FOR PUBLICATION

UNITED STATES COURT OF APPEALS

FOR THE NINTH CIRCUIT

MITCHELL GARRAWAY, No. 23-15482

Plaintiff-Appellee, D.C. No. 1:17-cv-00533- v. ADA-GSA

JACQUILINE CIUFO; K. MILLER;

J. ZARAGOZA, OPINION

Defendants-Appellants.

Appeal from the United States District Court for the Eastern District of California Ana de Alba, District Judge, Presiding

Argued and Submitted May 8, 2024 Pasadena, California

Filed September 3, 2024

Before: Richard C. Tallman, Danielle J. Forrest, and Patrick J. Bumatay, Circuit Judges.

Opinion by Judge Tallman; Dissent by Judge Bumatay SUMMARY *

Bivens / Collateral Order Doctrine

In an action brought by a federal inmate against prison officials pursuant to Bivens v. Six Unknown Named Agents of Fed. Bureau of Narcotics, 403 U.S. 388 (1971), the panel dismissed defendants’ interlocutory appeal and, joining three other circuits, held that district court orders extending Bivens, absent a denial of qualified immunity, are not immediately appealable under the collateral order doctrine. Plaintiff alleged that prison officials were deliberately indifferent to his safety in violation of his Eighth Amendment rights. The district court denied defendants’ motion for judgment on the pleadings, in which defendants argued that that no Bivens remedy is available for failure to protect an inmate from a risk of prisoner violence. After the Supreme Court issued Egbert v. Boule, 596 U.S. 482 (2022), defendants filed a motion to reconsider, which the district court denied. Defendants filed an interlocutory appeal. The panel noted that the collateral order doctrine is a narrow exception to be strictly applied. Appellate courts may consider an underlying Bivens remedy when reviewing an interlocutory order denying qualified immunity—and may even consider it as a matter antecedent to qualified immunity. However, it does not necessarily follow that appellate courts can review on an interlocutory basis an order recognizing a Bivens remedy standing alone.

* This summary constitutes no part of the opinion of the court. It has been prepared by court staff for the convenience of the reader.

GARRAWAY V. CUIFO 3

The panel concluded that an order recognizing a Bivens remedy is not one of the small classes of collateral rulings that is “effectively unreviewable” upon final judgment. Delaying review does not threaten significant public interests. Any alleged improper judicial intrusion into the legislative function resulting from recognizing a Bivens cause of action can be effectively rectified upon review of a final judgment. And delaying review does not so imperil the efficiency and effectiveness of the Executive Branch as to warrant immediate review when the next logical step in this litigation is for the federal-officer defendant to claim qualified immunity, the denial of which on a question of law would be immediately appealable. Dissenting, Judge Bumatay wrote that under the collateral order doctrine federal appellate courts may consider interlocutory appeals whenever denying immediate review would imperil a substantial public interest. Because preserving the separation of powers is a value of the highest order and authorizing an improper Bivens suit erodes that value, the court had jurisdiction to immediately review the district court’s Bivens ruling.

COUNSEL

D. Dangaran (argued) and Samuel Weiss, Rights Behind Bars, Washington, D.C.; Mitchell Garraway, Pro Se, United States Penitentiary, Coleman, Florida; for Plaintiff- Appellee. Weili J. Shaw (argued) and Barbara L. Herwig, Appellate Staff Attorney, Civil Division; Brian M. Boynton, Principal Deputy Assistant Attorney General; United States Department of Justice, Washington, D.C.; Victoria L. Boesch, Assistant United States Attorney, Office of the United States Attorney, Sacramento, California; for Defendants-Appellants.

OPINION

TALLMAN, Circuit Judge:

Plaintiff-Appellee Mitchell Garraway brought this Bivens action against three prison officials alleging they were deliberately indifferent to his safety in violation of his Eighth Amendment rights while he was incarcerated at U.S. Penitentiary, Atwater, California. Defendants-Appellants prison officials filed this interlocutory appeal after the district court denied their motion for reconsideration of an earlier motion for judgment on the pleadings in which they argued no Bivens remedy exists for failure to protect an inmate from a risk of prisoner violence. For the reasons set forth herein, we dismiss for lack of jurisdiction. 1 I Mitchell Garraway is a federal inmate currently incarcerated in Coleman, Florida. In March 2016, while Garraway was housed in Atwater, he allegedly informed

1 Unlike the dissent, we do not reach the merits of the district court’s order recognizing a Bivens remedy for Eighth Amendment deliberate indifference to inmate safety under Farmer v. Brennan, 511 U.S. 825 (1994). We may not assume jurisdiction for the purpose of deciding the merits of the case. See Ex parte McCardle, 74 U.S. 506, 514 (1868) (“Without jurisdiction the court cannot proceed at all in any cause. Jurisdiction is power to declare the law, and when it ceases to exist, the only function remaining to the court is that of announcing the fact and dismissing the cause.”).

GARRAWAY V. CUIFO 5

three prison officials that his cellmate cut his nose with a razor and asked to be moved to another cell. Garraway says that although officials were aware his cellmate had a history of assaulting other inmates, all three officials refused his request. Approximately two weeks later, Garraway’s cellmate allegedly struck him on the left side of his jaw, causing it to swell and inflicting lasting pain. Garraway filed this Bivens action against the three officials, arguing that their refusal to change his cell given his documented safety concerns and the cellmate’s violent history constituted deliberate indifference to his safety in violation of the Eighth Amendment. Defendants-Appellants moved for judgment on the pleadings arguing that no Bivens remedy is available for failure to protect an inmate from a risk of prisoner violence. On February 21, 2020, the district court denied the motion on the basis that it agreed with Garraway’s argument that his case “does not differ in a meaningful way from Farmer, nor do [his] claims arise in a new [Bivens] context.” The case proceeded for two and a half years. During this time, though Defendants-Appellants’ Answer listed the defense of qualified immunity, they have not yet formally invoked qualified immunity in a dispositive motion. The Supreme Court issued its decision in Egbert v. Boule on June 8, 2022. 596 U.S. 482 (2022) (holding that Bivens does not extend to create causes of action for Fourth Amendment excessive-force claims and First Amendment retaliation claims). On August 24, 2022, this case was reassigned from District Judge Dale A. Drozd to then- District Judge Ana de Alba. 2 On December 8, 2022,

2 Judge de Alba was elevated to the United States Court of Appeals for the Ninth Circuit on November 15, 2023. Defendants-Appellants filed a motion to reconsider the denial of their previous motion for judgment on the pleadings in light of Egbert. Judge de Alba denied Defendants-Appellants’ motion for reconsideration on February 1, 2023, reiterating that Farmer controlled, determining that Farmer remained intact after Egbert, and reasoning that Egbert “does not mention Farmer.” On March 31, 2023, Defendants-Appellants timely filed a notice of interlocutory appeal of the order denying their motion for reconsideration. In his papers and at oral argument, Garraway argued that we lacked jurisdiction over this appeal on two alternative grounds. First, he insists that we lack jurisdiction over the interlocutory appeal of a district court order recognizing a Bivens cause of action, untethered from a denial of qualified immunity. Second, he argues that under Hanson v. Shubert, 968 F.3d 1014 (9th Cir. 2020), the denial of a motion for reconsideration is not an appealable final order standing alone. Today we join three sister circuits in holding that district court orders extending Bivens, absent a denial of qualified immunity, are not immediately appealable under the collateral order doctrine. 3 II In all matters, the threshold question is one of jurisdiction. See Steel Co. v. Citizens for a Better Env’t, 523 U.S. 83, 95 (1995) (“Every federal appellate court has a special obligation to satisfy itself . . . of its own jurisdiction.” (cleaned up)); In re Martinez, 721 F.2d 262, 264 (9th Cir. 1983) (holding that federal courts have both the

3 Accordingly, we decline to answer whether we have jurisdiction over the denial of a motion to reconsider an order denying judgment on the pleadings and recognizing a Bivens remedy where we lack jurisdiction over the underlying order. See Hanson, 968 F.3d at 1018.

GARRAWAY V. CUIFO 7

inherent authority and the responsibility to consider their own jurisdiction). Garraway argues that we lack jurisdiction over an interlocutory appeal of a district court ruling recognizing a constitutional damages remedy under Bivens v. Six Unknown Named Agents of Fed. Bureau of Narcotics, 403 U.S. 388 (1971), absent an ancillary denial of qualified immunity.4 Though a matter of first impression for our Court, three of our sister circuits have declined to give collateral order treatment to this class of orders. See Himmelreich v. Fed. Bureau of Prisons, 5 F.4th 653, 662–63 (6th Cir. 2021); Graber v. Doe, 59 F.4th 603, 608 (3d Cir. 2023), cert. denied sub nom. Boresky v. Graber, 144 S. Ct. 681 (2024); Mohamed v. Jones, 100 F.4th 1214, 1215 (10th Cir. 2024). 5 By statute, U.S. Courts of Appeals “have jurisdiction of appeals from all final decisions of the district courts of the United States.” 28 U.S.C. § 1291 (emphasis added). Though a final decision is generally one “by which a district court disassociates itself from a case,” Swint v. Chambers Cnty. Comm’n, 514 U.S. 35, 42 (1995), the Supreme Court “has long given” the statute a “practical rather than a technical construction,” Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541, 546 (1949). As such, § 1291 has been

4 As noted supra, at p. 5, Defendants-Appellants’ Answer lists the affirmative defense of qualified immunity, however, immunity has not yet been formally invoked in a dispositive motion or ruled upon. Accordingly, the question of qualified immunity is neither properly before us nor available to provide a jurisdictional hook to evaluate the Bivens question. 5 The Eleventh Circuit is currently considering this same issue as a matter of first impression. Fleming v. FCI Tallahassee Warden, No. 23-10252 (11th Cir. argued Aug. 15, 2024). interpreted to “encompass[] not only judgments that ‘terminate an action,’ but also a ‘small class’ of collateral rulings that, although they do not end the litigation, are appropriately deemed ‘final.’” Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 106 (2009) (quoting Cohen, 337 U.S. at 545–46). As the Court established in Cohen and restated in Swint, “[t]hat small category includes only decisions that are conclusive, that resolve important questions separate from the merits, and that are effectively unreviewable on appeal from the final judgment in the underlying action.” Swint, 514 U.S. at 42 (citing Cohen, 337 U.S. at 546). On multiple occasions, the Supreme Court has admonished that the collateral order doctrine is a “narrow exception,” Firestone Tire & Rubber Co. v. Risjord, 449 U.S. 368, 374 (1981), to be “strictly applied,” Richardson- Merrell, Inc. v. Koller, 472 U.S. 424, 431 (1985). Out of “healthy respect for the virtues of the final-judgment rule,” and with an eye toward “efficient judicial administration,” Mohawk, 558 U.S. at 106 (quoting Firestone, 449 U.S. at 374), the doctrine must “never be allowed to swallow the general rule that a party is entitled to a single appeal, to be deferred until final judgment has been entered,” Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868 (1994) (citation omitted). Thus, “[t]he justification for immediate appeal must . . . be sufficiently strong to overcome the usual benefits of deferring appeal until litigation concludes.” Mohawk, 558 U.S. at 107. Specifically, the third Cohen factor inquiry— whether an order is “effectively unreviewable”—“cannot be answered without a judgment about the value of the interests that would be lost through rigorous application of a final judgment requirement.” Digit. Equip., 511 U.S. at 878–79.

GARRAWAY V. CUIFO 9

Accordingly, “the decisive consideration is whether delaying review until the entry of final judgment ‘would imperil a substantial public interest’ or ‘some particular value of a high order.’” Mohawk, 558 U.S. at 107 (quoting Will v. Hallock, 546 U.S. 345, 352–53 (2006)). Defendants-Appellants argue that “[t]he courts of appeals routinely consider the existence of Bivens remedies on immediate appeal from interlocutory district court decisions.” E.g., Wilkie v. Robbins, 551 U.S. 537, 549 n.4 (2007); Mejia v. Miller, 61 F.4th 663, 669 (9th Cir. 2023); Pettibone v. Russell, 59 F.4th 449, 452–54 (9th Cir. 2023); Ioane v. Hodges, 939 F.3d 945, 949, 951 (9th Cir. 2018); Martin v. Naval Crim. Investigative Serv., 539 F. App’x 830, 831–33 (9th Cir. 2013). They assert that “[i]n many of these appeals, defendants also challenge the district court’s denial of qualified immunity.” In fact, in every one of the cases Defendants-Appellants cite, defendants challenged the district court’s denial of qualified immunity. It is well-established that a district court’s denial of qualified immunity is a collateral order subject to immediate appeal “to the extent that it turns on an issue of law.” Mitchell v. Forsyth, 472 U.S. 511, 530 (1985). Further, “Wilkie establishes that, in an interlocutory appeal from a denial of qualified immunity, we necessarily have jurisdiction to decide whether an underlying Bivens cause of action exists.” Pettibone, 59 F.4th at 453. 6 In some

6 In Wilkie, the Court did not explicitly identify a theory on which jurisdiction was based. Pettibone, 59 F.4th at 453 (“In explaining why there was appellate jurisdiction to decide whether a Bivens cause of action existed, the [Wilkie] Court did not apply the pendent appellate jurisdiction test . . . [i]nstead, the Court said, without elaboration, that instances, upon review of denials of qualified immunity, courts have made the analytical choice to consider the Bivens remedy prior to considering qualified immunity. Compare, e.g., Wilkie, 551 U.S. at 549–62 (concluding no Bivens remedy existed and thus declining to reach qualified immunity), with Wood v. Moss, 572 U.S. 744, 757 (2014) (assuming without deciding that Bivens extends to First Amendment claims to reach the qualified immunity question). Appellate courts may consider the underlying Bivens remedy when reviewing an interlocutory order denying qualified immunity—and may even consider it as a matter antecedent to qualified immunity. However, it does not necessarily follow that appellate courts can review on an interlocutory basis an order recognizing a Bivens remedy standing alone. We cannot assume that which is necessary for our review to be proper—a jurisdictional hook. We must decide in the first instance whether the district court’s denial of Defendants-Appellants’ motion for reconsideration of that court’s earlier denial of Defendants- Appellants’ motion for judgment on the pleadings independently satisfies Cohen’s three requirements. See Swint, 514 U.S. at 40. Because the order fails to satisfy the third Cohen requirement—that the order be “effectively unreviewable on appeal from a final judgment”—we need not decide whether the other two prongs are met. See Lauro Lines S.R.L. v. Chasser, 490 U.S. 495, 498 (1989) (quoting Richardson-Merrell Inc. v. Koller, 472 U.S. 424, 431 (1985)) (explaining that “we need not decide” whether

the recognition of the underlying Bivens cause of action was ‘directly implicated by the defense of qualified immunity and properly before us on interlocutory appeal.’” (internal citations omitted)).

GARRAWAY V. CUIFO 11

the order meets the other prongs when it “fail[s] to satisfy the third requirement of the collateral order test”). A Defendants-Appellants argue that “such orders are effectively unreviewable after trial because delaying review threatens significant public interests that lie at the heart of the Supreme Court’s Bivens doctrine—protecting the separation of powers, and protecting the efficiency and effectiveness of the Executive Branch.” These interests are certainly implicated when any Bivens-style claim is brought and adjudicated. First, consideration of separation-of- powers principles are “central” to the analysis required of courts considering whether to fashion a Bivens remedy. Ziglar v. Abbasi, 582 U.S. 120, 135 (2017) (“When a party seeks to assert an implied cause of action under the Constitution itself . . . separation-of-powers principles are or should be central to the analysis. The question is ‘who should decide’ whether to provide for a damages remedy, Congress or the courts? The answer most often will be Congress.” (citation omitted)). Second, the recognition of a Bivens cause of action acts as an inherent limitation on the Executive Branch. See Corr. Servs. Corp. v. Malesko, 534 U.S. 61, 70 (2001) (“The purpose of Bivens is to deter individual federal officers from committing constitutional violations.”). But the question is not simply whether these interests are implicated by the Bivens inquiry or whether those interests are sufficiently weighty, but rather, whether delaying review of a district court order recognizing a Bivens remedy “so imperils th[ose] interest[s]” as to justify immediate appeal of that entire class of orders. Mohawk, 558 U.S. at 108 (“The crucial question, however, is not whether an interest is important in the abstract; it is whether deferring review until final judgment so imperils the interest as to justify the cost of allowing immediate appeal of the entire class of relevant orders.”). Stated differently, it is both necessary “that the right asserted be one that is essentially destroyed if its vindication must be postponed until trial is completed,” Chasser, 490 U.S. at 499, and that the right be “sufficiently important to overcome the policies militating against interlocutory appeals,” id. at 503 (Scalia, J., concurring). Neither condition, standing alone, is sufficient. See id. at 502–03 (recognizing that while the right to be sued only in a particular forum is “positively destroyed” by allowing trial to proceed in another jurisdiction, reversal after trial is “vindication enough because the right is not sufficiently important”); Mohawk, 558 U.S. at 108–09 (“acknowledg[ing] the importance of the attorney-client privilege,” but concluding “postjudgment appeals generally suffice to protect the rights of litigants and ensure the vitality of the attorney-client privilege”). While we recognize the relative importance of the interests implicated, we do not see what irreparable harm would occur, or what interest, right, or entitlement “would be lost through rigorous application of a final judgment requirement” such that this class of orders should be considered “effectively unreviewable.” Digit. Equip., 511 U.S. at 878–79 (emphasis added). 1 Regarding Defendants-Appellants’ concern for protecting separation-of-powers principles, improper judicial intrusion into the legislative function can be effectively rectified upon review of a final judgment, without immediate and irreparable harm being done to our system of governance as a result of the delay. In coming to this conclusion, we mean not to undermine the Supreme

GARRAWAY V. CUIFO 13

Court’s admonitions of the dangers of judicially created implied causes of action under the Constitution, but rather, we simply disagree with the argument that these admonitions amount to an immediate, concrete harm justifying interlocutory appeal. 7 Bivens and the doctrine it has spawned is not new—the Courts of Appeals and the Supreme Court have routinely, for many decades now, declined to extend Bivens to new contexts. See, e.g., Schweiker v. Chilicky, 487 U.S. 412, 414 (1988); Minneci v. Pollard, 565 U.S. 118, 120 (2012); Hernandez v. Mesa, 589 U.S. 93, 113–14 (2020); Adams v. Johnson, 355 F.3d 1179, 1181 (9th Cir. 2004); Pettibone, 59 F.4th at 457. In all these cases, the appellate process had adequate opportunity to review the Bivens determination without any irreparable harm being done to the separation of powers. Defendants-Appellants—and indeed, the dissent—

7 To further illustrate this point, the case(s) in which the Supreme Court has identified “honoring the separation of powers” as a “particular value of a high order” sufficient to warrant immediate review are inapposite. The class of orders in question in those cases—the denial of absolute Presidential immunity from suit—posed an immediate and tangible threat to the separation of powers. See Will, 546 U.S. at 352 (“Thus, in Nixon . . . we stressed the ‘compelling public ends,’ ‘rooted in . . . the separation of powers,’ that would be compromised by failing to allow immediate appeal of a denial of absolute Presidential immunity.” (citations omitted)); Nixon v. Fitzgerald, 457 U.S. 731, 743 (1982) (“In light of the special solicitude due to claims alleging a threatened breach of essential Presidential prerogatives under the separation of powers, we conclude that petitioner did present a ‘serious and unsettled’ and therefore appealable question to the Court of Appeals.” (citation omitted)). Denial of absolute Presidential immunity is the destruction of a right not to be sued, irreparable upon final judgment. The recognition of a Bivens remedy at the pleadings stage does not even resemble the denial of a claim as of right that is essentially destroyed if its vindication is postponed until final judgment. See infra § 3. fail to identify what immediate harm would be inflicted upon the separation of powers that cannot be effectively reviewed and remedied upon final judgment. Defendants-Appellants rely entirely on the conclusory assertion that “[a]n erroneous district court decision recognizing a Bivens remedy causes ‘immediate’ harm to separation-of-powers interests that are ‘essential.’” (Citing to Graber v. Doe II, 59 F.4th 603, 616 (3d Cir. 2023) (Hardiman, J., dissenting)). The dissent similarly relies upon dissents from other judges in other circuits proclaiming a “parade of horribles,” and ignoring that we are joining the majority opinions of the three other circuits who have addressed this issue before us. We choose to follow Supreme Court precedent, join the prevailing majority voices of our sister circuits, and decline to create the circuit split urged upon us. 2 Defendants-Appellants’ next argument, that orders recognizing a Bivens remedy threaten the efficiency and effectiveness of the Executive Branch, presents a more tangible harm, which, at first blush, has an immediacy to it. See Egbert, 596 U.S. at 499 (“Recognizing any new Bivens action ‘entail[s] substantial social costs, including the risk that fear of personal monetary liability and harassing litigation will unduly inhibit officials in the discharge of their duties.’” (alteration in original) (quoting Anderson v. Creighton, 483 U.S. 635, 638 (1987))); Ziglar, 582 U.S. at 134 (cautioning against the expansion of implied causes of actions under Bivens because “[c]laims against federal officials often create substantial costs, in the form of defense and indemnification . . . [and] the time and administrative costs attendant upon intrusions resulting from the discovery and trial process”). This argument has cogency, especially given, as Defendants-Appellants themselves recognize,

GARRAWAY V. CUIFO 15

“[t]hese concerns mirror those that prompted the Supreme Court to conclude that qualified-immunity rulings are collateral final orders.” But Defendants-Appellants’ attempt to justify extending the collateral order doctrine to orders recognizing Bivens remedies untethered from a denial of qualified immunity by comparing that class of orders to qualified immunity proves too much. Delaying review of an order recognizing a Bivens remedy does not so imperil the efficiency and effectiveness of the Executive Branch as to warrant immediate review when the next logical step in this litigation—and indeed, in nearly any Bivens action—is for the federal-officer defendant to claim qualified immunity, the denial of which on a question of law would be immediately appealable. See Mitchell, 472 U.S. at 525, 528–30 (“A major characteristic of the denial or granting of a claim appealable under Cohen’s ‘collateral order’ doctrine is that ‘unless it can be reviewed before [the proceedings terminate], it can never be reviewed at all.’” (alteration in original) (quoting Stack v. Boyle, 342

U.S. 1, 12 (1952))).

As the Court recognized in Mohawk, “[a]s long as the class of claims, taken as a whole, can be adequately vindicated by other means, the chance that the litigation at hand might be speeded, or a particular injustice averted, does not provide a basis for jurisdiction under § 1291.” 558 U.S. at 107 (cleaned up) (holding that disclosure orders adverse to the attorney-client privilege did not qualify for immediate appeal under the collateral order doctrine in part because “[p]ostjudgment appeals, together with other review mechanisms, suffice to protect the rights of litigants and preserve the vitality of the attorney-client privilege”). Protecting the efficiency and effectiveness of the Executive Branch is the foundation on which qualified immunity is grounded. As the Supreme Court stated in Mitchell:

The conception animating the qualified immunity doctrine . . . is that where an official’s duties legitimately require action in which clearly established rights are not implicated, the public interest may be better served by action taken with independence and without fear of consequences . . . [which is] not limited to liability for money damages; they also include the general costs of subjecting officials to the risks of trial— distraction of officials from their governmental duties, inhibition of discretionary action, and deterrence of able people from public service . . . even such pretrial matters as discovery are to be avoided if possible, as [i]nquiries of this kind can be peculiarly disruptive of effective government.

472 U.S. at 525–26 (second alteration in original) (internal citations and quotation marks omitted). Qualified immunity presents an alternative means by which the efficiency and effectiveness of the Executive Branch is protected. And because the denial of qualified immunity on issues of law is immediately appealable, whereby the underlying Bivens remedy itself is reviewable, any residual argument that qualified immunity fails to protect the effectiveness and efficiency of the Executive

GARRAWAY V. CUIFO 17

Branch from the threats of adverse rulings recognizing Bivens remedies loses much of its force. As such, we conclude that qualified immunity presents an alternative, if imperfect, means by which the potential harms posited can be vindicated. 3 Upon closer inspection, fitting the assertion that the “recognition of a Bivens remedy poses an immediate and irreparable threat to the efficiency and effectiveness of the Executive Branch” into the Cohen framework leads to absurd results and contravenes Supreme Court precedent. Insisting that an adverse Bivens ruling be subject to immediate appeal on the basis that government officials might be spared the burdens of litigation sooner essentially amounts to an assertion that government officials have a right to avoid litigation altogether—separate and apart from this right embodied by qualified immunity—that is essentially destroyed by an adverse Bivens ruling. However, that a ruling “may burden litigants in ways that are only imperfectly reparable by appellate reversal of a final district court judgment . . . has never sufficed” to independently justify collateral order treatment. Digit. Equip., 511 U.S. at 872. This is true even when the burdened litigant is a government actor or official. See Will, 546 U.S. at 354. As the Supreme Court announced in Will, if the government’s interest in avoiding litigation alone sufficed as justification for an interlocutory appeal, then “28 U.S.C. § 1291 would fade out whenever the Government or an official lost an early round that could have stopped the fight.” Id. Indeed, the logic of Will extends to the class of orders to which we are asked to give Cohen treatment today. In Will, the Supreme Court considered whether a district court order rejecting the Federal Tort Claims Act judgment bar as a defense to a Bivens claim was appealable on an interlocutory basis under the collateral order doctrine. Id. at 353–54. In holding it was not, the Court reasoned that in declining to construe the “claim of the customs agents in this case . . . as an immunity demanding the protection of a collateral order appeal,” the government did not have an absolute right to avoid trial. Id. Otherwise,

if simply abbreviating litigation troublesome to Government employees were important enough for Cohen treatment, collateral order appeal would be a matter of right whenever the Government lost a motion to dismiss under the Tort Claims Act, or a federal officer lost one on a Bivens action, or a state official was in that position in a case under 42 U.S.C. § 1983, or Ex parte Young.

Id. The dissent suggests that we read this language in Will to mean that because “an adverse Bivens action does not implicate the right not to stand trial . . . there’s no need for immediate appeal.” But we understand the Supreme Court to be making a more nuanced point. Where there is no right permanently destroyed, or harm irreparably done, simply abbreviating litigation burdensome to government officials does not suffice as justification for Cohen treatment. Bivens liability exists, albeit in a severely cabined form, not for the purpose of protecting the separation of powers or the efficiency and effectiveness of the Executive Branch, but in spite of those considerations—a tension the Supreme Court has repeatedly recognized. For example, in Egbert,

GARRAWAY V. CUIFO 19

the Court encapsulated the purpose of Bivens as “concerned solely with deterring the unconstitutional acts of individual officers—i.e., the focus is whether the Government has put in place safeguards to prevent constitutional violations from recurring.” 596 U.S. at 498 (cleaned up) (discussing alternative remedy processes). Importantly, Egbert also made plain that, given the stressors on the separation of powers, “recognizing a cause of action under Bivens is ‘a disfavored judicial activity.’” Id. at 491 (quoting Ziglar, 582 U.S. at 135). An adverse Bivens decision does not represent the denial of a government official’s right not to stand trial, as in the case of qualified immunity, or any denial of an affirmative right. Federal officials are harmed by an adverse Bivens ruling only insofar as they are required to litigate qualified immunity in a dispositive motion, the denial of which is immediately appealable. To elevate an adverse Bivens ruling to the class of orders immediately appealable would be to equate it with the denial of a federal officer’s right not to stand trial, in contravention of the spirit of Will. See Will, 546 U.S. at 353–54. As the Sixth Circuit trenchantly observed:

Unlike qualified immunity, Bivens provides a plaintiff’s remedy for unconstitutional conduct. It does not grant defendants an entitlement not to stand trial. To the extent that defendants are concerned about litigating meritless cases, qualified immunity more than adequately protects government officials from the burdens of litigation.

Himmelreich, 5 F.4th at 662. We agree. Because an order recognizing a Bivens remedy is not “effectively unreviewable” upon final judgment, we decline the invitation to extend the collateral order doctrine to allow for the immediate appeal of that class of claims. B Those who would oppose our holding today voice concern about implied causes of action as undermining the separation of powers, while in the same breath suggesting we expand the collateral order doctrine, a judicially created exception to 28 U.S.C. § 1291. Expansion of the class of collaterally appealable orders “has acquired special force in recent years with the enactment of legislation designating rulemaking, ‘not expansion by court decision,’ as the preferred means for determining whether and when prejudgment orders should be immediately appealable.” Mohawk, 558 U.S. at 113 (quoting Swint, 514 U.S. at 48). Congress has statutorily authorized the Supreme Court to adopt rules “defin[ing] when a ruling of a district court is final for the purposes of appeal under section 1291,” 28 U.S.C. § 2072(c), and empowered the Supreme Court to “prescribe rules, in accordance with [§ 2072], to provide for an appeal of an interlocutory decision to the courts of appeals that is not otherwise provided for under [§ 1292],” § 1292(e). The rule we announce today, if wrong, can be properly revisited via the rulemaking process “with the opportunity for full airing it provides.” Mohawk, 558 U.S. at 114.

III

The dissent urges that “we ought to use our power under the . . . collateral order doctrine” to reach the merits. With the utmost respect for our dissenting colleague’s views, our power to address the merits is synonymous with our

GARRAWAY V. CUIFO 21

jurisdiction, mandated by Congress and limited by the Constitution. We are not looking, as the dissent suggests, for any way out of applying the collateral order doctrine, but, rather, we are faithfully applying Supreme Court precedent disfavoring expanding the doctrine’s limited exception to § 1291’s final judgment rule. And we are unwilling to bend that precedent to the breaking point. DISMISSED for lack of jurisdiction.

BUMATAY, Circuit Judge, dissenting:

This case requires us to determine whether federal government officers may immediately appeal a district court’s recognition of a new Bivens claim—even though the Supreme Court has all but put us out of the business of creating new causes of action under that “zombi[fied]” doctrine. See Mohamed v. Jones, 100 F.4th 1214, 1240 (10th Cir. 2024) (Tymkovich, J., dissenting). The answer is “yes.” In this case, a federal prisoner, Mitchell Garraway, sued three federal prison officials under Bivens v. Six Unknown Named Agents of Federal Bureau of Narcotics, 403 U.S. 388 (1971), for the alleged violation of his Eighth Amendment right. On behalf of the officials, the government quickly moved for judgment on the pleadings, arguing that no Bivens remedy was available for Garraway’s novel context. The district court denied the motion. After the Supreme Court decided Egbert v. Boule, 596 U.S. 482 (2022), the government moved for reconsideration of the district court’s order. Once again, the district court denied the motion and allowed the Bivens claim to proceed. The government immediately appealed. The Supreme Court has been clear—no more freelanced Bivens claims. Simply, judicially created causes of action offend the “separation of legislative and judicial power.” Egbert, 586 U.S. at 491 (simplified). Outside of those already recognized by the Court, federal courts may not seize the legislative mantle to create a new Bivens remedy if there’s a single reason to oppose it. And the Court left us instructions: “[b]ecause recognizing a Bivens cause of action is an extraordinary act that places great stress on the separation of powers,” we have “a concomitant responsibility to evaluate any grounds that counsel against Bivens relief.” Id. at 497 n.3 (simplified) (emphasis added). Given this special “responsibility,” we ought to use our power under the well-established collateral order doctrine to foreclose the undue expansion of Bivens liability. Under that doctrine, federal appellate courts may consider interlocutory appeals whenever denying immediate review would imperil a “substantial public interest.” See Will v. Hallock, 546 U.S. 345, 349, 353 (2006). Because preserving the separation of powers is a value of the highest order and authorizing an improper Bivens suit erodes that value, we have jurisdiction to immediately review the district court’s Bivens ruling. And the posture of this appeal—from a denial of a motion to reconsider— doesn’t change our jurisdiction. See Hanson v. Shubert, 968 F.3d 1014, 1019 n.4 (9th Cir. 2020). Finally, on review of the Bivens order, we should have easily reversed because the district court inappropriately expanded the scope of Bivens. Instead, the majority declines jurisdiction and permits a wrong-headed Bivens action to continue. I thus respectfully dissent.

GARRAWAY V. CUIFO 23

I. Collateral Order Doctrine The courts of appeals “have jurisdiction of appeals from all final decisions of the district courts of the United States.” 28 U.S.C. § 1291. As the Supreme Court has stated, “[f]rom the very foundation of our judicial system, the general rule has been that the whole case and every matter in controversy in it must be decided in a single appeal.” Microsoft Corp. v. Baker, 582 U.S. 23, 36 (2017) (simplified); see also Bank of Columbia v. Sweeny, 26 U.S. 567, 569 (1828) (“If this motion could now prevail, it would be a plain evasion of the provision of the Act of Congress, that final judgments only should be brought before this Court for re-examination.”). This principle, commonly referred to as the final-judgment rule, “descend[ed] from the Judiciary Act of 1789, where the First Congress established the principle that only final judgments and decrees of the federal district courts may be reviewed on appeal.” Cunningham v. Hamilton Cnty., Ohio, 527 U.S. 198, 203 (1999) (simplified). But the Supreme Court has said that appellate courts also have jurisdiction over “a small class of collateral rulings that, although they do not end the litigation, are appropriately deemed final.” Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 106 (2009) (simplified). These immediate appeals are known as interlocutory appeals under the “collateral order doctrine.” Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 867 (1994). To satisfy the collateral order doctrine, a ruling must “[1] conclusively determine the disputed question, [2] resolve an important issue completely separate from the merits of the action, and [3] be effectively unreviewable on appeal from a final judgment.” Puerto Rico Aqueduct & Sewer Auth. v. Metcalf & Eddy, Inc., 506 U.S. 139, 144 (1993) (simplified). The first two prongs of this test are easily satisfied here. First, the district court’s Bivens ruling was not “tentative, informal[,] or incomplete.” Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541, 546 (1949). After denying reconsideration, the Bivens issue was “concluded,” “closed,” and “final” for this litigation. Id. Second, the Bivens issue resolves a crucial separation-of-powers issue apart from “the correctness of [either party’s] version of the facts.” Mitchell v. Forsyth, 472 U.S. 511, 528 (1985). So it’s separate from the merits of Garraway’s claim. The only question then is whether the Bivens ruling is “effectively unreviewable” on appeal from the eventual final judgment here. I turn to that next. A. Bivens Claims Are Effectively Unreviewable To be sure, the canonical example of an order “effectively unreviewable” on appeal is the denial of qualified immunity. That’s because the purpose of immunity is “an immunity from suit” which is “effectively lost if a case is erroneously permitted to go to trial.” Id. at 512 (emphasis removed). But immunity doctrines are not the only types of issues that are effectively unreviewable on appeal. See, e.g., Cohen, 337 U.S. at 541 (order denying applicability of state law to a stockholder’s derivative suit), Stack v. Boyle, 342 U.S. 1 (1951) (order denying a motion to reduce bail); United States v. Baker, 603 F.2d 759, 761–62 (9th Cir. 1979) (order requiring the federal government to pay the defendant’s attorney’s fees for deposition expenses); Gough v.

GARRAWAY V. CUIFO 25

Perkowski, 694 F.2d 1140 (9th Cir. 1982) (order disqualifying a party’s counsel in a civil case); Wiggins v. Alameda Cnty., 717 F.2d 466, 467–68 (9th Cir. 1983) (order requiring state prison officials to pay expenses associated with producing and guarding a state prisoner); Hunt v. Imperial Merch. Servs., Inc., 560 F.3d 1137 (9th Cir. 2009) (order imposing the costs of class notification on one party); Copeland v. Ryan, 852 F.3d 900, 904 (9th Cir. 2017) (order requiring a state prison official to reimburse petitioner for deposition expenses); see also Graber v. Doe II, 59 F.4th 603, 612 (3rd Cir. 2023) (Hardiman, J., dissenting) (collecting other cases). So “an immunity is neither sufficient nor necessary for an order denying a claim to be ‘effectively unreviewable on appeal.’” Graber, 59 F.4th at 612 (Hardiman, J., dissenting). Indeed, the Supreme Court has suggested that when it comes to whether an order is “effectively unreviewable on appeal,” the touchstone is not immunity, but whether appellate delay threatens “a substantial public interest.” Will, 546 U.S. at 353. Thus, rather than merely look to whether immunity is invoked, we must consider “the value of the interests that would be lost through rigorous application of a final judgment requirement,” Digit. Equip., 511 U.S. at 878– 79, or “whether delaying review . . . would imperil a substantial public interest or some particular value of a high order.” Mohawk Indus., 558 U.S. at 107 (simplified). So interlocutory appeal may be appropriate even when the “right to be free of all proceedings whatsoever” is not implicated. Will, 546 U.S. at 352. After all, the driving force of the “effectively unreviewable” prong is not immunity in itself but rather the interest “root[ed]” in the protection of the “separation of powers.” Id. (quoting Nixon v. Fitzgerald, 457 U.S. 731, 749 (1982)). In addressing the need for the collateral order doctrine, the Supreme Court “spoke of the threatened disruption of governmental functions, and fear of inhibiting able people from exercising discretion in public service if a full trial were threatened.” Id. And first among the “value[s] of a high order” supporting interlocutory appeal was “honoring the separation of powers” and “preserving the efficiency of government and the initiative of its officials.” Id. Although explained in the context of qualified immunity, these considerations overlap perfectly with the concerns for expansive Bivens liability. 1. Separation of Powers and Executive Branch Concerns Like qualified immunity, creating new Bivens causes of action imperils the separation of powers and hamstrings executive branch officials from performing their constitutional duties. So while not a formal immunity doctrine, see, e.g., Graber, 59 F.4th at 603, the caution in extending Bivens is rooted in the same concerns and should be entitled to the same immediate review. First, being subject to a proceeding in violation of the “separation of powers” presents a “here-and-now injury.” Axon Enter., Inc. v. FTC, 598 U.S. 175, 191 (2023) (simplified); see also Seila Law LLC v. CFPB, 591 U.S. 197, 212 (2020) (When “a provision violates the separation of powers it inflicts a ‘here-and-now’ injury on affected third parties that can be remedied by a court.”). Such an injury “is impossible to remedy once the proceeding is over” and so the Court has warned against providing review “too late to be meaningful.” Axon Enter., 598 U.S. at 191. By their nature, Bivens claims usurp the judicial process to hale executive branch officials into court—all without the

GARRAWAY V. CUIFO 27

authority of the legislature. So improper Bivens expansion creates a separation-of-powers problem that infects all three branches of government.

It irreparably harms the judiciary and the claimants by “hold[ing] out [a] kind of false hope, and in the process invit[ing] still more protracted litigation destined to yield nothing.” It irreparably harms the legislature by “arrogating legislative power” and upsetting “the careful balance of interests struck by the lawmakers.” It irreparably harms the executive branch, both abstractly by impairing government functioning, interfering with executive autonomy, and chilling high-level policy making, and tangibly by imposing “time and administrative costs attendant upon intrusions resulting from the discovery and trial process.” Finally, its zombie existence harms the public writ large because, absent its formal abrogation, Congress has no incentive to legislate in the space. Instead, potential claimants are left with a brain-dead cause of action sustained by life support.

Mohamed, 100 F.4th at 1239–40 (Tymkovich, J., dissenting) (simplified). So there are real costs to both the parties and our constitutional structure by delaying review here. Contrary to the majority’s view, the denial of interlocutory appeal would cause “here-and-now” injuries. As Judge Tymkovich memorably warns, “the judicial process itself is the injury, these harms are a bell that cannot be unrung later in the litigation.” Id. at 1240 (Tymkovich, J., dissenting). Second, the impact of new Bivens liability on the effective functioning of the executive branch warrants immediate review. “Recognizing any new Bivens action entails substantial social costs, including the risk that fear of personal monetary liability and harassing litigation will unduly inhibit officials in the discharge of their duties.” Egbert, 596 U.S. at 499 (simplified). Facing new personal liability, federal officials may stop fully carrying out their constitutional duties for fear of being haled into court. We risk the “distraction of officials from their governmental duties,” the “inhibition of discretionary action,” and “deterrence of able people from public service.” Harlow v. Fitzgerald, 457 U.S. 800, 816 (1982). In the federal prison context, as here, these concerns are especially troubling—even life-threatening. If Bivens is extended to give prisoners a remedy for failure to protect them from attacks from other inmates—as Garraway seeks—the implications for federal prison officials could be devastating. Under Garraway’s theory, a plaintiff can turn practically any prison altercation into grounds for a Bivens suit and lengthy entanglement in federal courts. Prison officials may then need to respond by diverting resources to ensure that inmates are monitored closely enough. Or the opposite may occur—encouraging prison officials to avoid inmates to sidestep anything giving rise to liability. Knowing that they have another cause of action in their back pocket, inmates may make inappropriate demands hoping that the mere threat of litigation may induce prison officials to comply.

GARRAWAY V. CUIFO 29

This parade of horribles may be speculative. But that’s the point. The judiciary is particularly ill-equipped to assess these questions and reach proper policy determinations. We simply have no competence to strike the balance between “economic and governmental concerns, administrative costs, and the impact on governmental operations systemwide” needed to fashion new remedies. Egbert, 596 U.S. at 496 (simplified). That’s why the Constitution leaves decisions like these to Congress. But more importantly, the threat of any impact on the executive branch from Bivens liability cries out for immediate review. While Garraway’s Bivens claim may not ultimately prevail in court, defendants and their fellow prison officials may still alter their conduct realizing that any Bivens suit may go to final judgment before any appellate review. So denying this interlocutory appeal may chill these defendants’ as well as other prison officials’ conduct. The ripple effects may then damage the efficiency and effectiveness of the executive branch. This prong is thus satisfied, and we should have asserted jurisdiction here. 2. Response to Majority First, the majority doesn’t dispute that Bivens creation causes irreparable harm to the executive branch. Rather it claims there’s no urgency to act now because “the next logical step . . . in nearly any Bivens action” is for the defendant to claim qualified immunity, which is immediately appealable under our collateral order doctrine. Maj. Op. 15 (emphasis added). Since this argument was not briefed, I am not sure this is correct. And we have said, “[n]ot every interlocutory appeal from a denial of a motion for summary judgment based on qualified immunity is immediately appealable.” Pauluk v. Savage, 836 F.3d 1117, 1121 (9th Cir. 2016) (holding that “we do not have jurisdiction over a district court’s determination that there are genuine issues of material fact”). And, of course, there’s a class of Bivens cases where raising a qualified immunity defense would be frivolous. Qualified immunity is only available when the constitutional law is not “clearly established.” See Wilk v. Neven, 956 F.3d 1143, 1148 (9th Cir. 2020). Take this case. The district court said that Garraway’s claim mimics the facts and applicable law of Farmer v. Brennan, 511 U.S. 825, 843 (1994), which set “clearly established” law on the Eighth Amendment. See Wilk, 956 F.3d at 1150. Perhaps the reason the government hasn’t asserted qualified immunity in the years since the district court rejected its motion for judgment on the pleadings is that it would be frivolous. But even assuming all Bivens claims will eventually be reviewed as part of a qualified-immunity appeal after summary judgment, such an appeal would be an insufficient alternative means of vindicating the irreparable harm to the executive branch. Having to wait to litigate qualified immunity at summary judgment means that federal officers, like defendants here, will have to submit themselves to the costs and strains of discovery. Yet the Supreme Court has barred recognition of Bivens remedies precisely to prevent the burdens posed by discovery, not just trial itself. For example, the Court reasoned that “[c]laims against federal officials often create substantial costs, in the form of defense and indemnification” and that “the time and administrative costs attendant upon intrusions resulting from the discovery and trial process are significant factors to be considered.”

GARRAWAY V. CUIFO 31

Ziglar v. Abbasi, 582 U.S. 120, 134 (2017). It is “the burden and demand of litigation,” including litigating qualified immunity outside the pleading stage, that “might well prevent [federal officials]—or, to be more precise, future officials like them—from devoting the time and effort required for the proper discharge of their duties.” Id. at 141. See also Egbert, 596 U.S. at 499 (raising concerns that Bivens claims may lead to “broad-ranging discovery” and may not be “amendable to summary disposition”) (simplified). Indeed, we have repeatedly held that forcing the government to make unnecessary and practically non- reimbursable expenditures is sufficient reason to have immediate appeal. See Baker, 603 F.2d at 761–62; Wiggins, 717 F.2d at 467–68; Copeland, 852 F.3d at 905. So we leave here-and-now injuries without remedy if we tell Bivens defendants that they must wait until the qualified-immunity appeal for any relief. Second, the majority contends that finding jurisdiction here would violate the “spirit of Will.” Maj. Op. 19. In Will, the Court resolved whether the refusal to apply the judgment bar to a Bivens claim under the Federal Tort Claims Act (“FTCA”) can be immediately appealed. Will, 546 U.S. at 347. Thus, Will is about the FTCA’s judgment bar—not the creation of a new Bivens cause of action, as here. In denying collateral appeal, the Court observed that the “avoidance of litigation for its own sake” supported the FTCA’s judgment bar—not any “public interest.” Id. at 353. The Court then cautioned, “if simply abbreviating litigation troublesome to Government employees were important enough for Cohen treatment, collateral order appeal would be a matter of right whenever the Government lost a motion to dismiss under the Tort Claims Act, or a federal officer lost one on a Bivens action.” Id. at 353–354. The majority then take this sentence to mean that an adverse Bivens decision does not implicate the right not to stand trial—so there’s no need for immediate appeal. To begin, it’s unclear whether Will was referring to established Bivens claims here, rather than the creation of new ones, which have different constitutional concerns. But even assuming the Court referred to new Bivens actions, it’s “drive-by dictum.” Graber, 59 F.4th at 612 (Hardiman, J., dissenting). Given everything the Court has said about the creation of Bivens claims being a matter of substantial public interest since Will, this one-off statement shouldn’t be dispositive here. Finally, the majority suggests that expanding the collateral order doctrine, as a “judicially created exception” to § 1291, creates the same separation-of-powers problems as fashioning new causes of actions under Bivens. Of course, we should always pause anytime we depart from the plain text of a congressional statute. And whether the collateral order doctrine is consistent with the text of § 1291 is complicated and is a question for another day. Compare Mohawk Indus., 558 U.S. at 115–16 (Thomas, J., concurring) (concluding that the collateral order doctrine deviates from § 1291’s text) with Adam Reed Moore, A Textualist Defense of a New Collateral Order Doctrine, 99 Notre Dame L. Rev. Reflection 1, 1 (2023) (arguing that “final judgments, other decisions that end litigation on the merits, and orders deciding issues that are ancillary to the merits and will not be revisited” all fall within the meaning of “final decisions.”). But even if the collateral order doctrine is judge-made law, Bivens expansion is orders of magnitude more destabilizing to the separation of powers. As the Court has said, the “judicial creation of a cause of action is an

GARRAWAY V. CUIFO 33

extraordinary act that places great stress on the separation of powers.” Nestlé USA, Inc. v. Doe, 593 U.S. 628, 636 (2021) (emphasis added). And creating new Bivens claims results in irreparable harm for each branch of government, as discussed above. In contrast, the collateral order doctrine’s impact is largely confined to the judicial branch. Departure from the final-judgment rule only “undermines efficient judicial administration and encroaches upon the prerogatives of district court judges, who play a special role in managing ongoing litigation.” Mohawk Indus., 558 U.S. at 106 (simplified). That’s nothing like the systemwide effect of Bivens-claims creation. In the end, we must follow Supreme Court precedent. And under that clear precedent, this court has jurisdiction to hear this appeal. B. Motion for Reconsideration So under the collateral order doctrine, we have jurisdiction to review this appeal. The only wrinkle here is that the government appeals not from the initial denial of the judgment on the pleadings—but from the denial of reconsideration. See Fed. R. Civ. P. 60(b). Ordinarily, motions for reconsideration are not subject to interlocutory appeals under the collateral order doctrine—even when the underlying order would have been. See Hanson, 968 F.3d at 1018 (“[W]e lack jurisdiction over an order denying a Rule 59(e) motion for reconsideration of a denial of qualified immunity, where we do not have jurisdiction over the appeal of the underlying order.”). But Hanson carved out an exception for motions for reconsideration based on “intervening law.” Id. at 1019 n.4. In those cases, the intervening law may “render[] the collateral order doctrine applicable.” Id. This rule applies so long as the new caselaw isn’t a “simple reiterat[ion]” of what’s been “explained many times” before. See id. (simplified). Admittedly, the Hanson exception was decided over my objection. Id. at 1019 (Bumatay, J., concurring). But the exception remains the binding precedent of our court and it was made precisely for the facts here. The district court first denied the motion for judgment on the pleadings under existing, pre-Egbert Ninth Circuit precedent. The Supreme Court then decided Egbert and overturned our precedent. See Egbert, 596 U.S. at 496 (calling our court’s analysis “deeply flawed”). Egbert then newly distilled Bivens analysis into “only one question: whether there is any rational reason (even one) to think that Congress is better suited to weigh the costs and benefits of allowing a damages action to proceed.” Id. at 496 (simplified). The government then sought reconsideration based on Egbert’s correction of Ninth Circuit Bivens doctrine. So this case falls squarely into the Hanson exception. We thus have jurisdiction over this interlocutory appeal.

II.

The District Court Erred in Recognizing a New Bivens Cause of Action On the merits, federal courts are no longer in the business of creating new Bivens remedies and so the district court should be reversed. As I noted, “[t]he text of the Constitution provides for no express cause of action for damages against federal officials for violations of its provisions.” Boule v. Egbert, 998 F.3d 370, 375 (9th Cir. 2021) (Bumatay, J., dissenting from the denial of rehearing en banc). “And for almost 200 years, no implied cause of action existed under the Constitution

GARRAWAY V. CUIFO 35

either.” Id. That’s until Bivens came along in the 1970s. So Bivens is both atextual and ahistorical. The Court soon recognized that Bivens was a “relic of the heady days in which [the] Court assumed common-law powers to create causes of action.” Corr. Serv. Corp. v. Malesko, 534 U.S. 61, 75 (2001) (Scalia, J., concurring). In Egbert, the Court all but sounded Bivens’ “death knell” by “creat[ing] a self- defeating test” for new Bivens causes of action. Mohamed, 100 F.4th at 1244 (Tymkovich, J., dissenting). The new test requires federal courts to proceed in two steps. First, we determine whether the claim “presents a new Bivens context”—one “meaningfully different” from the three Bivens claims recognized by the Court. Egbert, 596 U.S. at 492 (simplified) (referring to Bivens; Davis v. Passman, 442 U.S. 228 (1979); and Carlson v. Green, 446 U.S. 14 (1980)). Second, “if a claim arises in a new context, a Bivens remedy is unavailable if there are special factors indicating that the Judiciary is at least arguably less equipped than Congress to weigh the costs and benefits of allowing a damages action to proceed.” Id. (simplified). In the end, the Bivens inquiry is often reduced to one question: “whether there is any reason to think that Congress might be better equipped to create a damages remedy.” Id. In my view, the answer is always “yes.” In any case, I turn to the two Bivens steps next. A. New Context Garraway’s Bivens claim presents a new context. Garraway’s claim—that federal officers failed to protect him from another prisoner—is “meaningfully different” from the three recognized Bivens claims: Bivens (Fourth Amendment illegal search), Davis (Fifth Amendment sex discrimination), and Carlson (Eighth Amendment failure to provide adequate medical care). Garraway suggests that the Supreme Court in Farmer implicitly recognized a fourth Bivens claim under the Eighth Amendment. In that case, the Court held that prison officials acted with deliberate indifference to the health and safety of an inmate when they exposed the inmate to substantial risks of injury. Farmer, 511 U.S. at 843. But two years ago, the Supreme Court expressly limited the three recognized Bivens causes of action to Bivens, Davis, and Carlson— leaving out Farmer. Egbert, 596 U.S. at 492 (citing Ziglar, 582 U.S. at 131). Nor has this court found Farmer to be a recognized Bivens claim. See Marquez v. C. Rodriguez, 81 F.4th 1027, 1031 (9th Cir. 2023) (“We thus decline [the] invitation to recognize an implied fourth Bivens context arising from Farmer. . . . If the Court were inclined to recognize it as one of the few acceptable Bivens contexts, it would have done so. Instead, the Court continues to reaffirm that there are but three of these cases, and Farmer is not one of them.”); Chambers v. C. Herrera, 78 F.4th 1100, 1105 n.2 (9th Cir. 2023) (“The Supreme Court has never recognized Farmer as a Bivens action. We will not do so in the first instance.”). Thus, contrary to the district court’s decision, under both Supreme Court and Ninth Circuit precedent, Garraway’s claim presents a new Bivens context. B. Reasons to Defer to Congress Although only one is needed, several reasons support believing that Congress, rather than our court, is in the better position to authorize a damages remedy in this new prison

GARRAWAY V. CUIFO 37

context. See Egbert, 596 U.S. at 492. To begin, Garraway asserts claims that touch on detention policy, which is “peculiarly the province of the Legislative and Executive Branches of the Government.” See Bell v. Wolfish, 441 U.S. 520, 548 (1987); id. at 547 (“[T]he problems that arise in the day-to-day operation of a corrections facility are not susceptible of easy solutions. Prison administrators therefore should be accorded wide-ranging deference in the adoption and execution of policies and practices that in their judgment are needed to preserve internal order and discipline and to maintain institutional security.”). This is reason enough to defer to Congress. But there’s more. As the Supreme Court has stated, an existing “alternative remedial structure” precludes the judiciary from inferring a new Bivens cause of action. Egbert, 596 U.S. at 493 (simplified); see also Ziglar, 582 U.S. at 137. Congress has already authorized an administrative remedy for the harms that Garraway claims to have suffered—the Prison Litigation Reform Act, 42 U.S.C. § 1997(e). Through this Act, the Bureau of Prisons (BOP) established a program to allow inmates to seek formal administrative review of issues relating to their confinement. See 28 C.F.R. § 542.10(a). The Supreme Court has also recognized this program as a “means through which allegedly unconstitutional actions and policies can be brought to the attention of the BOP and prevented from recurring.” Malesko, 534 U.S. at 73. Because Congress is uniquely qualified to balance the interests of prison policy and has already provided an alternative remedy, the district court erred in recognizing Garraway’s new Bivens cause of action. The district court’s new-found Bivens cause of action violates both precedent and the Constitution. Too bad we pass on our chance to remedy this significant damage to the separation of powers. Because we refuse jurisdiction to review a district court ruling causing irreparable harm, I respectfully dissent.