M.D. Ala. 2016) (“The Supreme Court has approved of system-wide relief in prison cases involving systemwide violation[s] resulting from systemwide deficiencies” (quoting Plata, 563 U.S. at 532 (cleaned up))). The majority also determined that because there are differences in the frequency of visitation, the nature of recreation, and the quantity and quality of human interaction, the plaintiffs could not establish a predominating issue. Yet plaintiffs’ argument is not that there aren’t differences among the different housing assignments. Those distinctions are irrelevant. See Parsons, 754 F. 3d 657, 678 (9th Cir. 2014). Instead, they argue that the actual conditions of confinement in every instance, whatever the housing arrangements, or visitation options, which dictate that a person will spend twenty-two to twenty-four hours a day in a cell, for an indefinite time, violate Article I, Section 27 of the North Carolina Constitution. See id. at 678. ¶ 41 Plaintiffs’ argument is similar to the contentions advanced in Plata v. Brown. In Plata the class was composed of state prisoners who suffered an alleged constitutional violation based on “systemwide deficiencies” in prison “medical and mental health care that, taken as a whole, subject[ed] sick and mentally ill prisoners in California to ‘substantial risk of serious harm’ and cause[d] the delivery of care in the prisons to fall below the evolving standards of decency that mark the progress of a maturing society.” 563 U.S. at 505 n.3 (quoting Farmer v. Brennan, 511 U.S. 825, DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting 834 (1994)). In Plata the Court further stated that because the plaintiffs did “not base their case on deficiencies in care provided on any one occasion, [there was] no occasion to consider . . . particular deficienc[ies] in [the] medical care complained of.” Id. Similarly, because plaintiffs in this case do not allege a constitutional violation based on particular deficiencies but rather make allegations related to North Carolina’s policy “as a whole,” the majority’s analysis of differences in the ways in which different types of restrictive housing implement the policy is misplaced. Specifically, the majority’s recitation of variations in implementation of the policy’s administrative classifications for solitary confinement, the varied average length of time a person is kept in solitary confinement, the varied reasons a person may be subjected to such confinement, frequency of visitation, the nature of recreation and the quantity and quality of human interaction is irrelevant to the determination before us now. See id.; see also Parsons, 754 F. 3d at 678. ¶ 42 Furthermore, the trial court improperly assessed the merits of plaintiffs’ claims when it found there was not enough evidence to show the “Department’s [solitary confinement] policies and practices actually caused the complained of harm[.]” Addressing the merits of the plaintiffs’ case not only bypasses the process of discovery and trial but is also legal error. In North Carolina, Rule 23 does not ask whether the plaintiff will prevail on the merits and any inquiry into the merits of a case should be limited to the issue of class certification. Beroth Oil, 367 N.C. at 342, DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting 342 n. 5 At this stage plaintiffs are only required to show that North Carolina’s statewide solitary confinement policy and practice exposes class members to a common risk of harm, not whether this exposure occurred or rises to the level of a constitutional violation. See Beroth Oil, 367 N.C. 333 at 342. The evidence submitted by the plaintiffs meets this burden because at this stage all they seek to establish is that a group of people within North Carolina prisons may be exposed to a risk of harm because they spend twenty-two to twenty-four hours a day inside a cell. ¶ 43 In making its determination, the trial court considered two of the four reports submitted by the plaintiffs. One report detailed the increased risk during community reentry following the use of solitary confinement. The trial court and majority conclude alike that because the study involved observational data and correlational analysis, it could not “provide concrete support” for plaintiffs’ claim that solitary confinement increases the risk of post-release mortality. However, this does not address class certification under Rule 23, see Beroth Oil, 367 N.C. at 342 n. 5, and instead the trial court and majority’s reasoning addresses the central question in this case, namely whether defendants have in fact imposed a class wide policy that causes a substantial risk of serious harm. Yet at this point in the litigation, there is only one discreet question– whether class certification is met under Rule 23. See id. ¶ 44 Regarding the second study, the Vera Report, the majority recounts the trial court’s findings stating the report was “insufficient to connect DPS’s practices to the DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting alleged risk of harm.” In doing so, the majority notes that the Vera Report “commended DPS on its previous reform efforts, suggested that DPS continue implementation of both current and future reform,” and noted that DPS’s restrictive housing population had decreased by 10% one month after the study concluded.” However, this line of reasoning speaks to the merits of the plaintiffs alleged constitutional violation and is more properly addressed at a later stage in the litigation. See Beroth Oil, 367 N.C. at 342 n. 5. Thus, because at this stage plaintiffs only seek to establish a class of persons subjected to solitary confinement for twenty- two to twenty-four hours a day, their burden has been met. ¶ 45 Furthermore, although the majority does not reach this issue, the trial court found that the named representatives would not fairly and adequately represent the interests of all class members because (1) the plaintiffs do not represent the “wide spectrum of inmates potentially encompassed in the class,” and (2) “their own actions may compromise the viability of their own claims.” This conclusion was based upon the named plaintiffs being “placed in restrictive housing early in their sentence” and “being repeatedly assigned to restrictive housing or having had their assignment extended” due to “repeated disciplinary infractions.” However, the class is not based on the individual actions or circumstances of each plaintiff, instead it is based on a solitary confinement policy that subjects people to twenty-two to twenty-four hours a day in a cell, for an unlimited number of days. Thus, plaintiffs being placed in solitary DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting confinement early in their sentences, or the reason they were placed there or had their time there extended has no legal relevance. ¶ 46 The trial court also found that a class action was not superior to other available methods of adjudication because litigation would “devolve into a series of mini trials” about “each of the challenged restrictive housing assignments” and “the myriad of other relevant considerations and defenses that undoubtedly would not apply uniformly to all potential class members.” Here again the trial court mischaracterized plaintiffs’ arguments. Because plaintiffs challenge the policy as a whole there is no occasion to consider the individual circumstances of each plaintiff. See Plata v. Brown, 563 U.S. 493, 505 n.3 (2011). Instead, what is important is that the class is composed of people who spend twenty-two to twenty-four hours a day in a cell in social isolation. ¶ 47 Lastly, in upholding the trial court’s order, the majority repeatedly states that a trial court has broad discretion to decide whether to certify a class. Although it is true that under this Court’s precedent in Crow v. Citicorp Acceptance Co., Inc., 319 N.C. 274 (1987), a “trial court has broad discretion,” this discretion relates to balancing “[t]he usefulness of the class action device. . . against inefficiency or other drawbacks.” Id. at 284. Assessing the extent to which evidence proffered on the class certification motion proves that plaintiffs have suffered a violation of their constitutional rights is a legal error and does nothing to contemplate the required DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting balance. Instead, it evidences hostility to their claim on the merits, which is not the appropriate assessment at this point in the litigation. In other words, it is an abuse of discretion for the trial court to deny class certification on the grounds that the plaintiffs should lose on the merits. See Beroth Oil, 367 N.C. at 342. ¶ 48 The class here is not based on the individual circumstances of each plaintiff, instead it is based on a solitary confinement policy that subjects people to twenty-two to twenty-four hours a day in a cell, for an unlimited number of days. Like in Faulkenbury, 345 N.C. at 698, plaintiffs all seek the same type of relief, namely an injunction and declaratory judgment that the state constitutional guarantees mean that solitary confinement be used only as a last resort and for the shortest time necessary. See id. (“Each of the parties had a claim based on what he or she contends is underpayment of retirement benefits. This claim predominates over issues affecting only one individual class member. This establishes a class.”). Likewise, class certification is not based on an assessment of the plaintiffs’ allegations on the merits. See Beroth Oil, 367 N.C. at 342 n. 5. Thus, whether plaintiffs provided correlational or observational evidence cannot be relevant to this inquiry because all that is necessary to establish the grounds for class certification is that there is a group of people alleged to be exposed to the same treatment of little to no social interaction or environmental stimulation for twenty-two to twenty-four hours a day inside a cell. ¶ 49 “[C]onsideration of these issues is needed” and “[t]here are indications of a new DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting and growing awareness . . . of solitary confinement.” Ayala, 576 U.S. at 289 (Kennedy, J., concurring). Even years ago, it was “evident that some changes must be made in the system.” In re Medley, 134 U.S. 160, 168 (1890). As a result of the “terrible human toll” resulting from solitary confinement, Ruiz 137 S. Ct. at 1247 (Breyer J. dissenting), it has been suggested that if a case presents an issue of solitary confinement, “the judiciary may be required, within its proper jurisdiction and authority, to determine whether workable alternative systems for long term confinement exist, and, if so, whether a correctional system should be required to adopt them.” Ayala, 576 U.S. at 290 (Kennedy, J., concurring). Today this Court has the responsibility to apply the criteria for class certification to the claim that is actually being brought by plaintiffs, not to the claim as chopped up and reconstituted by defendants and the majority. ¶ 50 Plaintiffs are asking this Court to recognize that, as a group, they have state constitutional rights that are implicated by North Carolina’s solitary confinement practices. Those rights are equally violated by the whole policy, without regard to whether detainees are in RHAP, RHDP, HCON, or some other acronym for the same thing–solitary confinement in a single cell for twenty-two to twenty-four hours a day for an indefinite number of days. The majority essentially holds that because it does not agree with the constitutional claims on the merits, class certification is not appropriate. But our system of laws has long recognized the importance of the class DEWALT V. HOOKS 2022-NCSC-105 Earls, J., dissenting action vehicle for the resolution of disputes in which large numbers of individuals share a common claim and would all benefit from a common resolution.4 ¶ 51 Because plaintiffs challenge a widespread state policy and seek to establish a class of individuals who are subject to the same policy allowing for twenty-two to twenty- four hours inside a prison cell for an indefinite period, I would hold that the trial court based its ruling on a misapprehension of plaintiffs’ claim and a mistake of law. I would reverse the trial court’s order denying class certification, and remand the matter for further proceedings applying the correct understanding of class certification in these circumstances. Accordingly, I respectfully dissent. Justice HUDSON joins in this dissenting opinion. 4The English bill of peace, which originated in the middle ages to facilitate the adjudication of disputes involving common questions and multiple parties in a single action, was the basis for North Carolina’s early class action decisions in the late 1800s. See Chambers v. Moses H. Cone Mem'l Hosp., 374 N.C. 436, 440 (2020) (citing Bronson v. Wilmington N.C. Life Ins. Co., 85 N.C. 411, 414 (1881) (acknowledging the class action mechanism as a feature of civil procedure)).