Luther Poynter v. Aaron Shirley

READLER, Circuit Judge. Two inmates at Barren County’s detention center attacked Luther Poynter, a civil contemnor detained in the same facility. The incident left Poynter with permanent injuries. He later sued the County’s Jailer as well as the County itself, alleging that the jail failed to protect him against the attack in violation of the substantive component of the Fourteenth Amendment’s Due Process Clause.

Poynter’s failure-to-protect claim is one example of a broader set of due-process theories that a detained person like Poynter, who has not been convicted of a crime, may assert where the detainee believes his jailer has been deliberately indifferent to the conditions of his confinement. See Lawler ex rel. Lawler v. Hardeman County, 93 F.4th 919, 926 (6th Cir. 2024) (collecting types of deliberate-indifference claims). For nearly 30 years, the standard announced in Farmer v. Brennan, 511 U.S. 825 (1994), guided our analysis in those cases. Essential to a detainee’s deliberate-indifference claim under Farmer is that the jailer, at the very least, subjectively knew that his actions risked serious harm to the detainee.

Five years ago, we changed course. In Brawner v. Scott County, 14 F.4th 585 (6th Cir. 2021), we read Kingsley v. Hendrickson, 576 U.S. 389 (2015), a decision adopting a new legal standard for a pretrial detainee’s excessive-force claim, as also requiring a different approach in confinement cases pursued under a deliberate-indifference theory. Over time, Brawner has been understood to replace Farmer’s subjective inquiry with an objective one.

The change in doctrine announced in Brawner spawned a host of related developments in our Circuit. One was Westmoreland v. Butler County, 29 F.4th 721 (6th Cir. 2022), which relied on Brawner to craft a new, four-part framework for evaluating failure-to-protect claims like the one Poynter raises. Both the district court and a panel of our Court applied the Westmoreland test to Poynter’s claim. In reaching that conclusion, the panel deemed an earlier interpretation of Westmoreland no longer binding on the Court, emblematic of a larger debate over our deliberate- indifference jurisprudence. To clarify the matter, we granted rehearing en banc to consider whether Farmer or Kingsley should govern Fourteenth Amendment deliberate-indifference claims. We conclude that Farmer’s subjective standard continues to control claims brought by a detainee who has not been convicted of a crime. Because the district court employed Westmoreland to evaluate defendants’ summary-judgment motion, we vacate its order and remand to allow the district court to consider in the first instance whether Poynter’s claim is viable under Farmer.

I.

We begin with the somewhat lengthy history underlying our deliberate-indifference jurisprudence with respect to detained individuals.

A. While today’s case interprets the Fourteenth Amendment, the doctrine it explores originated in the Eighth Amendment’s prohibition on the infliction of cruel and unusual punishments. U.S. CONST. amend. VIII. For centuries, that provision was understood to prohibit “torturous punishments meted out by statutes or sentencing judges.” See Hudson v. McMillian, 503 U.S. 1, 18–19 (1992) (Thomas, J., dissenting) (discussing caselaw and historical sources). In Estelle v. Gamble, 429 U.S. 97 (1976), the Supreme Court extended the Amendment’s reach to hardships that might befall a prisoner while incarcerated. Among those hardships was the jailer’s lack of attention to a prisoner’s serious medical needs. Estelle, 429 U.S. at 104. Estelle reasoned that a jailer’s knowing inaction in the medical setting—what the Supreme Court described as his “deliberate indifference”—would “constitute[] the ‘unnecessary and wanton infliction of pain’” that the Eighth Amendment prohibits. Id. at 103–04 (quoting Gregg v. Georgia, 428 U.S. 153, 173 (1976) (plurality opinion)). In enshrining “deliberate indifference” as the benchmark in this area, Estelle emphasized the critical nature a jailer’s mental state plays in assessing liability: A jailer’s negligence alone “in diagnosing or treating a medical condition,” the Supreme Court made clear, is insufficient to state a constitutional violation. Id. at 106–07 (describing negligence as the province of state tort law).

Over the next 25 years, the Supreme Court refined its deliberate-indifference framework. Two decisions in particular deserve mention. One, Rhodes v. Chapman, 452 U.S. 337 (1981), extended Estelle’s logic to all conditions of confinement in prison, not just medical care. See Rhodes, 452 U.S. at 347. In so doing, Rhodes explained that, to run afoul of constitutional protections, a challenged condition must be objectively serious enough to “result in pain without any penological purpose.” Id. (citing Estelle, 429 U.S. at 103). The other is Wilson v. Seiter, 501 U.S. 294 (1991). Taking account of its past precedent, the Supreme Court in Wilson announced that conditions-of-confinement claims employ a two-part standard. The first inquiry asks whether the conduct at issue involved sufficiently serious harm (or risk thereof). Wilson, 501 U.S. at 298, 303. The second, a state-of-mind inquiry, asks whether the jailer’s conduct amounted to deliberate indifference. Id.

Having adopted a two-prong standard, the Supreme Court then turned to defining more precisely what satisfies the second (or deliberate-indifference) prong. In the years following Estelle, a circuit split developed over whether a jailer may be held constitutionally liable under a conditions-of-confinement theory for an incarcerated person’s injury where the jailer lacked subjective awareness of the potential harm, but nonetheless “should have known of a sufficiently serious danger to an inmate.” Farmer, 511 U.S. at 832 (quoting Young v. Quinlan, 960 F.2d 351, 360–61 (3d Cir. 1992)). In Farmer v. Brennan, the Supreme Court answered “no” to that question. With the Eighth Amendment’s focus on “punishments” and not “conditions,” the Supreme Court explained, “an official’s failure to alleviate a significant risk that he should have perceived but did not . . . cannot . . . be condemned as the infliction of punishment.” Id. at 837– 38. Demonstrating deliberate indifference, it followed, “requir[es] a showing that the [jailer] was subjectively aware of the risk.” Id. at 829; see also id. at 837 (“[T]he official must both be aware of facts from which the inference could be drawn that a substantial risk of serious harm exists, and he must also draw the inference.”).

Farmer was an Eighth Amendment case. And the Eighth Amendment applies only to prisoners who have been convicted of a crime. See Lovell v. County of Kalamazoo, 172 F.4th 931, 936 (6th Cir. 2026) (citing U.S. CONST. amend. VIII; Whitley v. Albers, 475 U.S. 312, 318 (1986)). Where does that leave “an individual within ‘some gray area’ between free citizen and convicted criminal,” such as a pretrial detainee? Hopper v. Plummer, 887 F.3d 744, 754 (6th Cir. 2018) (quoting Burgess v. Fischer, 735 F.3d 462, 472 (6th Cir. 2013)). Filling that gap is the substantive component of the Fourteenth Amendment’s Due Process Clause, see U.S. CONST. amend. XIV, § 1, which the Supreme Court understands to prohibit unlawful conditions of confinement, Bell v. Wolfish, 441 U.S. 520, 535 (1979) (citing Ingraham v. Wright, 430 U.S. 651, 671–72 n.40, 674 (1977); Kennedy v. Mendoza-Martinez, 372 U.S. 144, 165–67, 186 (1963); Wong Wing v. United States, 163 U.S. 228, 237 (1896)). In adopting substantive due process as the constitutional foundation for conditions claims, the Supreme Court tied those claims to the familiar language of the Eighth Amendment, with the crucial inquiry remaining whether those conditions are “imposed for the purpose of punishment.” Id. at 538 (citing Flemming v. Nestor, 363 U.S. 603, 613–17 (1960)); see also id. at 535 n.16. So although a claimant invoking the Due Process Clause by definition has not been convicted of a crime (as has a prisoner), that difference “has no application to a determination of the [detainee’s] rights . . . during confinement.” Id. at 533.

Heeding these words, we applied the Farmer deliberate-indifference standard for the next quarter century to pretrial detainees’ claims challenging the conditions of their confinement. See Griffith v. Franklin County, 975 F.3d 554, 569 (6th Cir. 2020) (citing Roberts v. City of Troy, 773 F.2d 720, 724–25 (6th Cir. 1985)). We were not alone in that understanding. Every other circuit, we note, did the same. See Brief of Indiana, Hawaii, and Texas as Amici Curiae in Support of Petitioners at 9, County of Orange v. Gordon, 586 U.S. 1069 (2019) (mem.) (No. 18- 337) (collecting cases where circuits applied Farmer’s framework to pretrial detainees pre- Kingsley).

B. In addition to governing the conditions of confinement, the Eighth Amendment has also been interpreted to prohibit a jailer’s use of excessive force. See Whitley, 475 U.S. at 320 (contrasting conditions-of-confinement and excessive-force claims and explaining that the elements of the respective claims must reflect “due regard for differences in the kind of conduct against which an Eighth Amendment objection is lodged”). Broadly speaking, excessive force is reflected by “physically abusive governmental conduct.” Graham v. Connor, 490 U.S. 386, 394 (1989). For Eighth Amendment claims alleging that a jailer employed excessive force, which typically involve an affirmative act “made in haste” and “under pressure,” the state of mind required differs from the deliberate-indifference standard. Whitley, 475 U.S. at 320. To be liable for excessive force, the jailer must act “maliciously and sadistically for the very purpose of causing harm.” Id. at 320–21 (quoting Johnson v. Glick, 481 F.2d 1028, 1033 (2d Cir. 1973)).

In 2015, the Supreme Court addressed the elements for proving excessive force when the plaintiff is a detainee not subject to the Eighth Amendment and must instead rely on the Fourteenth Amendment’s due-process guarantee. Kingsley, 576 U.S. at 391–92. “[T]he appropriate standard” with respect to the amount of force used against that class of detainees, the Supreme Court explained, “is solely an objective one.” Id. at 397. That said, the detainee must also demonstrate that the officer used force deliberately, even if the officer did not understand that the amount of force was excessive. Id. at 395–96. So “if an officer’s Taser goes off by accident or if an officer unintentionally trips and falls on a detainee,” the officer’s negligence would not amount to excessive force in violation of the Due Process Clause. Id. at 396 (citing County of Sacramento v. Lewis, 523 U.S. 833, 849 (1998)). This view aligned with the broader constitutional understanding that purely accidental or unconscious conduct cannot “shock[] the conscience” to amount to a substantive-due-process violation. Lewis, 523 U.S. at 846; see Daniels v. Williams, 474 U.S. 327, 330–34 (1986).

C. Following Kingsley, detainees who had not been convicted of a crime began asking federal courts to modify their deliberate-indifference framework for conditions-of-confinement claims in light of Kingsley’s refinement of what constitutes excessive force. Several circuits rejected Kingsley’s relevance in the distinct deliberate-indifference setting. See Cope v. Cogdill, 3 F.4th 198, 207 n.7 (5th Cir. 2021); Whitney v. City of St. Louis, 887 F.3d 857, 860 n.4 (8th Cir. 2018); Strain v. Regalado, 977 F.3d 984, 989–93 (10th Cir. 2020); Dang ex rel. Dang v. Sheriff, Seminole Cnty., 871 F.3d 1272, 1279 n.2 (11th Cir. 2017). Not all courts, however, took the same approach. The en banc Ninth Circuit viewed Kingsley as a watershed case that overturned decades of deliberate-indifference jurisprudence and required replacing Farmer with an objective-only inquiry for pretrial detainees’ conditions-of-confinement claims. See Castro v. County of Los Angeles, 833 F.3d 1060, 1069–71 (9th Cir. 2016) (en banc), vacating 797 F.3d 654 (9th Cir. 2015). The Second and Seventh Circuits followed suit, adopting Kingsley in differing forms with respect to deliberate-indifference claims by pretrial detainees, see Darnell v. Pineiro, 849 F.3d 17, 34–35 (2d Cir. 2017); Miranda v. County of Lake, 900 F.3d 335, 352 (7th Cir. 2018), as did the Fourth Circuit several years later, see Short v. Hartman, 87 F.4th 593, 605– 06 (4th Cir. 2023).

As for our Circuit, in 2021, we issued our decision in Brawner v. Scott County. See 14 F.4th at 596. Brawner understood Kingsley’s excessive-force framework to apply also to pretrial detainees’ medical-care claims. Id. at 596. “Kingsley’s clear delineation between claims brought by convicted prisoners under the Eighth Amendment and claims brought by pretrial detainees under the Fourteenth Amendment,” Brawner concluded, required “modification of” Farmer’s deliberate-indifference standard to incorporate objective considerations. Id. To implement that modification, Brawner crafted a new, two-part test governing Fourteenth Amendment indifference-to-medical-needs claims. First, the detainee must “present evidence from which a reasonable jury could find” an “objectively serious medical need.” Id. at 597. And second, the detainee must show that her jailer’s “action (or lack of action) was intentional (not accidental)” and the jailer either “acted intentionally” or “recklessly failed to act reasonably to mitigate the risk the serious medical need posed to [the detainee], even though a reasonable official . . . would have known that the serious medical need posed an excessive risk.” Id.

Our post-Brawner jurisprudence has not followed a straight line. Over five years, Brawner has produced dueling unanimous decisions reflective of our uncertainty over Brawner’s meaning. We have debated how to understand Brawner’s “modification” of the subjective prong. 14 F.4th at 596. Compare Trozzi v. Lake County, 29 F.4th 745, 755–58 (6th Cir. 2022), with Helphenstine v. Lewis County, 60 F.4th 305, 316 (6th Cir. 2023). We have similarly debated Brawner’s role in failure-to-protect deliberate-indifference cases. See Westmoreland, 29 F.4th at 729. Compare Buetenmiller v. Macomb Cnty. Jail, 53 F.4th 939, 945–46 (6th Cir. 2022), with Poynter ex rel. Fernandez v. Bennett, 162 F.4th 664, 678 (6th Cir. 2025), vacated, 169 F.4th 716 (6th Cir. 2026) (en banc) (mem.). All this has left our district-court colleagues with the unenviable job of attempting to reconcile these mixed signals. See, e.g., Campbell v. Riahi, No. 20-cv-678, 2023 WL 5979211, at *4 (S.D. Ohio Sep. 13, 2023) (“Since Brawner . . . changed the traditional test for deliberate indifference claims, district courts have faced conflicting guidance from the Sixth Circuit about the appropriate analytical framework. . . . Absent clarification, the safest course is perhaps to analyze such claims under both standards, with fingers crossed that they agree.” (footnote omitted)); Montiel v. Johnson, No. 25-cv-122, 2025 WL 807756, at *9 (W.D. Mich. Mar. 14, 2025) (similar); Bonham v. Fam. Outreach Ctr., Inc., No. 24-cv-1030, 2024 WL 4562685, at *9 (W.D. Mich. Oct. 24, 2024) (similar); Preston v. County of Macomb, No. 18-12158, 2022 WL 4590573, at *11, *15 (E.D. Mich. Sep. 29, 2022) (similar).

In fairness, we are not the only circuit that, having extended Kingsley beyond the excessive-force setting, has struggled to apply Kingsley’s reasoning to conditions-of-confinement claims. As we previously acknowledged, “our sister circuits” too “are all over the map on this issue.” Helphenstine, 60 F.4th at 316 (collecting different approaches). An illustrative example is the Seventh Circuit. In Miranda v. County of Lake, the Seventh Circuit held that Kingsley applied to pretrial detainees’ deliberate-indifference conditions-of-confinement claims. 900 F.3d at 352. Then came Pittman ex rel. Hamilton v. County of Madison (Pittman III), 970 F.3d 823 (7th Cir. 2020). There, then-Judge Barrett, writing for a unanimous panel, interpreted the post- Kingsley and Miranda framework for conditions claims as focusing on whether the jailers “‘were aware’ that their actions would be harmful” to a pretrial detainee or at least “‘strongly suspected’ that their actions would lead to harmful results.” 970 F.3d at 828. A later panel in the Seventh Circuit, however, found Judge Barrett’s test incompatible with what that panel understood to be Miranda’s Kingsley-inspired imposition of an objective-only test for conditions claims. See Pittman ex rel. Hamilton v. Madison County (Pittman IV), 108 F.4th 561, 570–71 (7th Cir. 2024). In resolving this purported tension, the panel recognized that “[t]he circumstance before us is one of our own making, as we (like many other courts) have struggled to implement Kingsley’s standards outside the context of a pretrial detainee’s claim of excessive force.” Id. at 569. The Seventh Circuit’s experience, in many ways, mirrors our own.

II.

With this doctrinal background in mind, we turn to the facts in this case. We recount them only briefly because our primary concern is with the appropriate legal standard, not its application to the facts of this case.

Luther Poynter entered Barren County’s detention center following his arrest for civil contempt of court. During his intake process, Poynter completed a risk-assessment form asking whether there was anyone in the detention center from whom he needed to be kept away. He did not list any names.

Poynter was eventually placed in a cell with Timothy Guess and Scotty Wix, among others. Video footage captured what happened next. Almost immediately after entering the cell, Poynter walked over to Wix while Guess walked around the cell. After a minute of conversation, Guess and Wix attacked Poynter, striking him multiple times in the head, causing Poynter to suffer a traumatic brain injury.

Poynter’s guardian sued Barren County, as well as Aaron Bennett, in his official capacity as Jailer, under 42 U.S.C. § 1983. A few housekeeping points regarding Poynter’s complaint bear mentioning at the start. One, Bennett is no longer the Jailer, so his successor, Aaron Shirley, has been automatically substituted as the named defendant. Fed. R. App. P. 43(c)(2). Two, all agree that the suit against Shirley in his official capacity functions as a suit against Barren County. And three, Poynter’s claims are premised solely upon municipal liability under Monell v. Department of Social Services, 436 U.S. 658 (1978). We thus refer to defendants as Barren County because no one contests that liability turns on the County’s policies and customs.

According to the amended complaint, by allowing Guess and Wix to stay in the general population, Barren County was deliberately indifferent to Poynter’s safety. To prevail on his Monell theory, Poynter must establish an underlying constitutional violation as well as a municipal “custom” that both “is so widespread as to have the force of law” and caused the violation of his constitutional rights. Bd. of the Cnty. Comm’rs v. Brown, 520 U.S. 397, 404 (1997) (citing Monell, 436 U.S. at 690–91). The custom Poynter highlights is Barren County’s practice of making housing decisions for its detention facilities without taking into account the inmates’ histories of violence in those facilities.

By way of background, under Kentucky state law, jails like Barren County’s detention center that house state prisoners as well as other detainees must develop a classification system for all individuals they hold. 501 Ky. Admin. Regs. 3:110, § 1(1). Among the considerations involved in determining the proper classification for one housed at the detention center are his “tendency to harm others” as well as his “[i]nstitutional behavioral history.” Id. § 1(2)(e), (3)(b). Barren County’s articulated classification policy met the requirements of state law. Nonetheless, Poynter alleges that the County had a custom of “ignor[ing] prior institutional violence in assigning housing” and “bypass[ing] the classification analysis set forth in state regulations and its own policy manual” for those arriving at the County’s detention center. Plaintiff-Appellant’s Suppl. En Banc Br. at 2–3. Thus, in practice, if a prisoner with the Kentucky Department of Corrections was placed in Barren County’s detention center—which, again, also housed local pretrial detainees—the County allegedly would not perform its own classification, nor would it update the prisoner’s classification based on the prisoner’s behavior at the detention center.

Guess’s and Wix’s histories illustrate the gist of this claim. At the time they attacked Poynter, both were state inmates housed at the detention center. According to Poynter, both Guess and Wix had histories of violence in prison dating back over a decade, which made it unreasonable for Barren County to house them with the general population.

Poynter frames his constitutional claim as well as his Monell argument as resting on an objective-reasonableness-only understanding of Brawner and Westmoreland. To his mind, he was at a serious risk of harm from Guess and Wix, making it objectively unreasonable to house him with the two inmates. For Monell purposes, Poynter acknowledges that to succeed on his inaction theory of municipal liability, he must show a “clear and persistent pattern of unconstitutional conduct” in Barren County. See Franklin v. Franklin County, 115 F.4th 461, 472 (6th Cir. 2024) (citation modified). To do so, he does not attempt to demonstrate a pattern of Farmer violations tied to Guess and Wix—for example, that Barren County jailers subjectively ignored instances where the two prisoners committed serious harm to another inmate. Instead, Poynter’s Monell claim rests exclusively on the premise that there was a pattern of objectively unreasonable behavior at the detention center with regard to Guess and Wix.

Applying Westmoreland, the district court granted summary judgment to defendants. A panel of this Court reversed. Poynter, 162 F.4th at 685. The panel applied the objective- reasonableness-only standard as to the underlying constitutional violation, id. at 680, as well as to the application of Monell liability, id. at 682.

The en banc Court vacated that opinion. Poynter, 169 F.4th at 716. Because liability hinged on the proper understanding of deliberate indifference and the Fourteenth Amendment, we ordered supplemental briefing on what standard we should apply to conditions-of- confinement cases: Farmer, Kingsley, or something else.

III.

Having considered the parties’ thoughtful arguments, we hold that Farmer’s deliberate- indifference standard remains the proper measure for evaluating conditions-of-confinement claims, including the one pursued by Poynter.

A. As Poynter proceeds under the Due Process Clause, we start, as one should, with the provision’s text. The Due Process Clause prohibits a state from “depriv[ing]” an individual of “liberty” without “due process.” U.S. CONST. amend. XIV, § 1. (The Fifth Amendment, we note, creates the same protection against the federal government. Id. amend. V; see also Fraihat v. U.S. Immigr. & Customs Enf’t, 16 F.4th 613, 647 (9th Cir. 2021) (explaining that the Fifth Amendment’s Due Process Clause provides analogous protections for those detained by the federal government).) With respect to its substantive guarantees, the Clause is understood to protect against “certain government actions regardless of the fairness of the procedures used to implement them,” leaving the word “deprive” as the key textual hook for our consideration. See Daniels, 474 U.S. at 331. The word “deprive” signals a “deliberate decision[]” by a government official. Id. (emphasis omitted); see Parratt v. Taylor, 451 U.S. 527, 548 (1981) (Powell, J., concurring in the result) (“A ‘deprivation’ connotes an intentional act denying something to someone, or, at the very least, a deliberate decision not to act to prevent a loss.”). This understanding aligns with the Clause’s role in protecting against governmental abuses of power. DeShaney v. Winnebago Cnty. Dep’t of Soc. Servs., 489 U.S. 189, 196 (1989); Daniels, 474 U.S. at 331 (noting the Due Process Clause “serves to prevent governmental power from being ‘used for purposes of oppression’” (quoting Murray’s Lessee v. Hoboken Land & Improvement Co., 59 U.S. (18 How.) 272, 277 (1856))). Respecting the “deliberate” component of the term “deprive,” a state thus customarily does not run afoul of due process where one’s “injury has been caused not by a state law, policy, or procedure, but by a random and unauthorized act that can be remedied by state law.” Albright v. Oliver, 510 U.S. 266, 285 (1994) (Kennedy, J., concurring in the judgment) (citing Parratt, 451 U.S. at 536); see also Daniels, 474 U.S. at 332 (distinguishing between an “abuse of power” and a “lack of due care”); Doe v. Jackson Loc. Sch. Dist. Bd. of Educ., 954 F.3d 925, 931 (6th Cir. 2020) (explaining that a state’s mere failure to protect one from private violence, with nothing more, is not a deprivation of liberty).

Against this textual backdrop, how do we determine when a state’s treatment of a detainee rises to the level of a deprivation of his liberty? On this point, we have the benefit of a host of instructive decisions from the Supreme Court, most notably Bell and Farmer, and we have nearly three decades’ worth of our own decisions applying those holdings. See, e.g., Lawler, 93 F.4th at 927 (collecting cases). From this collection of precedent, we know that the Due Process Clause prohibits punishment of those who have not been convicted of a crime— pretrial detainees, civil contemnors, immigration detainees, juvenile detainees, and the like. See Bell, 441 U.S. at 535 (citing Ingraham, 430 U.S. at 671–72 n.40; Kennedy, 372 U.S. at 165–67, 186; Wong Wing, 163 U.S. at 237). Yet we also know that not every risk or injury experienced by a detainee amounts to punishment for due-process purposes. “[A]n official’s failure to alleviate a significant risk that he should have perceived but did not . . . cannot . . . be condemned as the infliction of punishment.” Farmer, 511 U.S. at 838. Rather, before constitutional liability may arise, there must be “an intent to punish” by the jailer. Roberts, 773 F.2d at 725 (emphasis added) (citing Bell, 441 U.S. 520).

Our post-Farmer precedent sensibly employed this subjective inquiry as the benchmark for measuring when punishment occurs for purposes of a Fourteenth Amendment conditions-of- confinement claim. See Griffith, 975 F.3d at 569. To assess whether a pretrial detainee had been punished by the state, we asked the following: Was there a “substantial risk of serious harm” the state knew about and disregarded? See Farmer, 511 U.S. at 837–38; see also Villegas v. Metro. Gov’t of Nash. & Davidson Cnty., 709 F.3d 563, 568–69 (6th Cir. 2013). If not, the state has not been deliberately indifferent to the detainee’s conditions of confinement.

Farmer’s subjective-knowledge requirement plays an important gatekeeping role in this substantive-due-process setting. The subjective inquiry typically serves to capture the line between constitutional liability and normal jail administration. See Comstock v. McCrary, 273 F.3d 693, 703 (6th Cir. 2001) (recognizing that Farmer’s subjective standard “is meant to prevent the constitutionalization of medical malpractice claims”). In practice, it acts as a backstop against “the natural human tendency to confuse what [the Fourteenth] Amendment protects with our own ardent views about the liberty that Americans should enjoy” when interpreting the Amendment’s relatively open-ended language. Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228, 2247 (2022). Absent a subjective element, all that remains is a nebulous reasonableness test deployed by unelected judges to govern the administratively complex and inherently dangerous jail setting. See Farmer, 511 U.S. at 837–38 (distinguishing between that which “might well be something society wishes to discourage” through “tort liability” and that which can constitutionally “be condemned as the infliction of punishment”); see also Bell, 441 U.S. at 562 (cautioning courts against becoming “enmeshed in the minutiae of prison operations”). As we were recently reminded, a constitutional test that descends into a “vague” “reasonableness” principle is “neither limiting nor much of a principle.” Cf. Trump v. Slaughter, 146 S. Ct. 2283, 2310 (2026).

The proof of Farmer’s necessity is in the Brawner pudding. As a practical matter, cases interpreting Brawner have discarded a subjective inquiry in favor of an objective- reasonableness-only approach that lowers the threshold of constitutional liability perilously close to negligence—the very thing all of us otherwise seem to agree is insufficient for constitutional liability. Brawner, 14 F.4th at 596; Kingsley, 576 U.S. at 396. To see why, consider how the objective-reasonableness-only framework functions in practice. That framing requires an “objectively serious” risk to the detained individual as well as unreasonable conduct by the state in response to the risk. See Helphenstine, 60 F.4th at 315–16; Hehrer v. County of Clinton, 161 F.4th 955, 964 (6th Cir. 2025). The first requirement largely tracks Farmer, which itself requires that the deprivation by a jailer’s act or omission result in or pose a substantial risk of a “sufficiently serious” harm to the detained individual. Farmer, 511 U.S. at 834 (quoting Wilson, 501 U.S. at 298). But as the dangers intrinsic to the jail setting mean that the objective prong is somewhat easy to satisfy, the cases interpreting Brawner’s second requirement do the bar- lowering work. Id. at 858–59 (Thomas, J., concurring in the judgment) (explaining that danger or at least a risk of danger in jails is an inherent feature). In effect, an objective-reasonableness- only scheme shifts the inquiry from one that focuses on a jailer’s subjective state of mind as to a particular risk to what a reasonable official should have done in light of that risk. See Grote v. Kenton County, 85 F.4th 397, 405 (6th Cir. 2023); see also Castro, 833 F.3d at 1069–71 (understanding Kingsley as creating liability where the defendant fails to “take reasonable available measures” to limit a reasonably apparent “substantial risk of serious harm”). In so doing, it embraces the should-have-known standard Farmer expressly rejected, 511 U.S. at 838– 39, something indistinguishable from ordinary negligence, Doe, 954 F.3d at 932; see also Farmer, 511 U.S. at 836 n.4 (discussing how gross negligence “mean[s] little different from [civil] recklessness” (citing W. Page Keeton et al., Prosser and Keeton on the Law of Torts § 34 (5th ed. 1984))); Restatement (Third) of Torts: Liability for Physical and Emotional Harm § 3 & cmt. a (A.L.I. 2010) (defining “negligence” as when a “person does not exercise reasonable care under all the circumstances” and explaining that “reasonable care” is “conduct that avoids creating an ‘unreasonable risk of harm’”).

In addition to honoring long-standing due-process doctrine, Farmer’s approach also furthers the federalism principles underlying our broader constitutional order by respecting the role of state and local prison officials. See Brief of the Commonwealth of Kentucky, State of Michigan, State of Ohio, and State of Tennessee as Amici Curiae in Support of Appellees at 2 (supporting a return to Farmer). State tort law, of course, remains a potential avenue to remedy prison-related injuries, which are no small matter. See Kingsley, 576 U.S. at 408 (Scalia, J., dissenting). Yet an objective-reasonableness-only approach would go further: It empowers federal juries to determine which state and local pretrial detention policies are reasonable. See, e.g., Poynter, 162 F.4th at 680 (requiring the “trier of fact” to resolve whether Barren County took “reasonable action to abate” “high risk[s] of harm to Poynter”).

* * *

Were we starting from scratch, some of us might well chart a different course, given the long-standing criticism that liability in this setting is unmoored from the Constitution. See Farmer, 511 U.S. at 859–60 (Thomas, J., concurring in the judgment); DiFraia v. Ransom, 171 F.4th 622, 628 (3d Cir. 2026); Edmo v. Corizon, Inc., 949 F.3d 489, 506–07 (9th Cir. 2020) (Bumatay, J., dissenting from the denial of rehearing en banc). But until the Supreme Court says otherwise, Farmer’s recognition of a right against a jailer’s subjective decision to punish binds us. See Thompson v. Marietta Educ. Ass’n, 972 F.3d 809, 814 (6th Cir. 2020) (quoting Rodriguez de Quijas v. Shearson/Am. Express, Inc., 490 U.S. 477, 484 (1989)). We return to that familiar standard today.

B. The district court in this case applied the Westmoreland framework. And the parties likewise developed the record under Westmoreland. That approach potentially changed the course of the litigation, as the “difference” between Farmer’s subjective test and the objective test is one that “matters.” Lawler, 93 F.4th at 928. And it matters not just in degree, but also in kind. Case in point, while the panel’s understanding of Kingsley centered on the reasonableness of the prison’s customs, Farmer hinges significantly on subjective awareness. For these reasons, we are ill-positioned today to resolve whether Poynter has met his burden at summary judgment under the proper subjective standard. Cf. Farmer, 511 U.S. at 849 (remanding to the lower court to apply the proper framework in the first instance); Celotex Corp. v. Catrett, 477 U.S. 317, 327– 28 (1986) (remanding to allow the lower court to determine whether a party adduced sufficient evidence to survive summary judgment). As we have often said, we are “a court of review, not of first view.” Cavin v. Mich. Dep’t of Corr., 927 F.3d 455, 459 (6th Cir. 2019) (quoting Cutter v. Wilkinson, 544 U.S. 709, 718 n.7 (2005)).

Accordingly, vacatur is the most appropriate course of action. Both parties agreed at oral argument that vacating and remanding was a permissible outcome if we discarded Brawner. Doing so affords Poynter a chance to develop the evidence he needs to attempt to satisfy Farmer’s subjective recklessness standard. And it allows the district court to take the first crack at evaluating any new evidence Poynter presents, all of which will aid our review should a subsequent appeal follow. We therefore decline to apply Farmer in the first instance.

IV.

A. Poynter raises several counterpoints, beginning with the contention that this case is an improper vehicle for reconsidering Brawner.

1. To Poynter’s mind, the party-presentation principle precludes our reconsideration of which standard governs his claim. That essential aspect of appellate procedure reminds us that “[t]he parties ‘frame the issues for decision,’ while the court serves as ‘neutral arbiter of matters the parties present.’” Clark v. Sweeney, 146 S. Ct. 410, 412 (2025) (per curiam) (quoting United States v. Sineneng-Smith, 140 S. Ct. 1575, 1579 (2020)); see also Sineneng-Smith, 140 S. Ct. at 1579; Margolin v. Nat’l Ass’n of Immigr. Judges, 146 S. Ct. 1285, 1288 (2026) (per curiam).

But there are a host of reasons why the principle does not fit today’s circumstance. First, the panel opinion raised the issue itself when it said, “[t]o the extent that Buetenmiller can be read as requiring subjective knowledge to satisfy the intentional-conduct prong, we hold that it is inconsistent with Brawner, and that Brawner continues to control.” Poynter, 162 F.4th at 678. When a panel injects an issue into a case that the parties have not, the en banc Court may consider new arguments in response to the panel’s choices. See Popovich v. Cuyahoga Cnty. Ct. of Common Pleas, Domestic Rels. Div., 276 F.3d 808, 817 n.8 (6th Cir. 2002) (en banc). Second, the panel opinion was one in a long line of intra-Circuit conflicts across this area of law. With this degree of “intractable” conflict in our jurisprudence, we owe it to ourselves, the district courts, and the litigants in our Circuit alike to resolve the debate with finality. United States v. McCall, 20 F.4th 1108, 1116 (6th Cir. 2021) (Kethledge, J., dissenting), vacated, 29 F.4th 816 (6th Cir. 2022) (en banc) (mem.). And third, while parties may waive or forfeit claims by their actions, they cannot bind our legal analysis through stipulations on the law. Young v. United States, 315 U.S. 257, 259 (1942) (holding that the meaning of the law “cannot be left merely to the stipulation of parties”); United States v. Cabbage, 91 F.4th 1228, 1231 (6th Cir. 2024) (“Parties aren’t allowed to stipulate to legal conclusions because courts have an independent obligation to get the law right.” (citing Neuens v. City of Columbus, 303 F.3d 667, 671 (6th Cir. 2002))); see also Collins v. Yellen, 141 S. Ct. 1761, 1798 n.2 (2021) (Gorsuch, J., concurring in part) (“[P]arties cannot stipulate to the law.”); United States v. Fields, 53 F.4th 1027, 1057–58 (6th Cir. 2022) (Murphy, J., concurring) (“[L]itigants cannot force us to create bad law through their agreements in the lawsuit.”). “When an issue or claim is properly before the court, the court is not limited to the particular legal theories advanced by the parties, but rather retains the independent power to identify and apply the proper construction of governing law.” Kamen v. Kemper Fin. Servs., Inc., 500 U.S. 90, 99 (1991) (citing Arcadia v. Ohio Power Co., 498 U.S. 73, 77 (1990)). The parties put forward a claim centered on the Fourteenth Amendment. That the parties initially agreed to litigate under the Brawner rubric does not mean we must accept the parties’ understanding of that case (an understanding that itself was modified by the panel here). Our allegiance is to the Constitution, and its proper meaning fairly trumps the litigating choices of individual parties.

The cases Poynter calls forward are inapposite. Start with the Supreme Court’s most recent word on party presentation: Margolin. There, the Fourth Circuit rested its decision on a new argument of its own creation “without giving either side a chance to address its theory.” Margolin, 146 S. Ct. at 1288. Here, on the other hand, we ordered supplemental briefing directing the parties to address the proper standard to be applied to Poynter’s claim—a course of action the Supreme Court has blessed. See Sineneng-Smith, 140 S. Ct. at 1582 (collecting examples). Much the same is true for Clark. There, the Fourth Circuit “grant[ed] relief on a claim that Sweeney never asserted and that the State never had the chance to address.” Clark, 146 S. Ct. at 412. But again, we have given both sides a chance to address the legal standard, duties they and amicus parties ably tackled, and we have not injected any new claims into this case. Poynter’s final decision—Sineneng-Smith—is likewise far afield from this case. The Ninth Circuit in that case invited specific groups to participate as amici curiae, file briefs addressing specific questions, and argue before the court for a longer period than Sineneng- Smith’s own counsel. Sineneng-Smith, 140 S. Ct. at 1580–81. Suffice it to say, nothing of the sort occurred here.

The dissenting opinion has an added objection to our en banc review. Bringing clarity to our conditions-of-confinement jurisprudence is unnecessary, the dissenting opinion says, because it sees no “genuine and widespread confusion” in that collection of decisions, neither in our Circuit nor beyond. Dissenting Op. at 37. Yet the dissenting opinion, in practice, only adds to that confusion. Consider the following: The one point that all seem to agree on is that Brawner modified only the subjective prong of Farmer. See Brawner, 14 F.4th at 596 (“Kingsley requires modification of the subjective prong of the deliberate-indifference test for pretrial detainees.”); see also Dissenting Op. at 34. Our subsequent decision in Westmoreland, on the other hand, developed the separate “intentional-action requirement” that “match[ed] the objective prong” of the Farmer test, not the subjective prong. 29 F.4th at 729. Westmoreland described this “intentional-action” element as capturing an officer’s “intentional decision” despite awareness of the “danger” to the detainee and the “concerns about [the detainee’s] well-being.” Id. A third decision, Buetenmiller, echoed Westmoreland’s understanding of the intentional-act requirement in a case where the defendant had no warning about the danger to the detainee. Buetenmiller, 53 F.4th at 945–46 (quoting Westmoreland, 29 F.4th at 729).

Although the dissenting opinion would have us leave these decisions untouched, it cannot reconcile them. The dissenting opinion commends the panel opinion for abrogating Buetenmiller’s reading of Westmoreland’s intentional-act requirement, which, again, “match[ed] the objective prong” of the Farmer test. Westmoreland, 29 F.4th at 729; see also Dissenting Op. at 36–39. According to the dissenting opinion, that reading collided with Brawner’s broader “work” in modifying Farmer’s distinct subjective prong. Dissenting Op. at 37. How Buetenmiller’s interpretation of Farmer’s objective prong could be inconsistent with Brawner’s modification of Farmer’s subjective prong needs no further examination other than to say that this sort of back and forth simply underscores the longstanding confusion in this area of law, a matter we resolve today.

2. Next, Poynter argues that this case does not directly implicate Brawner because Brawner dealt with pretrial detainees and Poynter is a civil contemnor. True, in Brawner (as in Kingsley), the plaintiff was a pretrial detainee rather than a civil contemnor. See Brawner, 14 F.4th at 589; Kingsley, 576 U.S. at 392. But the Fourteenth Amendment’s Due Process Clause governs more than just pretrial detainees; it also provides protections for civil contemnors, juveniles, and others detained by the state without the usual procedures associated with a criminal conviction. See Hopper, 887 F.3d at 754. We see no reason why the same deliberate- indifference framework would not apply to all Fourteenth Amendment conditions-of- confinement claims, regardless of the authority for detainment. As evidence, consider the fact that we apply Kingsley to pretrial detainees and civil contemnors in the excessive-force context. Id. at 754–55. In short, Poynter’s case implicates the standard for all due-process conditions-of- confinement claims, not just those brought by individuals detained for civil contempt.

Poynter responds that a lower standard must apply to civil contemnors because the government ought not have the same security concerns for civil contemnors as compared to pretrial detainees. As a result, he says, a jail must employ “the least possible power adequate to the end proposed.” Anderson v. Dunn, 19 U.S. (6 Wheat) 204, 231 (1821) (emphasis omitted). We see several flaws in that logic. First, Poynter undercuts his own argument by conceding that “[t]he Kingsley test fairly approximates Anderson by identifying confinement conditions that are objectively unreasonable.” Plaintiff-Appellant’s Suppl. En Banc Reply Br. at 4. If Kingsley is the proper test for civil contemnors, then Brawner (which extended Kingsley) applies with equal force to detainees of all sorts.

Second, we have our doubts about Poynter’s assumption that the government has no security interests with respect to civil contemnors. Contemnors, it bears emphasizing, have refused to comply with a court order, suggesting a willingness to disregard the law. See Willy v. Coastal Corp., 503 U.S. 131, 139 (1992) (citing U.S. Cath. Conf. v. Abortion Rts. Mobilization, Inc., 487 U.S. 72, 79 (1988)). Pretrial detainees, by contrast, merely stand accused of violating the law. Either way, the government’s assessment of the security risk each class of detainee poses is not of a constitutional dimension such that it gives unelected judges the ability to create a different standard of liability for jailers. See Bell, 441 U.S. at 547 (first citing Jones v. N.C. Prisoners’ Lab. Union, Inc., 433 U.S. 119, 128 (1977); then citing Procunier v. Martinez, 416 U.S. 396, 404–05 (1974); then citing Cruz v. Beto, 405 U.S. 319, 321 (1972) (per curiam); and then citing Meachum v. Fano, 427 U.S. 215, 228–29 (1976)).

Nor, as a third response, do we agree with Poynter that Youngberg v. Romeo, 457 U.S. 307 (1982), establishes a distinct objective inquiry for civil contemnors. For one, Youngberg’s viability post-Farmer is questionable. The Supreme Court has never invoked Youngberg’s purported objective test as a basis for a decision after Farmer. And others have questioned Youngberg’s relationship with the deliberate-indifference framework. See, e.g., Lewis, 523 U.S. at 863 (Scalia, J., concurring in the judgment) (treating Youngberg as a deliberate-indifference case); Hare v. City of Corinth, 74 F.3d 633, 646–47 (5th Cir. 1996) (en banc) (explaining that later decisions may have undermined Youngberg’s logic). For another, Youngberg’s actual standard does not do the work Poynter suggests it does. Youngberg formulated the test this way: “[L]iability may be imposed only when the decision by the professional is such a substantial departure from accepted professional judgment, practice, or standards as to demonstrate that the person responsible actually did not base the decision on such a judgment.” 457 U.S. at 323. To our minds, that construction is just as consistent with requiring subjective knowledge (because the defendant must have known the relevant facts for any departure to be a substantial departure from usual practice) as it is with an objective-only test. Doubly so when “judges or juries are [no] better qualified than appropriate professionals in making such decisions.” Id. at 322–23. In short, this case, as is true of all Fourteenth Amendment conditions-of-confinement cases, fairly implicates the Farmer-Kingsley debate.

B. Turning to the merits, Poynter contends that Kingsley’s excessive-force holding reaches conditions-of-confinement cases as well, thereby requiring an objective-only test for all Fourteenth Amendment deliberate-indifference claims.

Kingsley did nothing of the sort. Compare a Fourteenth Amendment excessive-force claim, the subject of Kingsley, with a Fourteenth Amendment deliberate-indifference claim based on conditions of confinement. And start with how we resolve whether a defendant had a “sufficiently culpable state of mind” under each type of claim. Wilson, 501 U.S. at 297. For use-of-force cases, the answer is relatively easy. Absent evidence that the defendant’s act was simply an accident, the act itself—the deliberate use of force—largely establishes his state of mind. See Kingsley, 576 U.S. at 396. Deliberate-indifference claims, on the other hand, typically turn on a defendant’s failure to act. See Strain, 977 F.3d at 991 (citing Castro, 833 F.3d at 1069); see also Gibson v. Abate, No. 24-1929, 2025 WL 1913247, at *6 (6th Cir. July 11, 2025) (Batchelder, J., concurring in part and dissenting in part) (citing Castro, 833 F.3d at 1086 (Ikuta, J., dissenting)). That distinction matters for purposes of constitutional liability. Recall that punishment requires a subjective state of mind. See Roberts, 773 F.2d at 725; see also Wilson, 501 U.S. at 300 (explaining that “some mental element must be attributed to the inflicting officer before” any “pain inflicted” constitutes “punishment”). It perhaps makes sense to infer punitive intent when someone acts (or fails to act) with knowledge of the risks. But where an individual fails to act without knowledge of any risks (even if a reasonable person would have recognized the risks), “the mere failure to act does not raise the same inference.” Castro, 833 F.3d at 1086 (Ikuta, J., dissenting) (citing Farmer, 511 U.S. at 837–38).

Put another way, it is difficult to know why someone did not act. The inaction could be purposeful. It could be reckless. Or it could be negligent. Without proof of what an individual knew when he failed to act, we lack evidence demonstrating a sufficiently culpable mens rea and, it follows, satisfying punishment’s intent requirement. See Wilson, 501 U.S. at 297–300; cf. Lewis, 523 U.S. at 849 (“[C]onduct intended to injure in some way unjustifiable by any government interest is the sort of official action most likely to rise to the conscience-shocking level.” (citing Daniels, 474 U.S. at 331)). Farmer’s subjective mens rea component captures this important distinction. Remember that we are talking about deliberate indifference. It would be strange to say that a defendant chooses deliberately to be indifferent to risks of which he is unaware.

Indeed, even Brawner did not seem to think that Kingsley’s objective test wholly controlled in the deliberate-indifference context. In Brawner’s words, Kingsley merely required “modification” of Farmer’s subjective prong. Brawner, 14 F.4th at 593. But see Poynter, 162 F.4th at 678–79 (understanding the inquiry to be wholly objective); Dissenting Op. at 38 (understanding Brawner to be an “explicit rejection of” Farmer’s subjective prong). Yet this modification highlights the mismatch between excessive force’s action and deliberate indifference’s inaction. Even in modifying earlier jurisprudence, Kingsley continued to mandate that the application of force underlying an excessive-force claim be deliberate. See 576 U.S. at 396. But in the deliberate-indifference setting, it is difficult to see how a failure to act could be deliberate if a defendant lacked the knowledge to infer the consequences of his failure. In other words, Kingsley was a case about the amount of force, not the decision to use force. Id. So taking the purely objective test from Kingsley and making it the whole nine yards in the deliberate-indifference context distorts Kingsley’s logic, which—like Farmer—relies on both subjective and objective inquiries.

Kingsley is inapt in the conditions-of-confinement setting for other reasons too. Returning to excessive force, consider the fact that the parallel protection against excessive force in the Fourth Amendment is framed as an objective inquiry. See Graham, 490 U.S. at 397. Kingsley thus deemed it sensible to apply a similar inquiry under the Due Process Clause. See 576 U.S. at 399. By contrast, the parallel provision for conditions-of-confinement deliberate- indifference claims, the Eighth Amendment, dictates more than an objective inquiry. See Farmer, 511 U.S. at 837. There is little reason why the standard for basic human needs (like medicine) would turn on the detainee’s status, thereby justifying our imposition of different requirements for parallel constitutional protections. See Bell, 441 U.S. at 533, 535.

Perhaps most telling is that Kingsley did not mention Farmer, let alone conditions-of- confinement claims or the deliberate-indifference standard. Quite the opposite: Kingsley emphasized that its holding was cabined to excessive-force claims. See 576 U.S. at 396 (explaining that “[w]hether [recklessness] might suffice for liability in the case of an alleged mistreatment of a pretrial detainee need not be decided here” because it was only “with respect to this question [of whether the force was excessive] that we hold that courts must use an objective standard”). Relatedly, all accepted in Kingsley that the jailers in question “acted purposefully or knowingly with respect to the force they used against Kingsley,” meaning the Supreme Court never considered an objective-only rule with respect to excessive-force claims, let alone deliberate-indifference claims. Id. So it would be unusual to think that Kingsley silently modified how every circuit at the time had applied Farmer in the Fourteenth Amendment context. Especially so when the Supreme Court has suggested that conditions-of-confinement claims should not turn on the status of the prisoner. See Bell, 441 U.S. at 533. Nor are we quick to infer that the Supreme Court has authorized a new theory of substantive due process. That is an area of law for which we have “scarce” guideposts, and so the “doctrine of judicial self- restraint requires us to exercise the utmost care whenever we are asked to break new ground,” Collins v. City of Harker Heights, 503 U.S. 115, 125 (1992), particularly when the jurisprudential trend has moved well away from recognizing new rights, see Washington v. Glucksberg, 521 U.S. 702, 720 (1997) (quoting Collins, 503 U.S. at 125); see also Guertin v. Michigan, 924 F.3d 309, 312 (6th Cir. 2019) (Sutton, J., concurring in the denial of rehearing en banc) (noting that substantive-due-process claims are not “anchored in the U.S. Constitution’s text”); id. at 315 (Kethledge, J., dissenting from the denial of rehearing en banc) (describing substantive due process as “the vaguest of constitutional doctrines”).

On this point, we of course agree that the Supreme Court recognizes a substantive-due- process right against pretrial punishment. See Dissenting Op. at 48; see also supra at 5. And we likewise agree that a constitutional right is distinct from the test used to vindicate the right. See Dissenting Op. at 48–50 (arguing that rights are not “coterminous with tests”). But we do not agree that the relevant test used to “show that the right was violated” has no bearing on the scope of the right itself. Id. at 48. Rather, the bounds of the right understandably relate to the applicable test for determining whether the right has been infringed. See, e.g., Lewis, 523 U.S. at 852–54 (applying separate frameworks to substantive-due-process claims arising from different factual scenarios despite the fact that the underlying right is the same); Richard H. Fallon, Jr., The Supreme Court, 1996 Term—Foreword: Implementing the Constitution, 111 Harv. L. Rev. 54, 62–66 (1997) (discussing how the test used to implement a right shapes the right’s content); see also Dissenting Op. at 29 (agreeing that the right at issue “drives” the applicable test). So easing the manner in which one may prove a deliberate-indifference substantive-due-process claim naturally expands our understanding of the right protected (here, a right to reasonable care in the jail setting), just as doing the opposite would constrict that understanding. Cf. Dobbs, 142 S. Ct. at 2271, 2283 (replacing the “undue burden test” with “rational-basis review” as “the appropriate standard” for substantive-due-process challenges to abortion regulations). Said differently, if “what test detained persons must satisfy” does not go to the substance of the constitutional right at issue, we do not see why negligence would be off the table for these claims; civil recklessness, after all, is merely a “test.” Dissenting Op. at 48. Nor, contrary to the dissenting opinion’s suggestion, does the manner in which the federal courts treat a “federal common law right to Indian homelands” have any bearing on the issue here. See id. (quoting City of Sherrill v. Oneida Indian Nation, 544 U.S. 197, 213 (2005)) (“There is a sharp distinction between the existence of a federal common law right to Indian homelands and how to vindicate that right.” (citation modified)).

The dissenting opinion likewise has a very different view of Kingsley. In places, it sees “broad” “signal[s]” and “general reasoning” advancing the cause of substantive due process, see Dissenting Op. at 42, cues that it seems to believe displace contrary indicators in other substantive-due-process cases, see, e.g., Dobbs, 142 S. Ct. at 2247–48. Yet in other places, the dissenting opinion opts for a more granular approach. In particular, it takes an exceedingly cramped view of Kingsley’s command that “[i]t is with respect to this question [of whether the force was excessive] that we hold that courts must use an objective standard.” 576 U.S. at 396. One could fairly read that proclamation as reserving for another day the distinct category of deliberate-indifference claims, thereby foreclosing the dissenting opinion’s attempt to impose Kingsley’s excessive-force holding in the conditions-of-confinement context. See supra at 22. So the dissenting opinion instead characterizes this statement as making only the narrow point that Kingsley was not addressing excessive force’s intentional-use-of-force prong. Dissenting Op. at 41 n.10. Yet even that reading undermines the dissenting opinion’s overall bottom line. Here, the dissenting opinion largely agrees that Kingsley, by not discussing excessive force’s intentional-use-of-force prong, retained a subjective element for those claims related to the decision to use force. See 576 U.S. at 396. How Kingsley nonetheless commands that we abandon any meaningful subjective component for distinct deliberate-indifference claims, as the dissenting opinion would have one believe, is left unsaid. Again, we decline the dissenting opinion’s invitation to read Kingsley as silently overruling every circuit court’s deliberate- indifference framework for pretrial detainees by turning it into a wholly objective test.

Against all of this, Poynter and the dissenting opinion largely abandon Brawner’s reasoning in favor of a point never addressed in that decision. Specifically, they seize upon language in Kingsley where the Supreme Court approvingly cited Bell v. Wolfish’s rational-basis test, which seemingly includes no mens rea requirement. Kingsley, 576 U.S. at 398; Dissenting Op. at 39–54 (referencing Bell over a dozen and a half times). But we doubt that Bell requires an objective test in this context when for decades we (along with every other circuit court) applied Farmer—not Bell’s rational-basis review—for deliberate-indifference claims. And that makes good sense, as Bell itself counseled that conditions-of-confinement claims should not turn on the status of the prisoner. See Bell, 441 U.S. at 533. Nor did Poynter ever frame his challenge as a rational-basis claim, favoring instead the framework of deliberate indifference. See Castro, 833 F.3d at 1085 (Ikuta, J., dissenting) (explaining that Bell’s objective forms of liability are distinct from deliberate-indifference liability). In any event, reliance on Bell’s rational-basis test gets us to the same place. When a prison affirmatively adopts a policy, such as the “double-bunking” policy at issue in Bell, we can infer an intent to punish if the policy is unrelated to a legitimate government interest or excessive in relation to the asserted interest. 441 U.S. at 530 (quoting Kennedy, 372 U.S. at 168–69); id. at 538–39 (citing Flemming, 363 U.S. at 613–17). What other reason, after all, would there be for adopting such a policy? But, for reasons discussed, that inference does not extend forward to a jailer’s failure to act or a jail’s failure to adopt a given policy. See Hare, 74 F.3d at 644–45 (“A State’s imposition of a rule or restriction during pretrial confinement manifests an avowed intent to subject a pretrial detainee to that rule or restriction . . . . [But w]ith episodic acts or omissions, intentionality is no longer a given, and Bell offers an ill-fitting test.”). Nor would Bell’s rational-basis review be appropriate in the typical deliberate- indifference setting, where claims tend to center not on jail policy, but instead on inaction by individual jailers. And substantive-due-process challenges to such actions, of course, necessarily require a state-of-mind element. See Lewis, 523 U.S. at 846 (observing that the “criteria to identify what is fatally arbitrary differ depending on whether it is legislation or a specific act of a governmental officer that is at issue”).

For this reason, we likewise do not join the dissenting opinion’s critique of Barren County’s counsel for failing at oral argument to point to another type of due-process claim that requires proving a defendant’s subjective intent. See Dissenting Op. at 44–45. That critique, we note, seems to be aided by the dissenting opinion’s reconstruction of the actual exchange at argument. Contrast Oral Arg. at 29:24–29:31 (asking for “other areas for Fourteenth Amendment due-process claims where we use subjective intent”), with Dissenting Op. at 44–45 (framing the inquiry as “County’s counsel was asked if he could identify any due-process claims that involve a similar inquiry into the defendant’s subjective knowledge of the risk or intent to punish”). Either way, the fact remains that Kingsley itself did not obviate a plaintiff’s need to prove that the defendant used force intentionally. See 576 U.S. at 396. And Daniels, which Kingsley cited approvingly, likewise required a pretrial detainee “to allege sufficient facts to support a claim of intentional deprivation” such that he could establish a “pillow was left on the stairs with the intention of harming him.” 474 U.S. at 334; see also Kingsley, 576 U.S. at 396 (citing Daniels, 474 U.S. at 331). Indeed, Daniels went so far as to reject constitutionalizing a jailer’s “special duty” of reasonable care against unreasonable risks of physical harm to a detainee, a standard largely indistinguishable from the one advocated for by the dissenting opinion. Daniels, 474 U.S. at 335–36 (citing Restatement (Second) of Torts § 314A(4) (A.L.I. 1965)).

C. The dissenting opinion next finds fault in our textualist reading of the relevant constitutional provisions. See Dissenting Op. at 45–47. According to the dissenting opinion, our long-standing adherence to a subjective requirement for both Eighth and Fourteenth Amendment claims either reads the words “cruel and unusual” out of the Eighth Amendment’s text or reads them into the Fourteenth Amendment’s. Id. at 46. But Farmer, remember, made plain that failing “to alleviate a significant risk” about which a defendant did not know “cannot . . . be condemned as the infliction of punishment.” 511 U.S. at 838. In other words, subjective knowledge in this setting informs whether punishment occurred at all, not whether the punishment was cruel and unusual.

Reading Farmer in that way, the dissenting opinion responds, ignores the “discrete” issues addressed there, which, the dissenting opinion adds, set the “context” for that decision. Dissenting Op. at 51 n.13 (quoting Nat’l Pork Producers Council v. Ross, 143 S. Ct. 1142, 1155 (2023)). But query which approach better appreciates context: one that draws from a case about when deliberate indifference constitutes punishment (Farmer) and invokes nearly three decades of decisions applying that understanding, or one that draws from an excessive force case (Kingsley) that never mentioned deliberate indifference and expressly leaves untouched a separate subjective element.

Relatedly, the dissenting opinion believes that excessive force’s objective inquiry must also control in the deliberate-indifference context because both types of claims arise under the Due Process Clause, and because excessive force and deliberate indifference “can cause the same injuries.” Dissenting Op. at 43 (quoting Castro, 833 F.3d at 1070). In every other constitutional setting, however, those claims operate under different standards. See, e.g., Whitley, 475 U.S. at 320–21 (quoting Johnson, 481 F.2d at 1033); Farmer, 511 U.S. at 836 (creating distinct tests for excessive-force and failure-to-protect claims for convicted prisoners). The dissenting opinion’s approach thus has the odd feature of ordaining pretrial detainees as the lone set of claimants who can establish excessive-force and deliberate-indifference failure-to-protect claims with the same evidence (or lack thereof) of subjective knowledge. Nothing in the Due Process Clause compels that strange result.

D. Lastly, Poynter and the dissenting opinion characterize Brawner’s deliberate- indifference framework as requiring recklessness—albeit civil recklessness—not negligence. But that seems to be little more than a labeling choice when, in practice, Brawner requires only negligence. Recall that Poynter’s understanding of recklessness requires serious risks plus unreasonable action. All that formulation does is retack on the word recklessness to the ordinary negligence standard. Our experience with Poynter’s approach teaches that a negligence standard is almost always the end result, if not the starting place. Cf. Cummings v. Missouri, 71 U.S. (4 Wall.) 277, 325 (1867) (“The Constitution deals with substance, not shadows.”).

True, as the dissenting opinion emphasizes, its deliberate-indifference framework does include an intentionality component. In the dissenting opinion’s design, however, that component is exceedingly easy to satisfy. As evidence, look to our Circuit’s post-Brawner jurisprudence, which downplayed any intentionality requirement when describing an objective- reasonableness-only deliberate-indifference test. See, e.g., Howell v. NaphCare, Inc., 67 F.4th 302, 311–12 (6th Cir. 2023); Mercer v. Athens County, 72 F.4th 152, 161 (6th Cir. 2023). Any conscious act by a jailer seemingly qualifies as an intentional act, even if the act is divorced from any appreciation of the risks the detainee faces (for example, the act of housing an inmate, choosing a general monitoring practice of a detainee, and the like). Dissenting Op. at 37–38 & n.9, 53, 55–58.

We should not transform deliberate indifference’s intentionality requirement into a mere box-checking exercise. Recall that the Due Process Clause’s use of the word “deprive” requires a subjective state of mind. See supra at 11–12. The deprivation’s direct object, a “person,” in turn, must be deprived of the prepositional complements listed in the Clause: life, liberty, or property. U.S. CONST. amend. XIV, § 1. So the constitutional text requires the jailer to be aware of the challenged deprivation of liberty (which, in this case, is the beating by Guess and Wix, not the mere placement in the cell). Only this subjective awareness of the liberty deprivation (not of some other fact) differentiates conscience-shocking behavior that amounts to an unconstitutional abuse of power from mere accidental or negligent conduct by the jailer. See Lewis, 523 U.S. at 848–49. Yet the dissenting opinion would separate the state-of-mind element from its textual foundation, leaving a jailer liable so long as he has engaged in a conscious act, even if he is unaware that the action creates a risk to the detainee’s liberty. In practice, that purported limit on deliberate-indifference liability is no limit at all. After all, virtually every plaintiff can point to some intentional choice by the defendant that preceded the plaintiff’s injury. In Poynter’s case, for example, the panel derived an intentional act from the fact that there was no evidence Guess and Wix were accidentally placed in the same cell as Poynter. Poynter, 162 F.4th at 679. Tellingly, neither the panel nor the dissenting opinions tell us how far down the chain of causation the requisite intentional act can occur. Poynter’s booking assessment at the jail seems to be enough. Dissenting Op. at 38 & n.9. What about the decision to place him in Barren County’s detention center rather than another one? The examples are endless. See, e.g., id. at 57–58 (citing Kemp v. Fulton County, 27 F.4th 491, 497 (7th Cir. 2022); Pittman IV, 108 F.4th at 570–71 (noting that a guard failing to put in his hearing aid satisfies the intentional-act requirement)). At day’s end, the dissenting opinion’s conception of deliberate indifference all but eliminates any intentional-act requirement.

The dissenting opinion’s last-ditch attempt to constrain that seemingly limitless approach by invoking a “proximate cause” element—a phrase never used in Kingsley, Brawner, the Westmoreland majority opinion, the Poynter panel opinion, or plaintiff’s en banc briefs—only confirms the dissenting opinion’s fluid methods. See id. at 60 (suggesting that “merely booking Poynter” is “too far removed” to be a proximate cause of constitutional injury despite tying liability to Poynter’s security assessment, which occurred at the time of booking); id. at 59 (defending its “advocacy for proximate-causation principles”); id. (embracing the dissenting opinion in Westmoreland, which itself criticized the majority opinion for “omit[ting] a proximate cause requirement,” 29 F.4th at 740 (Bush, J., dissenting)). Nor, in any event, would the concept prove to be the limiting factor the dissenting opinion believes. In general, “the term ‘proximate cause’ is a poor one to describe limits on the scope of liability” because “it is confusing for a jury” and “multiple proximate causes are often present.” Restatement, supra, § 29 cmt. b; see also Palsgraf v. Long Island R.R. Co., 162 N.E. 99, 103 (N.Y. 1928) (Andrews, J., dissenting) (describing proximate cause as where “the law arbitrarily declines to trace a series of events beyond a certain point” “because of convenience, of public policy, [or] of a rough sense of justice”). That is particularly true in the jailhouse setting, as the dissenting opinion itself reflects. The dissenting opinion explains that placing a detainee in a cell will satisfy the intentionality prong, Dissenting Op. at 38 & n.9, 56, and it adds that doing so where harm later befalls the detainee satisfies the proximate-cause prong, id. at 60. But everyone in jail is told where they must be (whether it is their cell, the canteen, the yard, or somewhere else) by jail officials. So it is difficult to imagine a situation where the dissenting opinion’s conception of the intentional-act requirement is satisfied, but where its conception of the proximate-cause element is not.

The dissenting opinion’s malleable approach is all the more evident when one considers how it framed the intentional act here: the choice not to classify Guess and Wix based on their institutional histories and instead only consider their charges and behavior during arrest. Dissenting Op. at 47. But with “the line between action and inaction [often] blurry and a function of characterization,” id. at 53 (citing Koulta v. Merciez, 477 F.3d 442, 445–46 (6th Cir. 2007)), any inaction can be reframed as an intentional choice not to act. Look no further than the original panel opinion, which deemed defendants’ “failure to classify” Guess and Wix based on their institutional history as conduct relevant to the Due Process Clause’s deliberate-act requirement. Poynter, 162 F.4th at 679 (emphasis added). Again, if the failure to act can satisfy the intentional-act requirement, that simply returns our focus to an issue already addressed, namely, whether a failure to act permits an inference of punitive intent for the purpose of a deliberate-indifference claim. See Castro, 833 F.3d at 1086 (Ikuta, J., dissenting) (citing Farmer, 511 U.S. at 837–38). If it does, then the deliberate-indifference standard amounts to little more than negligence, a point all agree is at odds with settled Due Process Clause jurisprudence.

* * *

Farmer v. Brennan’s subjective standard properly governs deliberate-indifference cases brought under the Fourteenth Amendment. We vacate the district court’s order and remand to allow the district court to consider in the first instance how Farmer applies in this case.

DISSENT

KAREN NELSON MOORE, Circuit Judge, dissenting. In Brawner v. Scott County, 14 F.4th 585 (6th Cir. 2021), we held that pretrial detainees who bring deliberate-indifference claims under the Fourteenth Amendment need to prove that they faced an unjustifiably high risk of harm that was known or so obvious that it should have been known and that the defendant intentionally acted or failed to act in the face of that risk. Brawner’s test followed from the Supreme Court’s decision in Kingsley v. Hendrickson, 576 U.S. 389 (2015), and the majority of our sibling circuits that have addressed the question on the merits agree. Today, the majority discards that approach even though neither party asked us to do so. I respectfully dissent.

I. BACKGROUND

Barren County Detention Center (“BCDC”) is a jail that housed state and county detainees together. See R. 95-5 (Boston Dep. at 34) (Page ID #524); R. 100-4 (Sweeney Dep. at 101) (Page ID #785).¹1. “County” detainees include both pretrial detainees and incarcerated persons convicted of misdemeanors. “State” detainees are incarcerated persons who have been convicted of felonies but have been “assigned to serve their sentences in local jails.” D. 14 (Appellant Br. at 9 n.6); R. 95-3 (Bennett Dep. at 69) (Page ID #469) (testifying that once an inmate is sentenced, he becomes a state inmate). Luther Poynter was arrested and taken to BCDC on December 25, 2020, for civil contempt because he did not pay child support. R. 95-2 (Facility Admission Rep.) (Page ID #451); R. 100-2 (Fernandez Dep. at 33) (Page ID #714).²2. I recount the facts in the light most favorable to Poynter because we must view the facts in the light most favorable to the nonmoving party at summary judgment. See Greene v. Crawford County, 22 F.4th 593, 605 (6th Cir. 2022). On December 28, Poynter was placed in general-population cell 524, where Scotty Wix and Timothy Guess were already housed. R. 101-2 (12/28/2020 Incident Rep. at 1–2) (Page ID #890–91); Video Footage at 0:00– 1:30; R. 95-3 (Bennett Dep. at 86–95, 111) (Page ID #474–76, 480). In less than one minute and thirty seconds, Guess and Wix attacked Poynter, punching him repeatedly in the head even though he was seated and had not moved towards them or made any movements that indicated he was going to. Video Footage at 01:20–01:25. Poynter suffered a traumatic brain injury and now is partially paralyzed on his right side, struggles with his short-term memory, uses a wheelchair, and lives in an assisted-living facility. R. 100-2 (Fernandez Dep. at 7–8, 64–68) (Page ID #707, 721–22).

Guess and Wix were both state detainees and “frequent fl[iers]” in BCDC. R. 95-3 (Bennett Dep. at 32–33) (Page ID #460); R. 100-5 (Hawkins Dep. at 27) (Page ID #819). Both had extensive histories of violence in BCDC, which are reflected in incident reports in BCDC’s information system, JailTracker. All said, their JailTracker files together contained twenty-eight incident reports, prior to Poynter’s attack, that detail their erratic and often violent behavior in BCDC. R. 101-2 (Incident Reps.) (Page ID #856–94). Before the attack on Poynter, Guess had attacked his cellmates six times, fought with other detained persons twice, threatened others several times, and was involved in several verbal altercations.³3. R. 101-2 (5/19/2008 Incident Rep.) (Page ID #857); id. (12/26/2012 Incident Rep.) (Page ID #860); id. (5/4/2019 Incident Rep.) (Page ID #863); id. (5/11/2019 Incident Rep.) (Page ID #864); id. (9/3/2019 Incident Rep.) (Page ID #866); id. (1/24/2020 Incident Rep.) (Page ID #874); id. (8/26/2018 Incident Rep.) (Page ID #861); id. (10/7/2019 Incident Rep.) (Page ID #870); id. (7/6/2019 Incident Rep.) (Page ID #865); id. (7/7/2020 Incident Rep.) (Page ID #875); id. (9/11/2019 Incident Rep.) (Page ID #868); id. (12/19/2020 Incident Rep.) (Page ID #877). Guess was also aggressive with jail staff four times, leading BCDC staff to place him in restraints three times and to pepper spray him.⁴4. R. 101-2 (2/7/2007 Incident Rep.) (Page ID #856); id. (1/23/2010 Incident Rep.) (Page ID #858); id. (11/4/2019 Incident Rep.) (Page ID #872–73); id. (12/10/2020 Incident Rep.) (Page ID #876). Wix had a lengthy history of violence too. Before attacking Poynter, he had charged at and attempted to attack jail staff, attacked his cellmates five times, and fought with his cellmate.⁵5. R. 101-2 (9/28/2010 Incident Rep.) (Page ID #881); id. (12/27/2013 Incident Rep.) (Page ID #884); id. (2/15/2014 Incident Rep.) (Page ID #885); id. (7/19/2017 Incident Rep.) (Page ID #886); id. (5/11/2019 Incident Rep.) (Page ID #887); id. (2/15/2020 Incident Rep.) (Page ID #888); id. (10/14/2020 Incident Rep.) (Page ID #889). In the approximately eleven months leading up to Poynter’s assault, Guess and Wix were involved in the following incidents: Guess and Wix both assaulted other detained persons, id. (1/24/2020 Incident Rep.) (Page ID #874); id. (2/15/2020 Incident Rep.) (Page ID #888); Wix fought with and punched another detained person in his cell, id. (10/14/2020 Incident Rep.) (Page ID #889); Guess threatened another detained person and had to be removed from kitchen- work duty, id. (7/7/2020 Incident Rep.) (Page ID #875); Guess got into a verbal altercation with his cellmate, id. (12/19/2020 Incident Rep.) (Page ID #877); and Guess behaved violently and erratically towards jail staff, leading them to place him in restraints, id. (12/10/2020 Incident Rep.) (Page ID #876). Guess also had nearly thirty “keep aparts” in his file before the attack on Poynter, and Wix had nearly twenty.⁶6. A “keep-apart” or “keep-away” order reflects that the detained persons listed in the order had a history that required them to be separated from each other. R. 95-3 (Bennett Tr. at 74–76) (Page ID #471); R. 95-5 (Boston Dep. at 31) (Page ID #523). R. 101-3 (Offender Alerts at 1–4) (Page ID #895–98). In contrast, Poynter had one. Id. at 5 (Page ID #899).

Additionally, Guess and Wix were not detained in BCDC for the full eleven months before the December 28, 2020, attack. Guess had been most recently booked into BCDC on December 10, 2020, and Wix was booked the day before, on December 9, 2020. R. 95-3 (Bennett Dep. at 32–33, 68) (Page ID #460, 469). That makes sense because Bennett also testified that all the people housed in cell 525 on December 28, 2020, had been booked into BCDC in December. Id. at 96–97 (Page ID #476). That Guess and Wix were involved in fewer attacks in the months immediately preceding December 28, 2020, may therefore have been a product of their not being in BCDC, rather than an indication of improving behavior.

Despite their histories of violence, Guess and Wix were housed in a minimum-security general-population cell. BCDC produced no evidence that it ever classified or reclassified Guess and Wix. According to Poynter’s expert, Wix’s and Guess’s histories of violence demonstrated that they posed a threat to others and should not have been in a general-population cell. R. 95-7 (Eiser Dep. at 37–39, 51) (Page ID #550–51, 554).

Classification is a form of risk assessment that keeps detained persons and jail staff safe, and implementing it correctly is essential to minimize the risk of violence in jails. Id. at 29 (Page ID #548); R. 95-3 (Bennett Dep. at 136–37) (Page ID #486); R. 100-4 (Sweeney Dep. at 23–25) (Page ID #766). According to official BCDC policy and Kentucky law, each detained person must be classified based on their behavioral history at BCDC, among other factors, when they are initially booked into the jail. R. 101-1 (BCDC Classification Policy at 1–2) (Page ID #854– 55). Classifications also must be updated based on “[d]ay-to-day behavior as observed by jail staff,” disciplinary actions, and changes in mental or emotional status. Id. at 2 (Page ID #855). Institutional behavior is the “biggest” component of the classification analysis, R. 95-7 (Eiser Dep. at 30) (Page ID #549), and it must be considered to determine whether a detained person could harm others, R. 100-4 (Sweeney Dep. at 30) (Page ID #768).

BCDC’s custom, however, was to ignore its written policy. It classified detained persons without considering their institutional behavioral history.⁷7. The sole way that BCDC considered institutional behavior was if the detained person had an active “keep apart” from another individual in BCDC, which would result only in the two people subject to the order being separated. R. 95-5 (Boston Dep. at 25–33) (Page ID #522–24). Keep aparts would not result in changes to their classifications. Id. See R. 95-5 (Boston Dep. at 25–33) (Page ID #522–524). County detainees were classified based on only their charges and behavior during arrest and booking. Id. at 25 (Page ID #522). BCDC did not classify state detainees at all, let alone consider their institutional history. Instead, BCDC relied on the state Department of Correction’s classifications. Id. at 27–28 (Page ID #522); R. 100-5 (Hawkins Dep. at 19–20) (Page ID #817); R. 100-6 (Smith Dep. at 40) (Page ID #837). The state’s classifications, however, designated where a detained person could be assigned work duty, not where they should be housed. See R. 100-5 (Hawkins Dep. at 11) (Page ID #815); R. 95-3 (Bennett Dep. at 62) (Page ID #468).

Poynter, by and through his guardian and sister, Anita Fernandez, sued the Barren County Jailer and Barren County (“the County”) under 42 U.S.C. § 1983, alleging that their deliberate indifference to his safety violated the Fourteenth Amendment. See R. 32 (Am. Compl.) (Page ID #149–57). The district court granted summary judgment to the County, and Poynter appealed. A panel of this court reversed, holding that Poynter had presented sufficient evidence for a reasonable jury to conclude that he had established a deliberate-indifference claim. A majority of this court subsequently voted to grant en banc review and ordered the parties to submit supplemental briefing to address the appropriate legal standard for Fourteenth Amendment deliberate-indifference claims.

II. ANALYSIS

The Eighth Amendment protects convicted persons from “cruel and unusual punishments.” U.S. Const. amend. VIII. Based on the prohibition of “cruel and unusual punishments,” id., prisons “have a duty . . . to protect [convicted persons] from violence at the hands of other [incarcerated persons],” Westmoreland v. Butler County, 29 F.4th 721, 726 (6th Cir. 2022) (first alteration in original) (quoting Farmer v. Brennan, 511 U.S. 825, 833 (1994)). An incarcerated person’s constitutional rights are violated if the prison is “deliberate[ly] indifferen[t] to [their] health or safety.” Id. (first two alterations in original) (quoting Farmer, 511 U.S. at 834). Pretrial and civil detainees’ claims for deliberate indifference, however, “derive . . . from the Fourteenth Amendment,” and “not from the Eighth Amendment, with its focus on punishment.” Grote v. Kenton County, 85 F.4th 397, 405 (6th Cir. 2023); see also Hopper v. Plummer, 887 F.3d 744, 755 (6th Cir. 2018). This protection is derived from the Due Process Clause’s prohibition on deprivations of “liberty” without due process. Youngberg v. Romeo, 457 U.S. 307, 315, 320–22 (1982).

Despite the different sources of the rights, we “historically analyzed” deliberate- indifference claims brought under the Eighth and Fourteenth Amendments “under the same rubric.” Brawner, 14 F.4th at 591 (quoting Richmond v. Huq, 885 F.3d 928, 937 (6th Cir. 2018)). We required plaintiffs to prove an “objective” component—that the plaintiff faced a substantial risk of serious harm—and a “subjective” component—that the defendant actually knew of this substantial risk and disregarded it. Id. This standard was crafted by the Supreme Court in Farmer, a case addressing deliberate-indifference claims under the Eighth Amendment, not the Fourteenth. 511 U.S. at 828. I refer to this as the criminal-recklessness or Farmer standard.

We changed our approach in the wake of Kingsley, which counseled that there are important differences between the Eighth and Fourteenth Amendments. 576 U.S. at 400–02. In recognition of those differences, we held that a plaintiff who brings a Fourteenth Amendment deliberate-indifference claim must show that the defendant “acted deliberately (not accidentally), but also recklessly ‘in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.’” Brawner, 14 F.4th at 596 (quoting Farmer, 511 U.S. at 836). Brawner therefore modified the subjective-knowledge-of-the-risk requirement—plaintiffs did not have to prove that the defendant knew of the unjustifiably high risk, but only that a reasonable person in their position would have known. Put another way, Brawner requires intentional action but not subjective knowledge of the consequences of those actions. Pulling this all together, Brawner required plaintiffs to prove that the defendant intentionally acted in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known. I refer to this as the civil-recklessness or Brawner standard.

The majority now overrules Brawner and reverts to the Farmer criminal-recklessness standard for Fourteenth Amendment cases. I begin by discussing why this case is not a proper vehicle to do so in Part II.A and then turn to the merits and explain why Brawner’s test accurately captured and reflected the differences between the Eighth and Fourteenth Amendments in Part II.B.

A. Party-Presentation Principle

Until the en banc majority intervened, the County had never argued that Brawner should be overruled, that it was incorrectly decided, or even that the district court or the original panel applied it incorrectly. See R. 95-1 (County Mot. for Summ. J.) (Page ID #435–50); D. 15 (Appellee Br.); D. 27 (County Pet. Reh’g En Banc). Instead, the County cited Brawner approvingly. The district court and the panel followed the County’s lead and applied Brawner. Poynter ex rel. Fernandez v. Bennett, No. 1:21-CV-00162, 2025 WL 606183, at *4 (W.D. Ky. Feb. 25, 2025); Poynter ex rel. Fernandez v. Bennett, 162 F.4th 664, 677–78 (6th Cir. 2025). In its petition for rehearing en banc, the County again did not argue that we should overrule Brawner. See D. 27 (County Pet. Reh’g En Banc). In fact, the County’s en banc petition did not even argue that the panel erred in the portion of the deliberate-indifference test that Brawner modified. Id. Instead, it made only two arguments. First, that Poynter was not placed at a substantial risk of serious harm because the risk was not personal to Poynter. Id. at 10–13. Second, that the County could not be held liable under Monell v. Department of Social Services of City of New York, 436 U.S. 658 (1978), because there was not a pattern of similar constitutional violations. D. 27 (County Pet. Reh’g En Banc at 13–17). Therefore, the County forfeited any argument that Brawner should be overruled. See Citizens Coal Council v. EPA, 447 F.3d 879, 905 (6th Cir. 2006) (en banc). Nevertheless, the majority ordered the parties to brief the question and now overrules Brawner.

“In our adversarial system of adjudication, [courts] follow the principle of party presentation.” United States v. Sineneng-Smith, 590 U.S. 371, 375 (2020). Under that principle, “in the first instance and on appeal, [courts] rely on the parties to frame the issues for decision and assign to courts the role of neutral arbiter of matters the parties present.” Id. (alterations omitted) (quoting Greenlaw v. United States, 554 U.S. 237, 243 (2008)). “To put it plainly, courts ‘call balls and strikes’; they don’t get a turn at bat.” Clark v. Sweeney, 607 U.S. 7, 9 (2025) (per curiam) (quoting Lomax v. Ortiz-Marquez, 590 U.S. 595, 599 (2020)). Displeased with the parties’ framing of the issue and seemingly wanting a “turn at bat,” the majority overlooks the County’s forfeiture. Id. In fact, the County forfeited the issue three times over: in the district court, before the original panel, and in its petition for rehearing. See Scott v. First S. Nat’l Bank, 936 F.3d 509, 522–23 (6th Cir. 2019); Guyan Int’l, Inc. v. Pro. Benefits Adm’rs, Inc., 689 F.3d 793, 799 (6th Cir. 2012). The County addressed this question only after the court ordered it to. I would not endorse inviting the County to change its position via supplemental briefing as an acceptable way to raise the issue here. Cf. Valent v. Comm’r of Soc. Sec., 918 F.3d 516, 524 (6th Cir. 2019).

Additionally, we excuse forfeiture only “in exceptional cases or to avoid a miscarriage of justice.” Citizens Coal Council, 447 F.3d at 905 (quoting Mayhew v. Allsup, 166 F.3d 821, 823 (6th Cir. 1999)). This is not an exceptional case, and no miscarriage of justice would result from applying Brawner. The County is sophisticated and well-versed in the evolution of our caselaw on this issue. Id.; see R. 95-1 (County Mot. for Summ. J. at 8–10) (Page ID #442–44); D. 15 (Appellee Br. at 16–17, 19–20). The County “had a full and fair opportunity” to raise this issue before the district court, the original panel, and in its petition for rehearing. Citizens Coal Council, 447 F.3d at 905. It chose not to. I would hold the County to that choice.

Contrary to the majority’s assertion, the panel’s treatment of Buetenmiller v. Macomb County Jail, 53 F.4th 939, 945–46 (6th Cir. 2022), does not solve the party-presentation issue. Majority Op. at 16. The panel stated merely that “[t]o the extent that Buetenmiller can be read as requiring subjective knowledge [of the risk] to satisfy the intentional-conduct prong, we hold that it is inconsistent with Brawner, and that Brawner continues to control.” Poynter, 162 F.4th at 678. The panel was not breaking new ground or altering Brawner as the majority asserts, Majority Op. at 16, by holding that requiring the plaintiff to prove that the defendant knew of the risk was inconsistent with Brawner—another panel had already done just that in Helphenstine v. Lewis County, 60 F.4th 305, 317 (6th Cir. 2023). Helphenstine held that Trozzi v. Lake County, 29 F.4th 745, 757–58 (6th Cir. 2022), was “irreconcilable with Brawner” because it required the plaintiff to prove that “the prison official knew that his failure to respond would pose a serious risk to the pretrial detainee,” and Brawner did away with the “actual knowledge” requirement. Helphenstine, 60 F.4th at 316 (quoting Trozzi, 29 F.4th at 757–58). Nor was the panel reaching out to decide an issue that the parties did not present, as the majority contends. Majority Op. at 16. The panel discussed Buetenmiller to distill and apply the very standard that the parties agreed controlled.⁸8. And, again, the County had the opportunity to respond to the panel’s analysis in its en banc petition but chose not to.

Finally, no widespread confusion in our caselaw warrants such intervention. I am skeptical that Brawner created genuine and widespread confusion in this court. See Grote, 85 F.4th at 406 (discussing how Trozzi, which the majority cites to demonstrate confusion, “attempted to resurrect the pre-Brawner (and pre-Kingsley) treatment of pretrial detainees’ deliberate-indifference claims”). What work would Brawner’s rejection of the Farmer criminal- recklessness standard have done if it allowed later panels to resurrect the knowledge-of-the-risk requirement? The only logical answer is none, so such an interpretation clearly cannot be reconciled with Brawner.

Brawner’s intentional-act requirement never invited courts to inquire into the defendant’s knowledge of the risk. The intentional-action requirement asked only whether the defendant’s conduct itself was intentional. See Westmoreland, 29 F.4th at 729 (explaining that Brawner’s intentional-action requirement asked whether “[t]he defendant made an intentional decision with respect to the conditions under which the plaintiff was confined” (quoting Castro v. County of Los Angeles, 833 F.3d 1060, 1071 (9th Cir. 2016) (en banc))); Brawner, 14 F.4th at 597 (“Brawner needed to [show] . . . (1) that she had an objectively serious medical need; and (2) that [the defendant’s] action (or lack of action) was intentional (not accidental) and she either (a) acted intentionally to ignore Brawner’s serious medical need, or (b) recklessly failed to act . . . . (emphasis added)); Davis v. Chorak, No. 22-1839, 2023 WL 2487339, at *3 (6th Cir. Mar. 14, 2023) (holding that the plaintiff satisfied “the [intentional-conduct] element . . . [by] plausibly alleg[ing] that [the defendants] made intentional security-classification decisions” when they classified the plaintiff and his attacker as medium security, leading them to be housed together).⁹9. A few examples provided by the Ninth Circuit further illustrate this. If a plaintiff brought a claim that they were housed with a dangerous cellmate, “the inquiry at this step would be whether the placement decision was intentional.” Castro, 833 F.3d at 1070. Alternatively, “if the claim relates to inadequate monitoring of the cell, the inquiry would be whether the officer chose the monitoring practices rather than, for example, having just suffered an accident or sudden illness that rendered him unconscious and thus unable to monitor the cell.” Id. Reading the intentional-act requirement to ask whether the defendant took action with knowledge of the risk to the plaintiff would clearly render the Brawner civil-recklessness standard identical to the Farmer criminal-recklessness standard. But that cannot be right, because then Brawner’s explicit rejection of Farmer’s knowledge-of-the-risk requirement would have done no work at all.

The majority objects, asserting that my discussion here “adds to th[e] confusion.” Majority Op. at 17. “Brawner modified only the subjective prong of Farmer,” the argument goes, and Westmoreland, on the other hand, recognized a separate intentional-action requirement that was part of Farmer’s objective, rather than subjective, prong. Id. But, as already explained, Brawner itself also contained the intentional-act requirement. Brawner, 14 F.4th at 596 (holding that plaintiffs must show that the defendant “acted deliberately (not accidentally), but also recklessly ‘in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known’” (quoting Farmer, 511 U.S. at 836)). Consistent with that test, Westmoreland explained that the intentional-act requirement asked whether the defendant “made an intentional decision as to [the plaintiff’s] conditions of confinement.” Westmoreland, 29 F.4th at 729. In applying that requirement, Westmoreland evaluated what information the defendant had been exposed to, which perhaps more comfortably fits into the analysis of whether a reasonable person in the defendant’s position would have recognized the obvious risk. Id. But Westmoreland never required the plaintiff to present evidence to allow “a reasonable juror [to] conclude that officials knew about those dangers” or considered whether the individual defendant’s subjective “concerns” about the potential risk had been “allayed,” which Buetenmiller did. Buetenmiller, 53 F.4th at 945–46. Asking whether an “official[] knew about th[e] danger[]” is precisely what Brawner held was no longer required under the Fourteenth Amendment. See Helphenstine, 60 F.4th at 316–17 (holding that requiring the defendant to have known of the risk was “irreconcilable with Brawner”). So, indeed, Buetenmiller is the outlier.

The majority also makes much of the fact that Westmoreland described the intentional-act requirement as part of the “objective” component of the test. Westmoreland, 29 F.4th at 729. According to the majority, because the intentional-act requirement is labeled as “objective,” it is immune from criticism because Brawner did not modify the “objective” portion of the test. That is wrong. A panel cannot revive the precise knowledge-of-the-risk requirement that Brawner rejected by simply adding it to the “objective” prong of the analysis. Regardless of where the knowledge-of-the-risk requirement is placed, it is the precise element that Brawner rejected.

This “objective” versus “subjective” argument portends a recurring problem with the majority’s argument. The majority places great weight on these terms, to which the majority ascribes different definitions as necessary to advance its arguments, without considering the content of the purportedly “objective” or “subjective” requirements.

In the end, there is nothing about this case that makes it exceptional enough to excuse the County’s multiple forfeitures. Therefore, the en banc majority should have adhered to the party presentation-principle here and not reached out to decide an issue neither party raised or addressed until the en banc majority intervened.

B. Fourteenth Amendment Deliberate-Indifference Standard

On the merits, the Supreme Court has not directly addressed what standard governs Fourteenth Amendment deliberate-indifference claims. See, e.g., City of Canton v. Harris, 489 U.S. 378, 388 n.8 (1989). It has, however, provided relevant guidance. First, in Bell v. Wolfish, the Court evaluated various conditions of confinement that were challenged by pretrial detainees under the Fourteenth Amendment. 441 U.S. 520, 523 (1979). The Court held that because “a detainee may not be punished prior to an adjudication of guilt in accordance with due process of law,” if a condition of confinement does not advance a legitimate government interest, it violates the Due Process Clause. Id. at 535–39. Bell, therefore, created a test to analyze whether a condition of confinement violated the Fourteenth Amendment’s Due Process Clause that did not require proof of a subjective intent to punish, so the Court analyzed the conditions of confinement challenged by the plaintiffs, such as a policy of double bunking detainees, using purely objective evidence. Id. at 541–43.

In reaching this holding, the Supreme Court rejected the Second Circuit’s test that required the government to justify conditions of confinement by a compelling necessity, which it had derived primarily from the presumption of innocence. Id. at 532–34. The Supreme Court rejected the Second Circuit’s reliance on the presumption of innocence as the “source of the detainee’s substantive right to be free from conditions of confinement that are not justified by compelling necessity,” and held that the presumption of innocence “has no application to a determination of the rights of a pretrial detainee during confinement before his trial has even begun.” Id. at 532–33. But that did not mean that the status of pretrial detainees had no relevance in determining what conditions they could be subjected to. The Court instead held that the pretrial detainees’ rights derived from their right to be free from any punishment under the Fourteenth Amendment’s Due Process Clause and then contrasted it with the right to be free from cruel and unusual punishment under the Eighth Amendment. Id. at 534; see also id. at 535 n.16 (explaining that the Second Circuit “properly relied on the Due Process Clause rather than the Eighth Amendment in considering the claims of pretrial detainees” because “[d]ue process requires that a pretrial detainee not be punished” but “[a] sentenced inmate . . . may be punished, although that punishment may not be ‘cruel and unusual’ under the Eighth Amendment”).

Next, in Farmer, the Court held that to establish a deliberate-indifference claim under the Eighth Amendment, a plaintiff must show that the prison “official knows of and disregards an excessive risk to inmate health or safety.” 511 U.S. at 837. The Court explicitly tied that test to the text of the Eighth Amendment. Id. It explained that the criminal-recklessness standard “comports best with the text of the [Eighth] Amendment” since “an official’s failure to alleviate a significant risk that he should have perceived but did not . . . cannot . . . be condemned as the infliction of punishment” in violation of the Eighth Amendment’s prohibition on “cruel and unusual punishments.” Id. at 837–38.

The Court provided further guidance on the difference between claims brought under the Eighth and Fourteenth Amendments in Kingsley. There, the Court held that to establish an excessive-force claim under the Fourteenth Amendment, a pretrial detainee “must show only that the force purposely or knowingly used against him was objectively unreasonable.” Kingsley, 576 U.S. at 396–97. The Court explained that “there are, in a sense, two separate state-of-mind questions” in such cases. Id. at 395. “The first concerns the defendant’s state of mind with respect to his physical acts—i.e., . . . with respect to the bringing about of certain physical consequences in the world. The second . . . concerns the defendant’s state of mind with respect to whether his use of force was ‘excessive.’” Id.¹⁰10. In an effort to show that the Court explicitly limited Kingsley to the excessive-force context, the majority quotes the Court’s statement that it is “with respect to this question . . . that we hold that courts must use an objective standard.” Majority Op. at 23 (quoting Kingsley, 576 U.S. at 396). But, when read in context of the opinion, this statement was meant to make clear that the Court addressed only the second prong of the two-prong test. Kingsley, 576 U.S. at 395–96. The opinion’s context supports this reading because the statement immediately followed the Court’s discussion of the first prong of the analysis. Id. The majority nevertheless asserts that this reading is “exceedingly cramped.” Majority Op. at 23. If the majority believes that reading the Court’s statements in context is erroneous, then I am willing to commit such an “error.” As to the first, the Court explained that “the defendant must possess a purposeful, a knowing, or possibly a reckless state of mind” because “liability for negligently inflicted harm is categorically beneath the threshold of constitutional due process.” Id. at 396 (quoting County of Sacramento v. Lewis, 523 U.S. 833, 849 (1998)). “Thus, if an officer’s Taser goes off by accident or if an officer unintentionally trips and falls on a detainee, . . . the pretrial detainee cannot prevail on an excessive force claim” because they will not satisfy the first prong. Id.

But, as to the second state-of-mind requirement, the Court held that defendants can violate the Fourteenth Amendment without any showing of intent, even though intent to cause harm is required for excessive-force claims under the Eighth Amendment. Id. at 397–98. This conclusion was “consistent with [the Court’s] precedent,” such as Bell. Id. at 397. The Court explained that “Bell’s focus on ‘punishment’ does not mean that proof of intent (or motive) to punish is required for a pretrial detainee to prevail on a claim that his due process rights were violated.” Id. at 398. Instead, “as Bell itself shows . . . , a pretrial detainee can prevail by providing only objective evidence that the challenged governmental action is not rationally related to a legitimate governmental objective or that it is excessive in relation to that purpose.” Id.

The Court rejected the argument that it should rely on the standard from cases “brought by convicted prisoners under the Eighth Amendment’s Cruel and Unusual Punishment Clause” to craft the test for the Fourteenth Amendment’s Due Process Clause. Id. at 400. The Court viewed those cases as inapposite because “[t]he language of the two Clauses differs . . . [a]nd, most importantly, pretrial detainees (unlike convicted prisoners) cannot be punished at all, much less ‘maliciously and sadistically.’” Id. (quoting Ingraham v. Wright, 430 U.S. 651, 671–72, 671 n.40 (1977)). Therefore, in Fourteenth Amendment cases, “there is no need . . . , as there might be in an Eighth Amendment case, to determine when punishment is unconstitutional.” Id. at 401.

Kingsley’s rejection of Eighth Amendment precedent made clear that the Court “disapprove[s] [of] the uncritical extension of Eighth Amendment jurisprudence to” the Fourteenth Amendment’s Due Process Clause. Miranda v. County of Lake, 900 F.3d 335, 351 (7th Cir. 2018). Kingsley “signal[ed] that courts must pay careful attention to the different status of pretrial detainees” when crafting a standard for liability because it “cautioned that the Eighth Amendment and Due Process analyses are not coextensive.” Id. at 352. Kingsley also made “plain that punishment has no place in defining the mens rea element of a pretrial detainee’s claim under the Due Process Clause.” Darnell v. Pineiro, 849 F.3d 17, 35 (2d Cir. 2017). Instead, courts may consider whether the government action is rationally related to a legitimate government objective, among other inquiries, to determine if the Due Process Clause was violated. Kingsley, 576 U.S. at 398.

Kingsley also used “broad wording” and general reasoning that was not confined to excessive force. Castro, 833 F.3d at 1070; accord Short v. Hartman, 87 F.4th 593, 609 (4th Cir. 2023). The Court spoke “broad[ly]” of “challenged governmental action,” rather than merely excessive force, in explaining its holding. Castro, 833 F.3d at 1070 (quoting Kingsley, 576 U.S. at 398); accord Short, 87 F.4th at 609. For example, the Court stated that “a pretrial detainee can prevail by providing only objective evidence that the challenged governmental action is not rationally related to a legitimate governmental objective or that it is excessive in relation to that purpose.” Kingsley, 576 U.S. at 398. And rather than exclusively drawing upon excessive-force cases, the Court cited Bell and other cases that evaluated various conditions imposed on pretrial detainees: Block v. Rutherford, 468 U.S. 576 (1984) (denial of contact visitation and shakedown searches); Schall v. Martin, 467 U.S. 253 (1984) (juvenile-pretrial-detention scheme); and United States v. Salerno, 481 U.S. 739 (1987) (Bail Reform Act of 1984). Kingsley, 576 U.S. at 398.

Thus, “Kingsley did not decree on a whim that we must use an objective [reasonableness] test for excessive force claims.” Short, 87 F.4th at 610. Instead, Kingsley reached that conclusion “because ‘[the Court’s] precedent’ (above all, Bell) already recognizes that a pretrial detainee may state a due process claim against ‘a variety of prison conditions’ by an ‘objective standard’” that does not inquire into the defendant’s subjective knowledge or intent to punish. Id. (quoting Kingsley, 576 U.S. at 397–98). The majority believes that Kingsley’s failure to discuss Farmer is “telling” that Kingsley was confined to excessive-force claims. Majority Op. at 22. But I see no reason why the Court would have cited Farmer in Kingsley. Farmer is an Eighth Amendment case, and the Kingsley Court—addressing a Fourteenth Amendment claim— explicitly declined to rely on Eighth Amendment cases because “[t]he language of the two Clauses differs . . . [a]nd, most importantly, pretrial detainees (unlike convicted prisoners) cannot be punished at all, much less ‘maliciously and sadistically.’” Kingsley, 576 U.S. at 400 (quoting Ingraham, 430 U.S. at 671–72, 671 n.40).

Kingsley’s reasoning is therefore directly applicable to Fourteenth Amendment deliberate-indifference claims. In Fourteenth Amendment excessive-force cases and deliberate-indifference cases, “[t]he underlying federal right, as well as the nature of the harm suffered, is the same . . . . Both categories of claims arise under the Fourteenth Amendment’s Due Process Clause, rather than under the Eighth Amendment’s Cruel and Unusual Punishment Clause.” Castro, 833 F.3d at 1069–70. Both claims derive from the Due Process Clause’s prohibition on deprivations of liberty without due process. Bell, 441 U.S. at 535; Graham v. Connor, 490 U.S. 386, 395 n.10 (1989); Youngberg, 457 U.S. at 315, 320–22. Additionally, “[e]xcessive force applied directly by an individual jailer and force applied by a fellow inmate can cause the same injuries, both physical and constitutional,” and “[j]ailers have a duty to protect pretrial detainees from violence at the hands of other [incarcerated persons], just as they have a duty to use only appropriate force themselves.” Castro, 833 F.3d at 1070. Indeed, we have recognized that it can be difficult to distinguish between an excessive-force claim and a conditions-of-confinement claim because a single set of facts could serve as the basis of both claims, further suggesting that the source of the right, rather than the type of harm claimed, should govern the analysis. Moderwell v. Cuyahoga County, 997 F.3d 653, 662–63 (6th Cir. 2021) (noting that “claims of excessive force do not necessarily require allegations of assault, but rather can consist of the physical structure and conditions of the place of detention” and approvingly noting that the “[p]laintiff’s deliberate indifference claims . . . rely on the same factual predicate as the excessive force claims” (citation modified)). There is simply “nothing in the logic the Supreme Court used in Kingsley that would support . . . dissection of the different types of claims that arise under the Fourteenth Amendment’s Due Process Clause.” Miranda, 900 F.3d at 352.¹¹11. The majority objects to this because under the Eighth Amendment, excessive-force claims and deliberate-indifference claims are evaluated under different tests. Majority Op. at 23, 26. The difference between the Eighth Amendment tests for those claims, however, is that excessive-force claims require a higher mens rea (malicious and sadistic use of force) than deliberate-indifference claims (knowledge of the risk). Whitley v. Albers, 475 U.S. 312, 320–21 (1986); Farmer, 511 U.S. at 837. And from that premise, the majority suggests that excessive-force and deliberate-indifference claims under the Fourteenth Amendment should likewise apply different tests, Majority Op. at 23, 26, to mirror the approach under the Eighth Amendment, presumably with a higher mens- rea requirement for excessive-force claims than deliberate-indifference claims. That is not possible in the Fourteenth Amendment setting, however, because the test for excessive-force claims under Kingsley hovers above the standard for negligence by requiring an intentional act. Kingsley, 576 U.S. at 395–96. Lowering the Kingsley standard for deliberate-indifference claims by eliminating the intentional-act requirement would therefore create liability for mere negligence, which is prohibited under the Due Process Clause. Id. Therefore, it is not possible to lower the Kinglsey standard by removing the intentional-act requirement for deliberate-indifference claims.

Additionally, like in Kingsley, the civil-recklessness standard is consistent with related bodies of jurisprudence. Kingsley, 576 U.S. at 399 (noting that the objective-reasonableness “standard is also consistent with” excessive-force claims brought under the Fourth Amendment by plaintiffs not convicted of a crime). For example, the Court analyzes conditions-of- confinement claims, of which deliberate-indifference claims are a sub-species, Wilson v. Seiter, 501 U.S. 294, 302–04 (1991), under a test that does not consider the defendant’s intent, Bell, 441 U.S. at 535–43. The Court has also applied a standard that does not consider the defendant’s intent to determine whether the Bail Reform Act and a juvenile-detention scheme violated due process. Salerno, 481 U.S. at 747–48; Schall, 467 U.S. at 269–71. The Court did the same when it evaluated the lack of safety to which a civilly committed plaintiff was subjected. Youngberg, 457 U.S. at 320–23. In fact, at oral argument when the County’s counsel was asked if he could identify any due-process claims that involve a similar inquiry into the defendant’s subjective knowledge of the risk or intent to punish, he could not. See Oral Arg. at 29:20–30:47.

The majority responds that, in fact, Kingsley contains a subjective test—but that inquiry (which the Brawner framework retained) asks only whether the physical action taken was intentional rather than accidental, not whether the defendant had subjective knowledge of the risk or intent to punish. Kingsley, 576 U.S. at 395–96. Obviously, that someone acts intentionally does not mean that they subjectively know of the risk or intend to punish—so Kingsley does not solve this problem. And the majority’s use of subjective to describe the intentional-act requirement is part of majority’s clouding of the meaning of those terms. It is directly at odds with the majority’s contention that the civil-recklessness standard (which also includes an intentional-act requirement) is purely objective and the majority’s description of the intentional- act requirement elsewhere as objective. The majority cannot have it both ways and treat objective and subjective as linguistic chameleons that change meaning as needed to support its arguments—either the intentional-act requirement is subjective or it is objective.

The majority also contends that unlike excessive force, where “parallel protection” from the Fourth Amendment mandates an objective-reasonableness inquiry, “the parallel provision for conditions-of-confinement deliberate-indifference claims, the Eighth Amendment, dictates more than an objective inquiry.” Majority Op. at 21. But Kingsley rejected that exact reasoning. The Court acknowledged and explicitly refused to rely on the standard for “excessive force claims brought by convicted prisoners under the Eighth Amendment’s Cruel and Unusual Punishment Clause” because of the inherent differences between that Clause and the Fourteenth Amendment. Kingsley, 576 U.S. at 400.

My conclusion that the civil-recklessness standard applies to Fourteenth Amendment claims is consistent with the Second, Fourth, Seventh, and Ninth Circuits. Short, 87 F.4th at 611; Miranda, 900 F.3d at 353–54; Darnell, 849 F.3d at 34–35; Castro, 833 F.3d at 1068–73. On the other side of the scale, only the Tenth Circuit has reached the opposite conclusion on the merits. Strain v. Regalado, 977 F.3d 984, 991 (10th Cir. 2020). The majority cites cases from the Fifth, Eighth, and Eleventh Circuits to make the split appear more even. Majority Op. at 6. But those cases have, at most, a few sentences of reasoning and held (in footnotes) only that Kingsley “did not abrogate” their published precedent because it addressed “a different type of constitutional claim.” Cope v. Cogdill, 3 F.4th 198, 207 n.7 (5th Cir. 2021); see Whitney v. City of St. Louis, 887 F.3d 857, 860 n.4 (8th Cir. 2018); Dang ex rel. Dang v. Sheriff, Seminole Cnty., 871 F.3d 1272, 1279 n.2 (11th Cir. 2017). Given our en banc posture, those cases are of little help.

In the end, “Kingsley is clear”: there are important differences between the Eighth and Fourteenth Amendments. Short, 87 F.4th at 608–09. Critically, the texts of the Amendments differ, and as a result “there is no need [under the Fourteenth Amendment], as there might be in an Eighth Amendment case, to determine when punishment is unconstitutional.” Kingsley, 576 U.S. at 401. To reflect that, the standards for liability under the Eighth and Fourteenth Amendments must be different. Short, 87 F.4th at 609. Therefore, applying the civil- recklessness standard is “[t]he only way to respect the distinction Kingsley drew between the Eighth and Fourteenth Amendments.” Id. at 610.

C. Counter Arguments

The majority presents several arguments as to why the Farmer criminal-recklessness standard should control, but none of them persuade. First, the majority contends that the text of the Fourteenth Amendment’s Due Process Clause supports its conclusion that Farmer’s standard governs. Majority Op. at 11–12. Specifically, the majority reasons that the Clause’s use of “[t]he word ‘deprive’ signals a ‘deliberate decision[]’ by a government official.” Id. at 11 (quoting Daniels v. Williams, 474 U.S. 327, 331 (1986)). But that does not get the majority where it needs to go. Brawner required the defendant to act or fail to act intentionally, so it too complies with the deliberate-decision requirement. 14 F.4th at 596–97. After all, in common usage, one can deliberately “deprive” another of something without knowing the risk of that decision to deprive. The majority overlooks this distinction and instead makes an unexplained, and unsupported, logical leap from the fact that the defendant must deprive a person of their liberty to requiring the defendant to be subjectively aware that they are depriving a person of their liberty. Majority Op. at 27.

Poynter challenges the County’s deliberate decisions here: its decision (based on its custom) to classify detained persons based only on their charges and behavior during arrest, omitting consideration of institutional histories of violence, and its decision (based on its custom) to place detained persons with histories of violence in general-population cells where they could attack other detained persons. True, the County itself did not attack Poynter or even want Guess and Wix to attack him. But the majority’s criminal-recklessness test does not require that either—so that must not matter. What work does the deliberate-decision requirement do then? It means that mere negligence does not violate the Due Process Clause. Daniels, 474 U.S. at 330– 31 (holding that because “the word ‘deprive’ in the Due Process Clause connote[s] more than a negligent act,” the Due Process Clause is not violated by a “mere lack of due care by a state official”). All agree on this point, so the majority’s reliance on “deliberate” runs aground.

The majority next asserts that “Bell . . . counseled that conditions-of-confinement claims should not turn on the status of the prisoner.” Majority Op. at 24. But, as already explained, that is wrong. True, Bell rejected the Second Circuit’s reliance on the presumption of innocence as the “source of the detainee’s substantive right,” but that did not mean that the status of pretrial detainees had no relevance in determining what conditions they could be subjected to. 441 U.S. at 532–33. The Court explained that pretrial detainees’ rights derived from their right to be free from any punishment under the Fourteenth Amendment’s Due Process Clause and then contrasted that with the right to be free from cruel and unusual punishment under the Eighth Amendment. Id. at 534; 535 n.16. There would have been no reason for the Court to contrast the rights in this way if a detained person’s status was irrelevant.

Additionally, that view cannot be reconciled with subsequent caselaw that treats conditions-of-confinement claims differently depending on the plaintiff’s status as a pretrial detainee or convicted prisoner. In Wilson the Supreme Court considered the appropriate mens- rea requirement for Eighth Amendment conditions-of-confinement claims. 501 U.S. at 296. Instead of relying on Bell, the Court concluded that the deliberate-indifference test from Estelle v. Gamble, 429 U.S. 97 (1976), an Eighth Amendment case, was the appropriate standard. Id. at 303. If conditions-of-confinement claims do not turn on the status of the detained person, Majority Op. at 24, then surely the Court in Wilson would have applied the Bell standard to the Eighth Amendment. We have also evaluated conditions-of-confinement claims differently depending on whether they arise under the Eighth or Fourteenth Amendments, which of course turns on the status of the detained person. Under the Eighth Amendment, we require plaintiffs to prove that the defendants knew that the condition created a sufficiently serious risk of harm. Pearson v. Michigan Dep’t of Corr., 170 F.4th 1027, 1041 (6th Cir. 2026); Rafferty v. Trumbull County, 915 F.3d 1087, 1095 (6th Cir. 2019); Flanory v. Bonn, 604 F.3d 249, 254–55 (6th Cir. 2010). In contrast, when we evaluate a conditions-of-confinement claim under the Fourteenth Amendment, we apply Bell’s test and do not require the defendant to subjectively intend to punish or know of the risk. J.H. v. Williamson County, 951 F.3d 709, 717–19 (6th Cir. 2020). How this could be true if “Bell itself counseled that conditions-of-confinement claims should not turn on the status of the prisoner,” Majority Op. at 24, is anyone’s guess.

The majority also asserts that my reading of Kingsley and the civil-recklessness test is in tension with “the jurisprudential trend” moving away from “recognizing new” substantive-due- process rights. Majority Op. at 22. That is a bit of a head scratcher. The civil-recklessness test does not create a new substantive-due-process right. In fact, all agree that the Supreme Court has already recognized that pretrial detainees have a right to be free from punishment. Id. at 14. The question here is only what test detained persons must satisfy to show that the right was violated. “[T]here is a sharp distinction between the existence of a . . . right . . . and how to vindicate that right.” City of Sherrill v. Oneida Indian Nation of N.Y., 544 U.S. 197, 213 (2005) (first alteration in original) (quoting Oneida Indian Nation of N.Y. v. County of Oneida, 199 F.R.D. 61, 90 (N.D.N.Y. 2000)).

In response, the majority asserts that the “bounds” of a right turn on “the applicable test for determining whether the right has been infringed.” Majority Op. at 22–23. For starters, the majority has it backwards. Rights help inform tests, not the other way around. As the Court has clearly explained, when defining the elements of a § 1983 claim, courts begin by “‘identify[ing] the specific constitutional right’ at issue.” Manuel v. City of Joliet, 580 U.S. 357, 370 (2017) (quoting Albright v. Oliver, 510 U.S. 266, 271 (1994)). The right, in turn, informs the relevant test. Id. (explaining that “courts must closely attend to the values and purposes of the constitutional right at issue” when defining the elements of a § 1983 claim). The Court, moreover, has consistently distinguished the right undergirding a § 1983 claim from the claim’s elements in this fashion. See, e.g., id.; Chiaverini v. City of Napoleon, 602 U.S. 556, 561–62 (2024) (recognizing that § 1983 claims are “of course premised on a constitutional violation,” but distinguishing the right at issue from the claim’s “elements and rules,” which “may also be shaped by common-law tort principles” “to the extent” that they accord with the right’s “values and purposes” (citation modified)); Thompson v. Clark, 596 U.S. 36, 43 (2022) (same). We have drawn this very distinction, too. See, e.g., King v. Harwood, 852 F.3d 568, 580 n.4 (6th Cir. 2017). Consequently, recognizing that the civil-recklessness standard applies to deliberate- indifference claims vivified by the Due Process Clause does not entail the creation of a new right.

The sources cited by the majority do not demonstrate that rights are coterminous with tests, such that altering a test necessarily alters the associated right. Straining to support its position, the majority cites County of Sacramento v. Lewis, a law review article, and Dobbs v. Jackson Women’s Health Organization, 597 U.S. 215 (2022). To begin, it is altogether unclear how Lewis supports the majority. Majority Op. at 23 (citing Lewis, 523 U.S. at 852–54). The majority claims that Lewis applies different “frameworks to substantive-due-process claims arising from different factual scenarios despite the fact that the underlying right is the same.” Id. (emphasis added). It cannot both be true that the Court can alter tests while keeping “the underlying right . . . the same” and that “easing the manner in which one may prove a deliberate- indifference substantive-due process claim naturally expands our understanding of the right protected.” Majority Op. at 23. How the majority reconciles these Janus-faced propositions is anyone’s guess.

The law review article that the majority cites, moreover, does not suggest that tests shape the contents of rights. In fact, it explicitly recognizes not only that rights and tests are distinct, but also that courts can create tests that over- or under-enforce rights. Richard H. Fallon, Jr., Foreword: Implementing the Constitution, 111 Harv. L. Rev. 54, 57, 60 (1997) (articulating “the thesis that a gap frequently, often necessarily exists between the meaning of constitutional norms and the tests by which those norms are implemented); id. at 61 (explaining that the relationship between rights and tests “frequently is not and sometimes could not be one of identity”); id. at 63 (describing how the Court has crafted prophylactic tests to “ensure ‘breathing space’ for First Amendment freedoms” (citation modified)).

And finally, the majority’s account of Dobbs is, again, backwards. The Court did not alter the status of the right to elect an abortion by subjecting abortion restrictions to rational-basis review: it applied rational-basis review to an abortion restriction because it held that the right to elect an abortion is not fundamental. Dobbs, 597 U.S. at 240, 300–02. To the extent that the majority intends to claim that the right at issue drives the applicable test (here, the elements of a Fourteenth Amendment deliberate-indifference claim brought pursuant to § 1983), I agree. But the majority simply does not respect the profound differences between the Eighth and Fourteenth Amendments. Indeed, the majority next contends that because the right at issue in Fourteenth Amendment deliberate-indifference cases is the right to be free from punishment, we must apply the Farmer test to determine whether jail officials punished the detained person. Majority Op. at 12–14, 20–21. According to the majority, for a jail official to punish a detained person, they must intend to punish them. And because deliberate-indifference claims involve inaction, the argument goes, unless courts require the plaintiff to prove that the defendant knew about the substantial risk the plaintiff faced, courts cannot determine whether the defendant had an intent to punish. There are several issues with this argument.

The first is that it is hard to square with the majority’s admonition that we must start with the text of the Constitution. Majority Op. at 11. Farmer adopted the criminal-recklessness standard for Eighth Amendment claims because it “comports best with the text of the [Eighth] Amendment,” which prohibits “cruel and unusual ‘punishments.’” 511 U.S. at 837. But the text of the Fourteenth Amendment’s Due Process Clause, unlike the Eighth Amendment, does not prohibit “cruel and unusual” punishment. Instead, as the majority agrees, the Fourteenth Amendment’s prohibition on deprivations of “liberty” without due process prevents jail officials from punishing detained persons at all, not only cruelly and unusually.¹²12. I acknowledge that “punishment” is not in the text of the Fourteenth Amendment, but, as the majority agrees, the Supreme Court has read the Clause’s prohibition on deprivations of liberty without due process to encompass the right to be free from punishment. See Majority Op. at 12. See Kingsley, 576 U.S. at 400 (“The language of the [Eighth and Fourteenth Amendment] Clauses differs . . . [a]nd, most importantly, pretrial detainees (unlike convicted prisoners) cannot be punished at all. . . .”). Despite this, the majority believes that Farmer’s Eighth Amendment test applies to Fourteenth Amendment claims because it is necessary to show that a detained person was punished. But that cannot be right. Applying the Farmer test to both the Fourteenth and Eighth Amendments either reads “cruel and unusual” out of the Eighth Amendment or reads “cruel and unusual” into the Fourteenth Amendment.

The majority responds that under Farmer, “subjective knowledge . . . informs whether punishment occurred at all, not whether the punishment was cruel and unusual.” Majority Op. at 26. As I explain, that is erroneous. But even accepting that as true for the sake of argument, the majority still fails to clarify what aspect of the Farmer test isolates liability for only cruel and unusual punishment, rather than merely punishment. It cannot be the unjustifiably high risk of harm the plaintiff faces because the majority retains that aspect of the Farmer test for claims under the Fourteenth Amendment, which prohibits any punishment. I am therefore still left to wonder how the majority accounts for the total absence of the phrase “cruel and unusual” in the Due Process Clause.¹³13. In response, the majority retreats to language from Farmer, which it claims defines what can and “cannot . . . be condemned as the infliction of punishment.” Majority Op. at 26. In other words, the majority reads this language into the Fourteenth Amendment context because the Court used the word “punishment,” not the phrase “cruel and unusual punishment.” But this reading commits the cardinal sin of “pars[ing]” “[t]he language of an opinion . . . as though we were dealing with language of a statute.” Nat’l Pork Producers Council v. Ross, 598 U.S. 356, 373 (2023) (quoting Reiter v. Sonotone Corp., 442 U.S. 330, 341 (1979)). The Court has “emphasize[d],” however, that its “opinions dispose of discrete cases and controversies and they must be read with a careful eye to context.” Id. at 373–74. “The language [the majority] highlight[s] appeared in a particular context and did particular work.” Id. at 374. Farmer was an Eighth Amendment case, and defined the term punishment in light of the fact that the Amendment “outlaws cruel and unusual ‘punishments.’” 511 U.S. at 837.

The second issue with the majority’s argument is that the Court has held that a plaintiff can establish that they were punished in violation of the Due Process Clause without showing an intent to punish. Kingsley, 576 U.S. at 398–99. In evaluating conditions-of-confinement claims, the Court has applied an “objective standard” and looked to “objective evidence,” not “the prison officials’ subjective beliefs about the policy” or intent to punish, to determine if a condition violated the Due Process Clause. Id. at 398; see Bell, 441 U.S. at 541–43. Indeed, the Court has told us that “Bell’s focus on ‘punishment’ does not mean that proof of intent (or motive) to punish is required for a pretrial detainee to prevail on a claim that his due process rights were violated.” Kingsley, 576 U.S. at 398.

Even so, the majority asserts that Kingsley’s citation of Daniels proves that even Kingsley retained a subjective-intent requirement. Majority Op. at 25. According to the majority, Daniels held that a pretrial detainee was required to plead that jail officials left a pillow “on the stairs with the intention of harming him.” Id. (quoting Daniels, 474 U.S. at 334). This is a somewhat baffling misreading of Daniels. There, the plaintiff argued in favor of a negligence standard and claimed that a higher mens rea requirement would prove unworkable, as he could have simply “alleged that the pillow was left on the stairs with the intention of harming him.” Daniels, 474 U.S. at 334. The Court rejected that argument and, in doing so, never held that the plaintiff had to plead that jail officials intended to harm him. To the contrary, the Court explicitly noted that: “[d]espite his claim about what he might have pleaded, [the plaintiff] concedes that [the defendant] was at most negligent. Accordingly, this case affords us no occasion to consider whether something less than intentional conduct, such as recklessness or ‘gross negligence,’ is enough to trigger the protections of the Due Process Clause.” Id. at 334 n.3.

Returning to Poynter, Kingsley’s discussion of Bell matters because “the protection [an incarcerated person] is afforded against other inmates” and “the medical care a prisoner receives” are “‘condition[s]’ of his confinement,” and thus the same standard that governs conditions-of-confinement claims applies to Poynter’s claim that the County’s classification customs placed him at a substantial risk of harm from Guess and Wix. Wilson, 501 U.S. at 303. Consequently, Bell’s holding that plaintiffs asserting conditions-of-confinement claims can establish that the government violated due process without proving the defendant had a subjective intent to punish should apply equally to plaintiffs asserting deliberate-indifference claims. That makes especially good sense here because Poynter challenges the County’s classification custom itself—its custom of classifying and housing detained persons by considering only their charges and behavior during booking, not their institutional histories. There is no reason to think that, if the policy of double bunking in Bell can be evaluated without looking to the defendant’s subjective intent to punish, Poynter’s challenge to the County’s classification custom cannot be. It is hard to see why one could not infer an intent to punish from a classification custom but could infer such an intent from a policy of double bunking in Bell.

The majority’s response is that deliberate-indifference claims involve inaction whereas conditions-of-confinement claims involve affirmative action. Majority Op. at 21. But, again, Poynter challenges the County’s classification custom itself, even though he challenges it as insufficient because it failed to consider institutional history. The decision to consider certain factors in the classification analysis but not others is undoubtedly an intentional one. Additionally, the line between action and inaction can be blurry and a function of characterization. Koulta v. Merciez, 477 F.3d 442, 445–46 (6th Cir. 2007) (discussing how, because “it is sometimes difficult to distinguish action from inaction,” we ask “whether [the victim] was safer before the state action than he was after it,” “[r]ather than focusing on the often metaphysical question of whether officer behavior amounts to affirmative conduct or not,” to determine if there was a violation of due process (first alteration in original) (quoting Cartwright v. City of Marine City, 336 F.3d 487, 493 (6th Cir. 2003))).

Even if one construes Poynter’s claim as exclusively about inaction, Kingsley gave no indication that its reasoning would not apply equally to inaction when the defendant has a duty to act. And contrary to the majority’s assertion, Majority Op. at 20, Kingsley’s reasoning was not premised on the idea that punitive intent could be inferred from the affirmative act of applying force that is excessive. The Court’s opinion never once mentioned that the application of force allows for such an inference. See Kingsley, 576 U.S. at 391–404. And if an intent to punish can be inferred from the application of force alone in excessive-force cases, then why would the Eighth Amendment inquiry also require a showing “that officials applied force ‘maliciously and sadistically for the very purpose of causing harm?’” Farmer, 511 U.S. at 835 (quoting Hudson v. McMillian, 503 U.S. 1, 6 (1992)).

Even assuming that Kingsley was premised on the idea that punitive intent could be inferred from the application of force, Bell makes clear that such an inference can be drawn from conditions of confinement. 441 U.S. at 535–39. I see no reason why punitive intent cannot be inferred from Brawner’s civil-recklessness standard if it can be inferred from conditions of confinement. Civil recklessness exists when a person “acts or (if the person has a duty to act) fails to act in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.” Farmer, 511 U.S. at 836. Our caselaw applying this standard required that the defendants be exposed to the facts which create an unjustifiably high risk of harm, because we asked whether a reasonable person in the defendant’s position would have known of the risk. See Greene v. Crawford County, 22 F.4th 593, 609–14 (6th Cir. 2022); Poynter, 162 F.4th at 680 (discussing how the County “had all the relevant information about Guess’s and Wix’s violent pasts”). Therefore, if an official was exposed to the facts which create an unjustifiably high risk of harm, it is equally fair to infer an intent to punish as it is to infer that intent from imposing a condition of confinement that is not “rationally related to a legitimate nonpunitive governmental purpose.” Kingsley, 576 U.S. at 398 (quoting Bell, 441 U.S. at 561). There is, after all, “no legitimate nonpunitive goal . . . served by” placing a detained person at an unjustifiably high risk of harm. Swink v. S. Health Partners Inc., 160 F.4th 438, 451 (4th Cir. 2025) (quoting Short, 87 F.4th at 606), petition for cert. filed, (U.S. Mar. 13, 2026) (No. 25-1091).

The majority contends that under Bell we can infer punitive intent from the implementation of a condition of confinement that does not advance a legitimate government interest because there is no other reason to impose the condition. Majority Op. at 24. But a prison official may implement a condition for various reasons even if a court eventually determines that it does not advance a legitimate government interest. For one, the official could themself subjectively believe that the condition did in fact advance a legitimate interest even though a court disagrees. Or the official could implement the condition because it makes their job easier with their limited resources even though a court eventually determines that the negative tradeoffs of the choice mean that it does not actually advance a legitimate government interest. Under Bell, we can infer punitive intent from conditions of confinement without ruling out other potential explanations such as those. Therefore, the inference from a condition of confinement sheds no more light on punitive intent than the civil-recklessness standard. Punitive intent might not be the only inference one can draw from an official subjecting a detained person to an unjustifiably high risk of harm even though the official was exposed to the information that would inform a reasonable person of the risk, Farmer, 511 U.S. at 836; Greene, 22 F.4th at 609– 14, but it is certainly one of the compelling inferences that can be drawn. Bell teaches that is all that is required.

The third issue with the majority’s argument that punitive intent is necessary to establish that Poynter was punished in violation of the Fourteenth Amendment is that it overlooks that Poynter had a right to personal security as a civil detainee, which the Court established in Youngberg. In Youngberg, the plaintiff was civilly detained in a state institution. 457 U.S. at 309. He “was injured on numerous occasions” by himself and others and brought suit alleging that, among other things, the defendants’ failure to prevent such injuries violated the Fourteenth Amendment’s Due Process Clause. Id. at 310. The Court held that the plaintiff had “a constitutionally protected liberty interest in safety” and that “the right to personal security constitutes a ‘historic liberty interest’ protected substantively by the Due Process Clause,” which “is not extinguished by lawful confinement, even for penal purposes” or by “involuntary commitment proceedings.” Id. at 315–16 (quoting Ingraham, 430 U.S. at 673).

Poynter, too, was detained under civil authority. Just like the plaintiff in Youngberg, Poynter had his liberty restrained in a state facility following civil proceedings that were not a part of the criminal-justice process. Therefore, Poynter’s Due Process rights derive not only from his right to be free from punishment, but also from his right to physical safety while detained by the state pursuant to civil proceedings, as recognized by the Supreme Court in Youngberg. See Benjamin v. Fraser, 343 F.3d 35, 50 (2d Cir. 2003), overruled on other grounds by Caiozzo v. Koreman, 581 F.3d 63 (2d Cir. 2009). The majority contends that Youngberg’s viability is “questionable” because it has been undermined by subsequent cases, including Farmer.¹⁴14. Farmer is, of course, an Eighth Amendment case, so it is unclear how it undermined Youngberg, which is about the Fourteenth Amendment’s Due Process Clause. Majority Op. at 19. Even assuming that is true, the Supreme Court has reminded us time and time again that “lower courts [are] bound by even [the Court’s] crumbling precedents.” Loper Bright Enters. v. Raimondo, 603 U.S. 369, 406 (2024); see also Bosse v. Oklahoma, 580 U.S. 1, 3 (2016) (per curiam) (“Our decisions remain binding precedent until we see fit to reconsider them, regardless of whether subsequent cases have raised doubts about their continuing vitality.” (quoting Hohn v. United States, 524 U.S. 236, 252–53 (1998))).

The majority next asserts that the civil-recklessness test is purely objective and therefore inconsistent with Kingsley, which contained both a subjective and an objective component because it required the defendant to have intentionally applied physical force. Majority Op. at 21. But Brawner, and the original panel (contra the Majority’s assertion, Majority Op. at 21), retained the intentional-act requirement. Brawner, 14 F.4th at 596 (“A defendant must have not only acted deliberately (not accidentally), but also recklessly ‘in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.’” (quoting Farmer, 511 U.S. at 836)); Poynter, 162 F.4th at 678–79 (discussing and applying the intentional-action element). Here, there is no contention that the County unintentionally failed to classify and separate Guess and Wix based on their institutional histories of violence. Evidence supports that the County did so pursuant to its ubiquitous custom of classifying detained persons as violent or nonviolent based on only their charges and behavior during booking, without considering their institutional histories. Additionally, the County does not argue, and there is no evidence to support, that Poynter, Guess, and Wix were accidentally, rather than intentionally, placed in the same general-population cell. Yet again, this “objective” versus “subjective” argument is part of the majority’s larger obfuscation of the meaning of those terms—which shift depending on what definition would best serve the majority’s argument. Either the intentional-act requirement is “subjective” or it is “objective.”

The majority also purports to find support for its position from the word “deliberate” in the phrase “deliberate indifference,” asserting that “[i]t would be strange to say that a defendant chooses deliberately to be indifferent to risks of which he is unaware.” Majority Op. at 21. But Farmer explicitly rejected that reasoning. 511 U.S. at 840 (“Our decision that Eighth Amendment liability requires consciousness of a risk is thus based on the Constitution and our cases, not merely on a parsing of the phrase ‘deliberate indifference’ . . . [because] on the crucial point (whether a prison official must know of a risk, or whether it suffices that he should know) the term does not speak with certainty. Use of ‘deliberate’ . . . arguably requires nothing more than an act (or omission) of indifference to a serious risk that is voluntary, not accidental. . . . And even if ‘deliberate’ is better read as implying knowledge of a risk, the concept of constructive knowledge is familiar enough that the term ‘deliberate indifference’ would not, of its own force, preclude a scheme that conclusively presumed awareness from a risk’s obviousness.”).

The majority next contends that the civil-recklessness standard creates liability for mere negligence, which is below the threshold of constitutional protection. Majority Op. at 13–14, 26–27. It is certainly true that negligent conduct does not violate the Due Process Clause. Kingsley, 576 U.S. at 396. The majority’s argument, therefore, is based on its belief that there is no difference between recklessness and negligence. That would be news to the Supreme Court and to the common law. In Farmer, the Supreme Court recognized that civil recklessness required more culpable conduct than mere negligence. The Court explained that “deliberate indifference l[ay] somewhere between the poles of negligence at one end and purpose or knowledge at the other” and the middle ground could be “fair[ly]” described as “recklessly disregarding that risk.” Farmer, 511 U.S. at 836. That conclusion, however, “d[id] not . . . fully answer the pending question about the level of culpability deliberate indifference entails” because there were two potential definitions of recklessness that fell within the continuum from negligence to purposefulness.¹⁵15. The Court explained that “gross negligence” was distinct from, and a higher standard of liability than, mere negligence because it fell “[b]etween the poles” of negligence and purposefulness. Farmer, 511 U.S. at 836 n.4. The Court also equated gross negligence to civil recklessness. Id. The majority seizes on that equation to argue that civil recklessness is akin to mere negligence. Majority Op. at 14. But that holds no water because, as the Court acknowledged, gross negligence is a higher standard of liability than mere negligence. Id. One of those definitions was the civil-recklessness standard—which “calls a person reckless who acts or (if the person has a duty to act) fails to act in the face of an unjustifiably high risk of harm that is either known or so obvious that it should be known.” Id. (first citing W. Keeton et al., Prosser and Keeton on Law of Torts § 34, 213–14 (5th ed. 1984); and then citing Restatement (Second) of Torts § 500 (A.L.I 1965)).¹⁶16. As discussed, the Court rejected that standard not because it was below the constitutional threshold for liability, but because of the text of the Eighth Amendment. That is the precise standard that the majority now opines is equivalent to negligence, but the majority is incorrect. Put another way, civil recklessness requires intentional action but not subjective knowledge of the consequences of those actions, whereas negligence does not require intentional action at all. See Kemp v. Fulton County, 27 F.4th 491, 497 (7th Cir. 2022) (explaining that civil recklessness requires “the defendant officer [to] intend to carry out a certain course of actions; negligence is not enough”); Pittman ex rel. Hamilton v. Madison County, 108 F.4th 561, 570–71 (7th Cir. 2024) (discussing how in Kingsley, “the Court distinguished between intentional acts . . . that can lead to liability, and negligent [or accidental] acts . . . that cannot” and that, in light of that, “in Kemp, it was enough to show that the defendant intentionally chose not to wear his hearing aid on the day of the fight, even if he did not appreciate the risk of harm from that choice” (citation modified)), reh’g denied, No. 23-2301, 2024 WL 3889635 (7th Cir. Aug. 21, 2024), cert. denied, 145 S. Ct. 1154 (2025).

The common law has likewise long recognized the distinction between negligence and recklessness. See Hart v. Hillsdale County, 973 F.3d 627, 637 (6th Cir. 2020). Under the common law, “[f]or recklessness, the risk must be ‘substantially greater than that which is necessary to make [a defendant’s] conduct negligent.’” Hehrer v. County of Clinton, 161 F.4th 955, 964 (6th Cir. 2025) (second alteration in original) (quoting Restatement (Second) of Torts § 500 (A.L.I. 1965)). “That is, the ‘difference’ in the risk must be ‘so marked as to amount substantially to a difference in kind’ rather than degree.” Id. (quoting Restatement (Second) of Torts § 500 cmt. g (A.L.I. 1965); see also Safeco Ins. Co. of Am. v. Burr, 551 U.S. 47, 69 (2007) (recognizing that “[i]t is this high risk of harm, objectively assessed, that is the essence of recklessness at common law,” which must be “substantially greater than that which is necessary to make his conduct negligent” (quoting Restatement (Second) of Torts § 500 (A.L.I. 1963– 1964))). Clearly “[r]ecklessness is a lower bar than intent, but a higher bar than negligence,” so the civil-recklessness standard does not impose liability for mere negligence. Short, 87 F.4th at 611.

Finally, the majority argues that the intentional-act requirement does not impose any meaningful limitation on liability because “virtually every plaintiff can point to some intentional choice by the defendant that preceded the plaintiff’s injury.” Majority Op. at 27. According to the majority, the civil-recklessness test would impose liability for intentional actions such as booking Poynter into BCDC or the decision to detain him in BCDC instead of another jail. Id. at 27–28. The majority’s parade of “endless” examples of intentional decisions that could purportedly give rise to lability is easily addressed by the traditional proximate-causation principles that we apply in this context. Jones v. Kent County, 115 F.4th 504, 513–15 (6th Cir. 2024); Rhinehart v. Scutt, 894 F.3d 721, 747 (6th Cir. 2018); Parsons v. Caruso, 491 F. App’x 597, 604 (6th Cir. 2012).

The majority claims that my advocacy for proximate-causation principles is evidence of my “fluid methods.” Majority Op. at 28. Yet the Court has instructed that “[c]ommon-law principles . . . guide . . . the definition of § 1983 claims”—specifically, “the common law of torts.” Manuel, 580 U.S. at 370. “For centuries, it has been ‘a well established principle of [the common] law, that in all cases of loss, we are to attribute it to the proximate cause, and not to any remote cause.’” Lexmark Int’l, Inc. v. Static Control Components, Inc., 572 U.S. 118, 132 (2014) (alteration in original) (quoting Waters v. Merchants’ Louisville Ins. Co., 36 U.S. 213, 223 (1837)). Unsurprisingly, the Court has acknowledged that “the familiar proximate cause standard” plays a role in the § 1983 context. County of Los Angeles v. Mendez, 581 U.S. 420, 430–31 (2017). Our court has recognized this anodyne principle as well. See, e.g., Zimmermann v. Labish, No. 25-1734, ___ F.4th ____, 2026 WL 2529819, at *5 (6th Cir. Aug. 27, 2026) (recognizing that § 1983 claims are “undergird[ed] by “general common-law tort principles,” including “proximate cause”); Marvaso v. Sanchez, 971 F.3d 599, 606 (6th Cir. 2020) (“Like a tort plaintiff, a § 1983 plaintiff must establish both causation in fact and proximate causation.” (citation modified)); see also Westmoreland, 29 F.4th at 740 (6th Cir. 2022) (Bush, J., dissenting) (“[O]ur precedent for § 1983 claims requires proximate causation—not merely cause- in-fact.”) What is “fluid” about relying on binding precedent escapes me.

Instead, flowing directly past this precedent, the majority rests its argument on its policy preferences, contending that proximate causation is simply too confusing and too ineffective at curbing liability. Majority Op. at 28. These arguments, however, are built on sand. See id. (relying on comments to a Restatement provision and a nearly century-old dissent).¹⁷17. Even though the Restatement (Third) of Torts criticizes proximate cause principles, no less an authority than the Supreme Court has recognized that the very section of the Restatement that the majority relies on nevertheless articulates a proximate cause standard. Paroline v. United States, 572 U.S. 434, 445 (2014) (recognizing that Restatement (Third) of Torts: Liability for Physical Injury and Emotional Harm § 29 (A.L.I. 2010) articulates a proximate causation principle). Although “the principles of proximate causation” are sometimes confusing, “it is also true that proximate causation principles are generally thought to be a necessary limitation on liability.” Exxon Co., U.S.A. v. Sofec, Inc., 517 U.S. 830, 838 (1996). Indeed, proximate cause is such a “familiar doctrine,” Metro. Edison Co. v. People Against Nuclear Energy, 460 U.S. 766, 774 (1983), that claiming that it is too unworkable is akin to claiming that “our legal system . . . [is] a charade.” Ralph by Ralph v. Nagy, 950 F.2d 326, 328 (6th Cir. 1991). Here, applying those proximate cause principles shows that merely booking Poynter into BCDC or deciding to detain him there rather than another jail would, of course, not be sufficient to establish proximate cause because those actions are too far removed from the harm he suffered. In reality, the causation question here is simple. The County intentionally placed Guess, Wix, and Poynter in the same cell pursuant to its classification custom. In less than a minute and thirty seconds, Guess and Wix attacked Poynter. Had the County’s classification custom been sufficient, Poynter would not have been placed in the same cell as Wix and Guess and therefore would not have been attacked. R. 95-7 (Eiser Dep. at 96:5–15) (Page ID #565). Indeed, Poynter’s injuries “result[ed] from the risks that made” the County’s “conduct tortious” under the civil-recklessness standard. Restatement (Third) of Torts: Phys. & Emot. Harm § 29 (A.L.I. 2010).

III. CONCLUSION

For the foregoing reasons, I would adhere to Brawner and hold that plaintiffs bringing deliberate-indifference claims under the Fourteenth Amendment need to establish civil- recklessness. Because the majority casts that standard aside even though neither party asked us to do so, I respectfully dissent.