CHURCH AUTONOMY AND COLLATERAL-ORDER APPEALS

CHURCH AUTONOMY AND COLLATERAL-ORDER APPEALS

In recent years, a growing number of litigants and scholars have argued that—despite the usual rule in federal court that only final orders are appealable—interlocutory orders denying church-autonomy defenses under the First Amendment can be appealed immediately. Proponents ground their claims in the belief that church autonomy provides religious institutions with an immunity from suit, rather than with a mere defense to liability. As a result, the argument goes, orders denying church-autonomy defenses fall within the collateral-order doctrine, which allows for immediate appeal of certain interlocutory decisions, mostly concerning immunities from suit.

Every federal court of appeals to address the issue in the past three years has rejected this “immunity theory” of church autonomy and dismissed attempts at interlocutory appeal. But none of these courts engaged in the historical inquiry that the Supreme Court has instructed must guide interpretation of the Religion Clauses. This Comment fills that analytical gap, arguing that historical understandings of church–state relations in the early Republic favor reading church autonomy as a deference doctrine that does not give rise to immediately appealable collateral orders.

The full text of this Piece can be found by clicking the PDF link to the left.

Introduction

Since 2022, three federal courts of appeals—the Second, Seventh, and Tenth Circuits—have held that they lack jurisdiction to review interlocutory decisions denying church-autonomy defenses under the First Amendment. 1 See Garrick v. Moody Bible Inst., 95 F.4th 1104, 1106 (7th Cir. 2024) (holding that a district court’s denial of a motion to dismiss a religious college’s church-autonomy defense in a Title VII suit was not immediately appealable); Belya v. Kapral, 45 F.4th 621, 625 (2d Cir. 2022) (holding that a district court’s denial of a motion to dismiss a defamation claim on church-autonomy grounds was not immediately appealable), cert. denied, 143 S. Ct. 2609 (2023) (mem.); Tucker v. Faith Bible Chapel Int’l, 36 F.4th 1021, 1025 (10th Cir. 2022) (holding that a district court’s denial of summary judgment on a religious employer’s ministerial exception defense in a Title VII suit was not immediately appealable), cert. de­nied, 143 S. Ct. 2608 (2023) (mem.).
         In an unpublished opinion in 2024, the Eleventh Circuit also dismissed an interlocutory appeal of an order denying a church-autonomy defense for lack of jurisdic­tion. See Klein v. Oved, No. 23-14105, 2024 WL 1092324, at *1 (11th Cir. Mar. 13, 2024) (per curiam). But that opinion is not precedential. See 11th Cir. R. 36-2 (“Unpublished opinions are not considered binding precedent, but they may be cited as persuasive authority.”).
These outcomes are, in some sense, unsurprising. In general, circuit courts have jurisdiction to review only final decisions. 2 See 28 U.S.C. § 1291 (2018). Parties do not have a right to immediately appeal most interlocutory orders. 3 See id. § 1292 (specifying the limited circumstances in which courts of appeals can review interlocutory orders). But the consistency of the outcomes in these cases belies a still-simmering debate about the nature of the Religion Clauses’ protections for church autonomy, which could have significant consequences for plaintiffs’ ability to seek recourse from religious institutions.

In each case, a defendant religious institution argued that, notwith­standing the final-judgment rule, a district court’s interlocutory denial of a church-autonomy defense was immediately appealable under the collateral-order doctrine. 4 See Appellant’s Brief at 15–26, Garrick, 95 F.4th 1104 (No. 21-2683) [hereinafter Garrick Appellant’s Brief]; Defendants-Appellants’ Opening Brief at 49–58, Belya, 45 F.4th 621 (No. 21-1498), 2021 WL 3856216 [hereinafter Belya Defendants-Appellants’ Brief]; Ap­pellant Faith Bible Chapel International’s Opening Brief at 44–46, Tucker, 36 F.4th 1021 (No. 20-1230), 2020 WL 6077116 [hereinafter Tucker Appellant’s Brief]. That doctrine treats a small class of interlocutory decisions, mostly concerning immunities from suit, as “final” and therefore appealable on an immediate basis. 5 Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868–74 (1994) (explaining the requirements for immediate appeal under the collateral-order doctrine and noting that “orders denying certain immunities are strong candidates for prompt appeal”). The defendants in these cases contended that church autonomy not only insulates religious institutions from liability for ecclesiastical decisions but also immunizes them from the burdens of litigation over those decisions. 6 See Garrick Appellant’s Brief, supra note 4, at 18–23 (“[T]he Religion Clauses limit the process of litigation, not merely liability.”); Belya Defendants-Appellants’ Brief, supra note 4, at 51–54 (claiming that, like qualified immunity, church autonomy implicates a pro­tected right that must be resolved prior to discovery and trial); Tucker Appellant’s Brief, supra note 4, at 44–45 (advocating for appellant’s First Amendment protection from the burdens of trial in matters relating to its internal management of clergy). As a result, they insisted, inter­locutory decisions denying church-autonomy defenses are appealable as collateral orders. 7 See Garrick Appellant’s Brief, supra note 4, at 26 (concluding that church auton­omy is an “immunity sufficient to justify interlocutory review” (citing Mitchell v. Forsyth, 472 U.S. 511, 525 (1985); Rubin v. Islamic Republic of Iran, 637 F.3d 783, 790 (7th Cir. 2011))); Belya Defendants-Appellants’ Brief, supra note 4, at 50 (asserting that an appeal of a denial of a church-autonomy defense “falls squarely within the collateral order doctrine, as it involves the denial of a First Amendment immunity from discovery and trial”); Tucker Appellant’s Brief, supra note 4, at 46 (arguing that “[d]efenses under the church autonomy doctrine . . . provide an immunity from merits discovery and trial, the denial of which can be immediately appealed”).

Though the Second, Seventh, and Tenth Circuits rejected that theory, 8 See Garrick, 95 F.4th at 1109–17 (“No court has ever held that the First Amendment doctrine of church autonomy establishes a constitutional right to immunity from trial in cases where non-ministerial employees allege non-religious discrimination.”); Belya, 45 F.4th at 630–34 (finding no support for collateral-order appeals rooted in the church-autonomy doctrine in decisions of either the Supreme Court or sister circuits); Tucker, 36 F.4th at 1036–47 (finding that orders preliminarily denying summary judgment to religious employers in employment discrimination claims fall outside the collateral-order doctrine). religious institutions continue to argue that interlocutory church-autonomy decisions are immediately appealable. 9 See, e.g., Opening Brief of Defendant-Appellant at 18–30, O’Connell v. U.S. Conf. of Cath. Bishops, No. 23-7173 (D.C. Cir. filed Sept. 10, 2024) [hereinafter O’Connell Opening Brief of Defendant-Appellant] (“[USCCB] asserts a First Amendment autonomy from being ‘deposed, interrogated, and haled into court’ to question the Pope’s judgment about how to steward a religious offering.” (quoting Equal Emp. Opportunity Comm’n v. Cath. Univ. of Am., 83 F.3d 455, 467 (D.C. Cir. 1996))). And they are gaining traction. Several judges and scholars have already expressed strong sup­port for the religious institutions’ view of church autonomy, 10 See, e.g., Garrick, 95 F.4th at 1117–25 (Brennan, J., dissenting) (“The First Amendment’s protection of church autonomy provides immunity precluding litigation of religious questions.”); Belya v. Kapral, 59 F.4th 570, 577–80 (2d Cir. 2023) (Park, J., dissenting from the order denying rehearing en banc) (“Rejections of church autonomy defenses should be immediately appealable, in the same way that denials of qualified immunity are appealable.”); Tucker, 36 F.4th at 1049–59 (Bacharach, J., dissenting) (“As a structural safeguard, the ministerial exception protects religious bodies from the suit itself—unlike most affirm­ative defenses that protect only against liability.”); Carl H. Esbeck, An Extended Essay on Church Autonomy, 22 Federalist Soc’y Rev. 244, 266–67 (2021) (“When a church has timely raised the ministerial exception by pleading or motion and the affirmative defense has been denied by the trial court, the structural nature of church autonomy calls for an interlocutory appeal.”); Peter J. Smith & Robert W. Tuttle, Civil Procedure and the Ministerial Exception, 86 Fordham L. Rev. 1847, 1880–81 (2018) (arguing that the application of the collateral-order doctrine to ministerial exception cases “would better guard against Establishment Clause violations by trial courts”); Lael Weinberger, Is Church Autonomy Jurisdictional?, 54 Loy. U. Chi. L.J. 471, 503–05 (2022) [hereinafter Weinberger, Is Church Autonomy Jurisdic­tional?] (“Interlocutory appeals from church autonomy are necessary to protect interests at the heart of the church autonomy doctrine.”). and the Second and Tenth Circuits’ decisions narrowly avoided rehearing en banc. 11 See Belya, 59 F.4th at 572 (2d Cir. 2023) (denying rehearing by a 6-6 vote); Tucker v. Faith Bible Chapel Int’l, 53 F.4th 620, 622 (10th Cir. 2022) (denying rehearing by a 6-4 vote). Should the judicial tide change, religious institutions could make it much harder for plaintiffs to get to trial by subjecting them to costly, years-long interlocutory appeals. 12 As the Supreme Court has noted, the general rule against interlocutory appeal serves to avoid “the obstruction to just claims that would come from permitting the harass­ment and cost of a succession of separate appeals from the various rulings to which a litiga­tion may give rise.” Will v. Hallock, 546 U.S. 345, 350 (2006) (internal quotation marks omitted) (quoting Firestone Tire & Rubber Co. v. Risjord, 449 U.S. 368, 374 (1981)). Recent cases involving interlocutory appeals of decisions denying church-autonomy defenses demonstrate the risk of obstruction and delay. In Belya v. Kapral, for example, a church’s attempted appeal took two years to resolve. Compare Notice of Appeal at 1, Belya, 45 F.4th 621 (No. 21-1498) (noting that appeal was initiated on June 17, 2021), with Synod of Bishops of the Russian Orthodox Church Outside of Russ. v. Belya, 143 S. Ct. 2609 (2023) (mem.) (denying certiorari on June 12, 2023). And in Garrick v. Moody Bible Institute, a similar appeal took two and a half years to be resolved. Compare Notice of Appeal by Defendant, the Moody Bible Institute of Chicago at 1, Garrick v. Moody Bible Inst., 494 F. Supp. 3d 570 (N.D. Ill. 2020) (No. 1:18-cv-00573) (initiating appeal on September 13, 2021), with Garrick v. Moody Bible Inst., No. 21-2683, 2024 WL 1892433, at *1 (7th Cir. Apr. 30, 2024) (denying rehearing).

This Comment argues that interlocutory orders denying church-autonomy defenses are not immediately appealable, but for reasons not addressed in the Second, Seventh, and Tenth Circuits’ decisions. Part I introduces the collateral-order doctrine and explains that whether an in­terlocutory order is immediately appealable turns on whether it concerns an immunity from suit or a mere defense to liability. Part II describes com­peting arguments among judges, scholars, and litigants over whether church autonomy provides an immunity that is subject to collateral-order review. Part III argues that—though largely absent from the Second, Seventh, or Tenth Circuits’ reasoning—historical understandings of church–state relations in the early Republic show that church autonomy does not operate as an immunity from suit. Instead, church autonomy is best understood as a deference doctrine that cannot give rise to immedi­ately appealable collateral orders.

I. The Collateral-Order Doctrine

This Part provides an overview of the collateral-order doctrine, its requirements, and the limited set of rights for which immediate appeal is warranted. 13 For a more detailed discussion of the doctrine’s development and scope, see Lloyd C. Anderson, The Collateral Order Doctrine: A New “Serbonian Bog” and Four Proposals for Reform, 46 Drake L. Rev. 539, 548–68 (1998); Michael E. Harriss, Note, Rebutting the Roberts Court: Reinventing the Collateral Order Doctrine Through Judicial Decision-Making, 91 Wash. U. L. Rev. 721, 728–34 (2014).

Under 28 U.S.C. § 1291, the jurisdiction of the federal courts of ap­peals is generally limited to reviewing “final decisions of the district courts.” 14 29 U.S.C. § 1291 (2018). A final decision typically “‘terminate[s the] action’ or ‘ends the litigation on the merits and leaves nothing for the court to do but execute the judgment.’” Ritzen Grp., Inc. v. Jackson Masonry, LLC, 140 S. Ct. 582, 586 (2020) (alteration in original) (quoting Gelboim v. Bank of Am. Corp., 574 U.S. 405, 409 (2015)). Interlocutory decisions are, with few exceptions, not appealable until after a final judgment on the merits. 15 See 28 U.S.C. § 1292(a) (2018) (granting jurisdiction to review certain orders con­cerning injunctions, receiverships, or admiralty issues); id. § 1292(b) (granting discretionary jurisdiction to review interlocutory orders when a district judge certifies “that such order involves a controlling question of law as to which there is substantial ground for difference of opinion and that an immediate appeal . . . may materially advance the ultimate termination of the litigation”); id. § 1292(e) (empowering the Supreme Court to “prescribe rules . . . to provide for an appeal of an interlocutory decision to the courts of appeals that is not otherwise provided for”). By requiring that “the whole case . . . [be] decided in a single appeal,” 16 Microsoft Corp. v. Baker, 137 S. Ct. 1702, 1712 (2017) (second alteration in origi­nal) (internal quotation marks omitted) (quoting McLish v. Roff, 141 U.S. 661, 665–66 (1891)). this final-judgment rule “pre­serves the proper balance between trial and appellate courts, minimizes the harassment and delay that would result from repeated interlocutory appeals, and promotes the efficient administration of justice.” 17 Id.

But the Supreme Court “has long given [§ 1291] a practical rather than a technical construction,” treating a “small class” of “collateral” or­ders as final even though they do not end litigation on the merits. 18 Cohen v. Beneficial Indus. Loan Corp., 337 U.S. 541, 546 (1949) (citing Cobbledick v. United States, 309 U.S. 323, 328 (1940); United States v. River Rouge Co., 269 U.S. 411, 414 (1926); Bank of Columbia v. Sweeny, 26 U.S. (1 Pet.) 567, 569 (1828)). That small class “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.” 19 Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 106 (2009) (internal quotation marks omitted) (quoting Swint v. Chambers Cnty. Comm’n, 514 U.S. 35, 42 (1995)). Whether these requirements are satisfied is “determined for the entire category to which a claim belongs, without regard to the chance that the liti­gation at hand might be speeded, or a ‘particular injustic[e]’ averted.” 20 Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 868 (1994) (alteration in original) (quoting Van Cauwenberghe v. Biard, 486 U.S. 517, 529 (1988)).

The Supreme Court’s application of the collateral-order doctrine has often been criticized as unpredictable and incoherent. 21 See Bryan Lammon, Rules, Standards, and Experimentation in Appellate Jurisdiction, 74 Ohio St. L.J. 423, 431 n.38 (2013) (compiling criticisms); Adam N. Steinman, Reinventing Appellate Jurisdiction, 48 B.C. L. Rev. 1237, 1238–39 (2007) (same). But see Timothy P. Glynn, Discontent and Indiscretion: Discretionary Review of Interlocutory Orders, 77 Notre Dame L. Rev. 175, 204–05 (2001) (admitting that “criticism may have been warranted a decade ago” but insisting that “the collateral order doctrine is now both coherent and easy to apply”); Aaron R. Petty, The Hidden Harmony of Appellate Jurisdiction, 62 S.C. L. Rev. 353, 385 (2010) (arguing that the contemporary Court “appears to be applying the collateral order doctrine in a predictable and uniform manner”). But for the past forty years, the Court has articulated at least one consistent principle un­dergirding its finality jurisprudence: A decision denying “an immunity from suit rather than a mere defense to liability” necessarily satisfies the requirements for a collateral order. 22 Swint, 514 U.S. at 42 (emphasis omitted) (quoting Mitchell v. Forsyth, 472 U.S. 511, 526 (1985)); see also P.R. Aqueduct & Sewer Auth. v. Metcalf & Eddy, Inc., 506 U.S. 139, 144 (1993) (“Once it is established that [defendants] are, in effect, immune from suit . . . it follows that the elements of the Cohen [v. Beneficial Industry Loan Corporation] collateral order doctrine are satisfied.”). For a more detailed description of the Supreme Court’s “line of cases concluding that immunities from suit that encompass a right not to be tried are im­mediately appealable,” see Petty, supra note 21, at 383–86. Such decisions conclusively deter­mine that defendants “have no right not to be sued.” 23 P.R. Aqueduct & Sewer Auth., 506 U.S. at 145 (discussing the implications of the denial of state sovereign immunity under the Eleventh Amendment). That right not to be sued is conceptually distinct from the underlying merits of an action. 24 Plumhoff v. Rickard, 572 U.S. 765, 772 (2014); see also Mitchell, 472 U.S. at 527–28 (explaining that “[a]n appellate court reviewing the denial of the defendant’s claim of im­munity need not consider the correctness of the plaintiff’s version of the facts, nor even determine whether the plaintiff’s allegations actually state a claim” and instead need only determine “a question of law”). And if erroneously denied, the right “could not be effectively reviewed on appeal from a final judgment because by that time the immunity from standing trial will have been irretrievably lost.” 25 Plumhoff, 572 U.S. at 772.

The Supreme Court has strictly limited the rights it recognizes as cre­ating immunities from suit that are subject to collateral-order review. “[V]irtually every right that could be enforced appropriately by pretrial dismissal might loosely be described as conferring a ‘right not to stand trial.’” 26 Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 873 (1994) (citations omitted). But such bare characterizations of a right do not alone warrant immediate appeal. 27 See Swint, 514 U.S. at 43 (“§ 1291 requires courts of appeals to view claims of a ‘right not to be tried’ with skepticism, if not a jaundiced eye.” (internal quotation marks omitted) (quoting Digit. Equip., 511 U.S. at 873)). Otherwise, the Court has warned, the collateral-order doctrine would “swallow the general rule that a party is entitled to a single appeal, to be deferred until final judgment has been entered.” 28 Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 106 (2009) (internal quotation marks omitted) (quoting Digit. Equip., 511 U.S. at 868). Instead, a right confers a true immunity only if it protects “a substantial public inter­est” 29 Id. at 107 (internal quotation marks omitted) (quoting Will v. Hallock, 546 U.S. 345, 353 (2006)). that is “weightier than the societal interests advanced by the ordinary operation of final judgment principles.” 30 Digit. Equip., 511 U.S. at 879. Such immunities usually arise from one of two sources. First, they may stem from “an explicit statutory or constitutional guarantee that trial will not occur,” 31 Midland Asphalt Corp. v. United States, 489 U.S. 794, 801 (1989); see also Digit. Equip., 511 U.S. at 879 (“When a policy is embodied in a constitutional or statutory provision entitling a party to immunity from suit (a rare form of protection), there is little room for the judiciary to gainsay its ‘importance.’”). like the Speech or Debate Clause, 32 See Helstoski v. Meanor, 442 U.S. 500, 506–09 (1979) (“[T]he Speech or Debate Clause was designed to protect Congressmen ‘not only from the consequences of litigation’s results but also from the burden of defending themselves.’” (quoting Dombrowski v. Eastland, 387 U.S. 82, 85 (1967))). the Double Jeopardy Clause, 33 See Abney v. United States, 431 U.S. 651, 661–62 (1977) (“[I]f a criminal defendant is to avoid exposure to double jeopardy and thereby enjoy the full protection of the Clause, his double jeopardy challenge to the indictment must be reviewable before that subsequent exposure occurs.”). the Eleventh Amendment, 34 See P.R. Aqueduct & Sewer Auth. v. Metcalf & Eddy, Inc., 506 U.S. 139, 144–45 (1993) (finding that the protection of Eleventh Amendment immunity from suit “is for the most part lost as litigation proceeds past motion practice”). or the Foreign Sovereign Immunities Act. 35 Though the Supreme Court has not decided whether decisions denying claims of immunity under the Foreign Sovereign Immunities Act are immediately appealable, every circuit to address the issue has held that such decisions are collateral orders. See Steinman, supra note 21, at 1250 n.95 (collecting cases). Second, an immunity may arise from the need to preserve the proper functioning of, or allocation of power within, the government. 36 See Mitchell v. Forsyth, 472 U.S. 511, 525–26 (1985) (explaining that qualified im­munity serves to avoid “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” (internal quotation marks omitted) (quoting Harlow v. Fitzgerald, 457 U.S. 800, 819 (1982))); see also Will v. Hallock, 546 U.S. 345, 352 (2006) (explaining that public interests warranting immediate appeal include “honoring the separation of powers, preserving the efficiency of government and the initi­ative of its officials, [and] respecting a State’s dignitary interests”). Qualified immunity, for ex­ample, protects government officials from the burdens of litigating unwar­ranted suits that “can be peculiarly disruptive of effective government,” regardless of the ultimate outcome. 37 Mitchell, 472 U.S. at 526 (internal quotation marks omitted) (quoting Harlow, 457 U.S. at 817). If a right does not flow from an ex­plicit guarantee or from governance concerns, it will rarely confer an immunity whose denial is appealable as a collateral order, even if construed as a “right not to stand trial.” 38 See, e.g., Swint v. Chambers Cnty. Comm’n, 514 U.S. 35, 42–43 (1995) (holding that a municipality’s claim of a “qualified right to be free from the burdens of trial” under § 1983 is not an immunity from suit warranting collateral-order review); Digit. Equip. Corp. v. Desktop Direct, Inc., 511 U.S. 863, 873–77 (1994) (holding that a purported “right not to stand trial” under a private settlement agreement is not immediately appealable); cf. Mohawk Indus., Inc. v. Carpenter, 558 U.S. 100, 109 (2009) (holding that disclosure orders adverse to the attorney–client privilege, which appellant portrayed as a “right not to disclose the privileged information in the first place,” are not collateral orders (internal quotation marks omitted) (quoting Brief for Petitioner at 25, Mohawk, 558 U.S. 100 (No. 08-678), 2009 WL 1155404)).

Beyond limiting immediate appeal to decisions denying narrowly de­fined immunities, the Supreme Court has in recent years disfavored any further expansion of the collateral-order doctrine. 39 See Mohawk, 558 U.S. at 114 (holding that the current doctrine provides “adequate protection to litigants” and any further expansion of the doctrine “should be furnished, if at all, through rulemaking”). The Court has in­sisted that deferring appeal until final judgment on the merits is appropriate in most cases. 40 See id. at 108–09 (noting that the Court “routinely require[s] litigants to wait until after final judgment to vindicate valuable rights, including rights central to our adversarial system”); cf. Will, 546 U.S. at 350 (noting the importance of maintaining “the substantial finality interests § 1291 is meant to further”). When immediate appeal is required to pre­vent injustice in particular cases, the Court encourages the use of alternate avenues to seek appellate review over “the blunt, categorical instrument of § 1291 collateral order appeal.” 41 Mohawk, 558 U.S. at 112 (internal quotation marks omitted) (quoting Digit. Equip., 511 U.S. at 883); see also Cunningham v. Hamilton County, 527 U.S. 198, 209–10 (1999) (Kennedy, J., concurring) (“Should [the] hardships [of denying immediate appeal] be deemed to outweigh the desirability of restricting appeals to ‘final decisions,’ solutions other than an expansive interpretation of § 1291’s ‘final decision’ requirement remain availa­ble.”). These alternatives include mandamus and permissive appeals under § 1292(b). Mohawk, 558 U.S. at 110–13. And to the extent it is necessary to des­ignate a class of orders as categorically appealable, the Court has explained, “The procedure Congress ordered for such changes . . . is not expansion by court decision” under the collateral-order doctrine, “but by rulemaking under” the Rules Enabling Act, 28 U.S.C. § 2072. 42 Swint, 514 U.S. at 48. That Act empowers the Court to adopt rules that “define when a ruling of a district court is final for the purposes of appeal under section 1291.” 28 U.S.C. § 2072(c) (2018). As the Court has noted, “the rulemaking process has important virtues” as compared to the collateral-order doctrine because “[i]t draws on the collective experience of bench and bar” and “facilitates the adoption of measured, practical solu­tions.” Mohawk, 558 U.S. at 114.
         At least two justices have indicated that rulemaking leaves no room or justification for recognition of new classes of collateral orders. Justice Clarence Thomas, for example, has criticized the collateral-order doctrine as a “judicial policy” that “subordinate[s] what the appellate jurisdiction statute says to what the Court thinks is a good idea.” Id. at 115, 119 (Thomas, J., concurring in part and concurring in the judgment). Rather than “needlessly perpetuate[]” that problematic policy, Justice Thomas “would leave [such] value judg­ments . . . to the rulemaking process.” Id. Justice Neil Gorsuch similarly remarked, when he was a circuit judge, that “any pleas to expand appellate jurisdiction ought be directed to the Rules Committee, not our doorstep.” McClendon v. City of Albuquerque, 630 F.3d 1288, 1297 n.2 (10th Cir. 2011).
As a result, the Court has recognized only three new categories of collateral orders in the twenty-first century. 43 See Shoop v. Twyford, 142 S. Ct. 2037, 2043 n.1 (2022) (agreeing with the court of appeals that orders under the All Writs Act requiring states to transport prisoners are ap­pealable as collateral orders); Osborn v. Haley, 549 U.S. 225, 239 (2007) (holding that an order denying a government employee’s immunity from suit under the Westfall Act is im­mediately appealable); Sell v. United States, 539 U.S. 166, 176–77 (2003) (holding that a pretrial order requiring a defendant to involuntarily receive medication in order to render him competent to stand trial was appealable as a collateral order because it “raise[d] ques­tions of clear constitutional importance” (citing Coopers & Lybrand v. Livesay, 437 U.S. 463, 468 (1978))). A purported immunity from suit is thus unlikely to spawn a new class of collateral orders unless it is rooted in a particularly clear textual guarantee or particularly strong concerns about effective governance. 44 Cf. Will, 546 U.S. at 353–54 (finding that the Federal Tort Claims Act’s bar to cer­tain claims against government employees did not create an immunity from suit or warrant collateral-order review where it aimed “simply to save trouble for the Government and its employees” rather than to avoid more serious threats to efficient governance).

II. Competing Views of Church Autonomy and Its Immediate Appealability

In recent years, religious institutions have argued that decisions deny­ing church-autonomy defenses under the First Amendment are immediately appealable as collateral orders. 45 See supra notes 4–9 and accompanying text. The church-autonomy doc­trine, also known as the ecclesiastical-abstention doctrine, 46 McRaney v. N. Am. Mission Bd. of the S. Baptist Convention, Inc., 966 F.3d 346, 347 (5th Cir. 2020). protects religious institutions’ “independence in matters of faith and doctrine and in closely linked matters of internal government.” 47 Our Lady of Guadalupe Sch. v. Morrissey-Berru, 140 S. Ct. 2049, 2061 (2020). The doctrine’s earliest roots were in intracongregational disputes over church property. See infra section II.A. But it is now frequently invoked as a defense against all sorts of government regulation and judicial interference in religious institutions’ affairs. Some churches have argued, for example, that the doctrine shields them from lawsuits based on religious offi­cials’ alleged sexual abuse of children. See, e.g., Doe v. Roman Cath. Bishop of Springfield, 190 N.E.3d 1035, 1038 (Mass. 2022). Other churches have asserted church-autonomy de­fenses to defamation and intentional infliction of emotional distress (IIED) claims by reli­gious officials and congregation members. See, e.g., McRaney, 966 F.3d at 347–49 (rejecting the district court’s use of the church-autonomy doctrine to dismiss plaintiff’s defamation and IIED claims because the complaint “involves a civil rather than religious dispute”); Hubbard v. J Message Grp. Corp., 325 F. Supp. 3d 1198, 1207–09 (D.N.M. 2018) (consider­ing a church’s use of the church-autonomy doctrine against a member’s IIED and defamation claims); In re Diocese of Lubbock, 624 S.W.3d 506, 509, 511 (Tex. 2021) (“[T]he deacon’s [defamation and IIED] claims relating to the Diocese’s publication and communi­cation of the results of its investigation cannot be severed from its policy to investigate its clergy in the first place.”). And religious institutions have relied on the church-autonomy doctrine to seek exemptions from antidiscrimination laws and mandates to provide certain forms of healthcare or insurance coverage. See, e.g., McMahon v. World Vision, Inc., 704 F. Supp. 3d 1121, 1135–37 (W.D. Wash. 2023) (discussing arguments that prohibiting religious institutions from discriminating in employment based on sex, sexual orientation, or marital status violates church autonomy); Cedar Park Assembly of God of Kirkland v. Kreidler, 683 F. Supp. 3d 1172, 1187–88 (W.D. Wash. 2023) (discussing arguments that requiring churches to cover abortion and contraception in employee health insurance plans violates church autonomy); Opening Brief of Appellant Christian Healthcare Centers, Inc. at 61–63, Christian Healthcare Ctrs., Inc. v. Nessel, 117 F.4th 826 (6th Cir. 2024) (No. 23-1769), 2023 WL 7040138 (arguing that requiring religious healthcare institutions to offer gender-affirming care violates church autonomy).
         The ministerial exception is a branch of this church-autonomy doctrine, as a church’s choice of minister is a matter of internal governance that relates closely to faith and doc­trine. See Our Lady of Guadalupe Sch., 140 S. Ct. at 2061 (locating the constitutional foundation for the ministerial exception in “the general principle of church autonomy”); Lael Weinberger, The Limits of Church Autonomy, 98 Notre Dame L. Rev. 1253, 1255 n.4 (2023) (defining the ministerial exception as an “application of church autonomy doc­trine”).
In attempting to appeal interlocutory church-autonomy decisions as collateral orders, reli­gious institutions have contended that church autonomy is akin to qualified immunity, a shield to the burdens of litigation that is not subject to the ordinary final-judgment rule. 48 See Belya Defendants-Appellants’ Brief, supra note 4, at 51 (arguing that, “like qualified immunity, the Religion Clauses’ rule against interference in internal religious af­fairs provides ‘immunity from the travails of a trial and not just from an adverse judgment’” (quoting McCarthy v. Fuller, 714 F.3d 971, 975 (7th Cir. 2013))); Tucker Appellant’s Brief, supra note 4, at 46 (same); see also Garrick Appellant’s Brief, supra note 4, at 18–19 (arguing that church autonomy necessitates early protections from “judicial interference”). The three federal courts of appeals to address the question since 2022—the Second, Seventh, and Tenth Circuits—have each rejected that argument. 49 See supra note 8. As noted above, a fourth court of appeals—the Eleventh Circuit—reached the same result in an unpublished, nonbinding opinion last year. See Klein v. Oved, No. 23-14105, 2024 WL 1092324, at *1 (11th Cir. Mar. 13, 2024) (per curiam) (holding that the lower court’s order was “not immediately appealable pursuant to the collateral order doctrine” because it “did not conclusively determine whether the ecclesiastical abstention doctrine could shield [the defendants] from liability”). But churches remain eager to litigate the issue and have repeatedly sought Supreme Court review. 50 See O’Connell Opening Brief of Defendant-Appellant, supra note 9; Petition for a Writ of Certiorari at 25–30, Faith Bible Chapel Int’l v. Tucker, 143 S. Ct. 2608 (2023) (No. 22-741), 2023 WL 1864484; Petition for a Writ of Certiorari at 23–29, Synod of Bishops of the Russian Orthodox Church Outside of Russ. v. Belya, 143 S. Ct. 2609 (2023) (mem.) (No. 22-824), 2023 WL 2339736; see also Conditional Cross-Petition for Writ of Certiorari at 16, Liberty Univ., Inc. v. Bowes, 144 S. Ct. 1030 (2024) (No. 23-703), 2023 WL 9064333 (asking the Supreme Court to resolve “whether th[e] ministerial exception represents an immunity to suit altogether, or whether it is merely a defense to liability”). This Part chronicles the development of the church-autonomy doctrine and surveys the competing arguments over whether church-autonomy deci­sions are immediately appealable as collateral orders.

A.      A Short History of the Church-Autonomy Doctrine

The Supreme Court first articulated a principle of church autonomy as a matter of common law in Watson v. Jones in 1871. 51 80 U.S. (13 Wall.) 679 (1871). The dispute in Watson arose from a schism within a local Presbyterian congregation be­tween pro- and antislavery factions, each of which claimed control over the local church’s property. 52 See id. at 717. In adjudicating which faction constituted the true church, the Court explained that churches must have independence in deciding matters of “theological controversy, church discipline, ecclesi­astical government, or the conformity of the members of the church to the standard of morals required of them.” 53 Id. at 733. As a result, when an ecclesiastical authority has decided such religious questions, civil courts “must accept such decisions as final, and as binding on them, in their application to the case before them.” 54 Id. at 727. In Watson, the highest ecclesiastical authority govern­ing the local congregation—the General Assembly of the Presbyterian Church of the United States of America—had already determined that the antislavery faction properly controlled the disputed property. 55 See id. at 694. The Court was bound by that decision. 56 Id. at 733.

Following the incorporation of the Religion Clauses against the states, 57 See Everson v. Bd. of Educ., 330 U.S. 1, 15–16 (1947) (incorporating the Establishment Clause); Cantwell v. Connecticut, 310 U.S. 296, 303–05 (1940) (finding that “[t]he Fourteenth Amendment has rendered the legislatures of the states as incompetent as Congress” to restrict the free exercise of religion). the Supreme Court constitutionalized Watson’s church-autonomy principle in 1952. 58 See Kedroff v. Saint Nicholas Cathedral, 344 U.S. 94, 116 (1952) (holding that Watson’s “spirit of freedom for religious organizations” in “matters of church government as well as those of faith and doctrine” have to “be said to have federal constitutional protec­tion”). Throughout the next two decades, the Court repeat­edly held that interference with the decisions of religious authorities on “questions of discipline, or of faith, or ecclesiastical rule, custom, or law” violates the First Amendment. 59 See id. at 113 (internal quotation marks omitted) (quoting Watson, 80 U.S. (13 Wall.) at 727). The Court thus invalidated a New York statute that transferred control of local churches from a foreign governing authority to a domestic authority; 60 Id. at 107–08. forbade Georgia courts from resolving church property disputes based on the courts’ evaluation of the parties’ adherence to religious doctrine; 61 Presbyterian Church v. Mary Elizabeth Blue Hull Mem’l Presbyterian Church, 393 U.S. 440, 449–50 (1969). and overturned an Illinois court order preventing the Serbian Orthodox Church from reorganizing its dioceses due to a judicial interpretation of church law. 62 Serbian E. Orthodox Diocese v. Milivojevich, 426 U.S. 696, 721–22 (1976). But the Court did not com­pletely bar civil adjudication of disputes involving religious institutions. In­stead, it reasoned that resolving such disputes using “neutral principles of law” is constitutional. 63 See Jones v. Wolf, 443 U.S. 595, 602–04 (1979) (holding “that a State is constitu­tionally entitled to adopt neutral principles of law as a means of adjudicating a church property dispute”). A Maryland court’s resolution of a church property dispute based on secular language in a deed, for example, “involved no inquiry into religious doctrine” and did not offend the First Amendment. 64 Md. & Va. Eldership of the Churches of God v. Church of God at Sharpsburg, Inc., 396 U.S. 367, 368 (1970) (per curiam); see also Jones, 443 U.S. at 602–04 (describing the “‘neutral principles of law’ approach”).

Though these decisions made clear that “the First Amendment se­verely circumscribes the role that civil courts may play” in resolving ecclesiastical disputes, 65 Milivojevich, 426 U.S. at 709 (internal quotation marks omitted) (quoting Presbyterian Church, 393 U.S. at 449). they failed to specify the procedural nature of that limitation. In some cases, the Court spoke of church autonomy as depriv­ing civil courts of “jurisdiction” to decide religious questions. 66 See Milivojevich, 426 U.S. at 713–14 (“But it is a very different thing where a subject-matter of dispute, strictly and purely ecclesiastical in its character,—a matter over which the civil courts exercise no jurisdiction, . . . becomes the subject of its action.” (quoting Watson v. Jones, 80 U.S. (13 Wall.) 679, 733 (1871))); Kedroff v. Saint Nicholas Cathedral, 344 U.S. 94, 114 (1952) (referring to church autonomy in terms of “civil jurisdiction over church adjudications”). But the Court nevertheless decided those cases on the merits after accepting ec­clesiastical authorities’ answers to any religious questions. 67 In Milivojevich, for example, the Court found that a minister’s claim that the reor­ganization of the Serbian Orthodox Church’s American dioceses was procedurally and sub­stantively defective under church law raised a purely ecclesiastical question. See 426 U.S. at 708–10. Rather than dismiss the case for lack of jurisdiction, the Court accepted the Holy Synod of the Serbian Orthodox Church’s determination that church law permitted the re­organization and held on the merits that the minister’s claim was groundless. See id. at 721–23. The Court then reversed the state court’s judgment for the minister and remanded for further proceedings. See id. at 722–25; see also Jones, 443 U.S. at 608–10 (vacating a state court’s judgment in a church property dispute and remanding with instructions to accept an ecclesiastical authority’s decision as to possible religious questions); Presbyterian Church v. Mary Elizabeth Blue Hull Mem’l Presbyterian Church, 393 U.S. 440, 449–52 (1969) (re­versing a state court decision awarding church property to a local congregation over the general Presbyterian church and remanding for further proceedings). In other words, the Court’s jurisdictional language was at odds with its treatment of church autonomy in practice. The result was a decades-long split among lower federal courts and scholars as to whether the church-autonomy doctrine operates procedurally as a jurisdictional bar or an affirmative defense. 68 See Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 478–80 (de­scribing the split); see also Michael A. Helfand, Religion’s Footnote Four: Church Autonomy as Arbitration, 97 Minn. L. Rev. 1891, 1904–18 (2013) (detailing the development and en­trenchment of the jurisdictional conception of church autonomy).

The Supreme Court finally addressed the split over church auton­omy’s procedural character in 2012. 69 See Hosanna-Tabor Evangelical Lutheran Church & Sch. v. Equal Emp. Opportunity Comm’n, 565 U.S. 171, 195 n.4 (2012). In Hosanna-Tabor Evangelical Lutheran Church and School v. EEOC, the Court, for the first time, recog­nized the existence of a “ministerial exception” that “precludes application of [antidiscrimination] legislation to claims concerning the employment relationship between a religious institution and its minis­ters.” 70 See id. at 188. The Hosanna-Tabor Court grounded the ministerial exception in its church-autonomy precedents, casting a religious institution’s choice of minister as “strictly a matter of ecclesiastical government” with which civil courts cannot interfere. 71 Id. at 185–87 (internal quotation marks omitted) (quoting Kedroff, 344 U.S. at 115). The Court has since explicitly confirmed that the “constitutional foundation” for the min­isterial exception is “the general principle of church autonomy.” See Our Lady of Guadalupe Sch. v. Morrissey-Berru, 140 S. Ct. 2049, 2061 (2020). And in a footnote, the Court declared that the ministerial exception, and the broader church-autonomy doctrine by ex­tension, is an affirmative defense rather than a jurisdictional bar. 72 See Hosanna-Tabor, 565 U.S. at 195 n.4. The split that necessitated this declaration arose from the more general problem of “loose language about jurisdiction in [the Court’s] civil procedure cases.” Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 494. Professor Howard Wasserman thus aptly contextualizes the Court’s announcement in Hosanna-Tabor as part of its broader project in recent years of “clarify[ing] the line between jurisdiction and merits.” Howard M. Wasserman, The Roberts Court and the Civil Procedure Revival, 31 Rev. Litig. 313, 350 (2012).
         But as legal scholar Lael Weinberger has documented, that announcement has not led all state and lower federal courts to treat church autonomy as an affirmative defense. See Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 481–85. Some state courts continue to treat church autonomy as a jurisdictional bar as a matter of state civil procedure. Id. at 481–82. And “[m]ultiple federal courts have cabined the Hosanna-Tabor footnote to ministerial exception cases while treating other church autonomy cases as unaf­fected.” Id. at 483. As Weinberger notes, “The distinction is questionable, to say the least, given the Supreme Court’s clarification that church autonomy is the larger category within which the ministerial exception fits.” Id.

Though the Supreme Court has resolved the conflict over the proce­dural character of the church-autonomy doctrine, its announcement that the doctrine operates as an affirmative defense leaves many open ques­tions about the substantive scope of that defense. 73 See Mark E. Chopko & Marissa Parker, Still a Threshold Question: Refining the Ministerial Exception Post-Hosanna-Tabor, 10 First Amend. L. Rev. 233, 243 (2012). Most notably, the Court has yet to clarify whether church autonomy protects religious institutions from the burdens of discovery and trial or only from liability. That question has recently come to the fore in litigation over the immediate appealability of interlocutory decisions denying church-autonomy defenses.

B.      Church Autonomy as an Immunity Subject to Collateral-Order Review

In the wake of Hosanna-Tabor, several religious institutions, scholars, and judges have argued that, while the church-autonomy doctrine does not impose a jurisdictional bar, it provides more than a mere defense to liability. 74 See, e.g., Tucker v. Faith Bible Chapel Int’l, 53 F.4th 620, 625 (10th Cir. 2022) (Bacharach, J., dissenting from the denial of en banc consideration) (“[T]he ministerial exception protects a religious body from the suit itself.”); Defendants-Appellants’ Amended Petition for Rehearing En Banc at 10, Belya v. Kapral, 59 F.4th 570 (2d Cir. 2023) (No. 21-1498) [hereinafter Belya Defendants-Appellants’ Amended Petition for Rehearing En Banc] (“[T]he [church autonomy] doctrine provides a non-jurisdictionalimmunity against claims arising from internal church leadership disputes.”); Smith & Tuttle, supra note 10, at 1871–72 (“Establishment Clause limitations on the authority of courts to resolve religious ques­tions requires courts to treat the ministerial exception quite differently from ordinary affirmative defenses.”). These religious institutions and scholars view church autonomy as an immunity from suit akin to qualified or sovereign immunity, both of which operate procedurally as affirmative defenses. 75 See, e.g., Tucker v. Faith Bible Chapel Int’l, 36 F.4th 1021, 1049–51 (10th Cir. 2022) (Bacharach, J., dissenting) (arguing that, like qualified immunity and absolute immunity for government actors, “[t]he ministerial exception also advances values of a high order” and should protect religious bodies from litigation), cert. denied, 143 S. Ct. 2608 (2023); Belya Defendants-Appellants’ Brief, supra note 4, at 50–51 (arguing that church-autonomy questions are analogous to qualified immunity and should be resolved at the outset of liti­gation); Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 496–97 (noting that sovereign immunity and church autonomy both concern the ability of government to exercise authority over a “distinct and autonomous authority”). And like decisions denying those other immunities, the argument goes, interlocutory decisions deny­ing church-autonomy defenses are immediately appealable as collateral orders. 76 See Garrick v. Moody Bible Inst., 95 F.4th 1104, 1124 (7th Cir. 2024) (Brennan, J., dissenting); Belya, 59 F.4th at 578–80 (Park, J., dissenting from the order denying rehearing en banc); Smith & Tuttle, supra note 10, at 1878–81.

This “immunity theory” of church autonomy stems from concerns that, regardless of the outcome as to liability, judicial inquiry into ecclesi­astical matters unduly influences churches’ religious affairs. 77 See Belya Defendants-Appellants’ Brief, supra note 4, at 52–54 (arguing that, “once exposed to discovery and trial, the constitutional rights of the church to operate free of judicial scrutiny would be irreparably violated” (internal quotation marks omitted) (quoting United Methodist Church v. White, 571 A.2d 790, 793 (D.C. 1990))); Esbeck, supra note 10, at 266–67 (“[T]o allow the case to continue to be prepared for trial and fully tried on the merits is to reoffend the First Amendment with new church-state entanglements, and to do so in a manner that can never be corrected on appeal.”). That is because subjecting churches to litigation over “sensitive religious deci­sions”—thereby “[a]llowing ‘[c]hurch personnel and records’ to ‘become subject to subpoena, discovery, [and] cross-examination’”—could “pres­sure churches to base religious decisions on ‘avoid[ing] litigation or bureaucratic entanglement’ instead of ‘doctrinal assessments.’” 78 Garrick Appellant’s Brief, supra note 4, at 20–22 (second and fourth alterations in original) (quoting Rayburn v. Gen. Conf. of Seventh-Day Adventists, 772 F.2d 1161, 1171 (4th Cir. 1985)); see also Tucker, 53 F.4th at 627–28 (Bacharach, J., dissenting from the de­nial of en banc consideration) (arguing that “religious bodies will undoubtedly hesitate before deciding whether to suspend or fire renegade ministers” if the ministerial exception shields them only from liability but not from trial); Tucker Appellant’s Brief, supra note 4, at 44–46 (“Merits discovery and trial ‘inevitably affect’ future internal leadership decisions, pressuring the church to make them ‘with an eye to avoiding litigation . . . .’” (quoting Equal Emp. Opportunity Comm’n v. Cath. Univ. of Am., 83 F.3d 455, 466–67 (D.C. Cir. 1996))); Chopko & Parker, supra note 73, at 294 (noting the coercive impact of “[f]orcing . . . parties through years of expensive litigation, where churches may weary of the diversion of resources away from mission”). Though the Religion Clauses do not explicitly protect against the burdens of litiga­tion, 79 Compare the First Amendment’s restriction on laws “respecting an establishment of religion, or prohibiting the free exercise thereof,” U.S. Const. amend. I, with constitu­tional provisions recognized as creating immunities from suit, like the Speech or Debate Clause, U.S. Const. art. I, § 6, cl. 1 (“[F]or any Speech or Debate in either House, [members of Congress] shall not be questioned in any other Place.”), Double Jeopardy Clause, U.S. Const. amend. V (“[N]or shall any person be subject for the same offense to be twice put in jeopardy . . . .”), and the Eleventh Amendment, U.S. Const. amend. XI (“The Judicial power . . . shall not be construed to extend to any suit . . . against one of the United States . . . .”). proponents of the immunity theory un­derstand the First Amendment as a “structural limit” that prohibits any government influ­ence over churches’ religious affairs in order to preserve total ecclesiastical independence. 80 See Tucker, 36 F.4th at 1051–53 (Bacharach, J., dissenting) (“[U]nder the Establishment Clause, the ministerial exception serves as a structural limit on governmental power over religious matters.”); see also Belya Defendants-Appellants’ Amended Petition, supra note 74, at 13 (“[C]hurch autonomy is not merely a ‘personal’ defense, but ‘a struc­tural limitation’ . . . which ‘categorically prohibits’ judicial ‘involve[ment] in religious leadership disputes.’” (second alteration in original) (quoting Conlon v. InterVarsity Christian Fellowship, 777 F.3d 829, 836 (6th Cir. 2015))). On that view, church autonomy must protect religious institutions from the undue influence that results from judicial review of ecclesiastical questions. And immediate appeal must be available to vindi­cate that immunity when it is wrongly denied. 81 See Garrick, 95 F.4th at 1122–23 (Brennan, J., dissenting) (arguing that an appeal from a final judgment cannot remedy the litigation-related costs that the church-autonomy doctrine was designed to avoid); Belya, 59 F.4th at 578 (Park, J., dissenting from the order denying rehearing en banc) (same); Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 503–05 (“Interlocutory appeals from church autonomy are necessary to protect interests at the heart of the church autonomy doctrine.”).

To bolster their arguments, proponents of the immunity theory analogize church autonomy to qualified immunity. 82 See, e.g., Garrick, 95 F.4th at 1123–24 (Brennan, J., dissenting); Belya Defendants-Appellants’ Brief, supra note 4, at 51–52; Smith & Tuttle, supra note 10, at 1880–81. They describe church autonomy and qualified immunity as rooted in similar “foundational con­stitutional interests—the former the interest in precluding governmental intervention in religious disputes, and the latter in separation of powers concerns.” 83 Garrick, 95 F.4th at 1123–24 (Brennan, J., dissenting) (quoting Belya, 59 F.4th at 578–79 (Park, J., dissenting from the order denying rehearing en banc)); see also Joshua Lollar, Comment, Prayer for Relief: Church, Court, and Immediate Appeal in Ministerial Exception Cases, 73 U. Kan. L. Rev. 217, 250 (2024) (arguing that “both protect potential defendants from suit in service to larger social and legal necessities: coherent functioning of government with qualified immunity and respect for First Amendment rights with the ministerial exception”). Where qualified immunity protects gov­ernment officials from suit “because the costs of litigation ‘can be peculiarly disruptive of effective government,’” they argue that church autonomy protects religious institu­tions from litigation that could disrupt their internal governance and reli­gious decisionmaking. 84 Tucker, 36 F.4th at 1050 (quoting Mitchell v. Forsyth, 472 U.S. 511, 526 (1985)); see also Adam Reed Moore, A Textualist Defense of a New Collateral Order Doctrine, 99 Notre Dame L. Rev. Reflection 1, 43–44 (2023), https://scholarship.law.nd.edu/cgi/viewcon
tent.cgi?article=1142&context=ndlr_online [https://perma.cc/6X67-9UJ7] (arguing that the fear of discovery and trial process “would chill free exercise” by forcing religious groups to define their “religious” functions within the bounds of what “a secular court would likely agree to be religious”).
And both qualified immunity and church au­tonomy, the argument continues, are threshold issues to be resolved at the outset of a case. 85 Garrick, 95 F.4th at 1123–24 (Brennan, J., dissenting); Belya, 59 F.4th at 579–80 (Park, J., dissenting from the order denying rehearing en banc); Tucker Appellant’s Brief, supra note 4, at 46.

Proponents of the immunity theory also analogize church autonomy to sovereign immunity, casting both as “implicit in the constitutional de­sign.” 86 Belya, 59 F.4th at 580 n.4 (Park, J., dissenting from the order denying rehearing en banc) (internal quotation marks omitted) (quoting Alden v. Maine, 527 U.S. 706, 730 (1999)); see also Tucker, 36 F.4th at 1050 (Bacharach, J., dissenting) (comparing the minis­terial exception to defenses that protect “values of a ‘high order,’”including “Eleventh Amendment immunity” (quoting Will v. Hallock, 546 U.S. 345, 352 (2006))); Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 496–97 (noting the similarities be­tween the church-autonomy doctrine and sovereign immunity). Legal scholar Lael Weinberger has argued, for example, that “[b]oth sovereign immunity and church autonomy have a structural char­acter” and “speak to the authority of government . . . to exercise its authority over a distinct and autonomous authority (the state government in the sovereign immunity context, the religious institution in the church autonomy context).” 87 Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 496–97; see also Tucker, 36 F.4th at 1050 (Bacharach, J., dissenting) (comparing the purposes of church au­tonomy and sovereign immunity). As Jacquelyn Oesterblad has explained, this “ill-defined ‘structural limitation’ theory” makes little sense given the Supreme Court’s explanations that application of the ministerial exception (and the church-autonomy doctrine) requires incredibly fact-intensive inquiries. See Jacquelyn Oesterblad, Note, If You’re a Minister and You Know It, Clap Your Hands: Contract Nondiscrimination Clauses as a Voluntary Waiver of the Ministerial Exception, 41 Yale L. & Pol’y Rev. 282, 313–15 (2023). Church autonomy “is no more a special structural limitation than the Thirteenth Amendment is a structural limitation on the state’s power to enforce enslavement, or the First Amendment is a structural barrier to prior restraints on the press. These are all merits doctrines dealing with constitutional rights.” Id. at 316.

If one accepts the analogy of church autonomy to either qualified or sovereign immunity, it follows that church autonomy satisfies the require­ments for immediate appeal. 88 As explained in Part I, decisions denying constitutional immunities are necessarily conclusive, involve important questions separate from the merits, and are unreviewable on appeal from a final judgment. That applies to church autonomy as construed by the immun­ity theory. Under that theory, even when district courts simply defer consideration of church-autonomy defenses, they conclusively deny churches’ immunity from suit by subjecting the churches to discovery and motion practice in the meantime. See Belya, 59 F.4th at 577 (Park, J., dissenting) (arguing that, by subjecting a church to litigation, a district court conclusively determines the church’s immunity from suit); Tucker, 36 F.4th at 1058 (Bacharach, J., dis­senting) (same); Belya Defendants-Appellants’ Brief, supra note 4, at 55 (same). As with qualified and sovereign immunity, the argument continues, church autonomy is “im­portant” in serving core constitutional principles and provides a protection from litigation that is conceptually distinct from a church’s liability on the merits. See Garrick, 95 F.4th at 1121–22 (Brennan, J., dissenting) (“Because religious autonomy ‘lies at the foundation of our political principles,’ a decision on a claim implicating church autonomy involves an important question.” (citation omitted) (quoting Watson v. Jones, 80 U.S. (13 Wall.) 679, 728 (1871))). And if church autonomy offers immunity from suit, its loss is effectively unre­viewable upon later appeal: “If a church autonomy defense is erroneously denied and litigation allowed to proceed, there is no way to undo the interference with the religious institution that occurs simply by virtue of the litigation itself.” Weinberger, Is Church Autonomy Jurisdictional?, supra note 10, at 504; see also Belya, 59 F.4th at 577–78 (Park, J., dissenting from the order denying rehearing en banc) (“The denial of a church autonomy defense is . . . effectively unreviewable on appeal after final judgment.”). Proponents of the immunity theory thus argue that interlocutory decisions denying church-autonomy defenses are final for the purposes of § 1291 and are appealable as collateral orders.

C.       Church Autonomy as a Defense to Liability Subject to the Final-Judgment Rule

Despite enthusiastic support from religious institutions and scholars, arguments that church autonomy is an immunity subject to collateral-order review have met with little success in the federal courts. In two extraordinary instances, the Fifth and Seventh Circuits each permitted in­terlocutory appeals of orders denying church-autonomy defenses, but both circuits have subsequently cabined those decisions to their unique facts. 89 In McCarthy v. Fuller, the Seventh Circuit accepted an immediate appeal of a district court order refusing to accept the Holy See’s determination that the plaintiff was not a nun. See 714 F.3d 971, 973–74 (7th Cir. 2013). To resolve the plaintiff’s defamation claim against a defendant who had called her a “fake nun,” the district court planned to instruct the jury to decide whether the plaintiff was a nun in good standing with the Catholic Church. See id. at 974. The Seventh Circuit held that the order was immediately appealable because it “require[d] a jury to answer a religious question.” See id. at 976. But the following year, the Seventh Circuit clarified that McCarthy permits collateral-order review only in the rare case that a district court determines that a jury can preempt an ecclesiastical authority on a question of religious doctrine. See Herx v. Diocese of Fort Wayne–S. Bend, Inc., 772 F.3d 1085, 1091 (7th Cir. 2014). And the court recently recast the “extreme facts of McCarthy” as “best understood as satisfying those exacting requirements for a writ of mandamus,” rather than the requirements of the collateral-order doctrine. Garrick, 95 F.4th at 1114.
         The Fifth Circuit permitted an immediate appeal of a district court decision that erro­neously ordered discovery against a third-party religious organization in Whole Woman’s Health v. Smith, 896 F.3d 362, 367–68 (5th Cir. 2018). Because the religious organization was a nonparty that could not “benefit directly from” any post–final judgment relief, the discovery order was effectively unreviewable if not appealed immediately. Id. But the Fifth Circuit has since confined Whole Woman’s Health to the unique context of orders “allow[ing] discovery against a nonparty with substantial First Amendment implications.” Leonard v. Martin, 38 F.4th 481, 487 (5th Cir. 2022) (citing Whole Woman’s Health, 896 F.3d at 368); see also Vantage Health Plan, Inc. v. Willis-Knighton Med. Ctr., 913 F.3d 443, 450 n.2 (5th Cir. 2019) (characterizing Whole Woman’s Health as a “case involving the collateral order doctrine and a third-party document production order”).
And the Second, Seventh, and Tenth Circuits have all recently re­jected the immunity theory, albeit over considerable dissent. 90 See supra note 8. The circuits instead held that church autonomy is a mere defense to liability that is subject to the final-judgment rule. 91 See Garrick, 95 F.4th at 1115–16 (“[T]he doctrine of church autonomy [does not] confer immunity from trial in every employment discrimination suit.”); Belya v. Kapral, 45 F.4th 621, 633 (2d Cir. 2022) (stating that “[t]he church autonomy doctrine provides reli­gious associations neither an immunity from discovery nor an immunity from trial on secular matters” and “serves more as ‘an ordinary defense to liability.’”(quoting Herx, 772 F.3d at 1090)); Tucker, 36 F.4th at 1039–41 (“[Religious entities] are protected by the First Amendment, certainly, but are generally not excused from complying with generally appli­cable government regulation or from being haled into court.”).

In holding that church-autonomy decisions do not give rise to collat­eral orders, the Second, Seventh, and Tenth Circuits each rebuffed the analogy of church autonomy to qualified and sovereign immunity. But none of the three circuits grounded their analyses in historical under­standings of church autonomy or the Religion Clauses. The Second Circuit found the comparison inapt because “a district court’s order denying qual­ified immunity is an immediately appealable collateral order . . . only ‘to the extent that it turns on an issue of law,’” and the church-autonomy order at issue involved substantial questions of fact. 92 Belya, 45 F.4th at 633–34 (quoting Mitchell v. Forsyth, 472 U.S. 511, 530 (1985)). The Seventh and Tenth Circuits distinguished church autonomy as serving different ends than qualified or sovereign immunity because it protects “only private par­ties,” not “public officials or unit[s] of government.” 93 See Garrick, 95 F.4th at 1115–16; see also Tucker, 36 F.4th at 1040 (noting that “[i]mmunity from suit is a benefit typically only reserved for governmental officials” (alteration in original) (internal quotation marks omitted) (quoting Gen. Steel Domestic Sales, L.L.C. v. Chumley, 840 F.3d 1178, 1182 (10th Cir. 2016))). The Seventh and Tenth Circuits thus reasoned that, unlike those immunities, church auton­omy does not implicate “the separation of powers, the dignity interest of a State, the efficient operation of the government, or any other public inter­est” that would be imperiled by the litigation process. 94 Garrick, 95 F.4th at 1115–16 (quoting Herx, 772 F.3d at 1090); see also Tucker, 36 F.4th at 1040 (“[The] rationales underlying qualified immunity—‘to safeguard government, and thereby to protect the public at large’—‘are not transferable to private parties.’” (quot­ing Wyatt v. Cole, 504 U.S. 158, 168 (1992))). And, both circuits noted, those are the only sorts of “higher interests” the Supreme Court has said “ought to be protected” by the collateral-order doctrine. 95 Garrick, 95 F.4th at 1115 (citing Herx, 772 F.3d at 1090); see also Tucker, 36 F.4th at 1026 (“[T]he reason that the Supreme Court permits immediate appeals from the denial of qualified immunity is to protect, not individual government officials, but rather the pub­lic’s interest in a functioning government. That public interest is not present [in suits against] a private religious employer . . . .”).

Heeding the Supreme Court’s thirty years of “increasingly emphatic instructions” to keep the class of collateral orders “‘small,’ ‘modest,’ and ‘narrow,’” 96 Tucker, 36 F.4th at 1034 (internal quotation marks omitted) (quoting Kell v. Benzon, 925 F.3d 448, 452 (10th Cir. 2019)); see also Garrick, 95 F.4th at 1110 (“The Supreme Court has repeatedly emphasized that the collateral order doctrine is limited and narrow. Its instructions are emphatic . . . .” (citations omitted)); Belya, 45 F.4th at 629 (“[T]he Supreme Court has admonished that ‘the class of collaterally appealable orders must remain “narrow and selective in its membership.”’” (quoting Mohawk Indus., Inc. v. Carpenter, 558 U.S. 110, 113 (2009))). the Second, Seventh, and Tenth Circuits held that—because church autonomy protects religious institutions only from liability for their religious decisions, not from the burdens of discovery and trial—decisions denying church-autonomy defenses are not immediately appealable as col­lateral orders. 97 None of the interlocutory orders at issue in these cases were “conclusive” because each district court had indicated that the defendant could raise the church-autonomy de­fense again at a later stage of the litigation, and no right to be free from the judicial process was lost in the meantime. See Garrick, 95 F.4th at 1114–15 (stating that the defendant could raise the church-autonomy defense “at later stages of litigation, including summary judg­ment”); Belya, 45 F.4th at 631 (“[T]he [district court’s] orders are not conclusive because they do not bar any defenses, they did not rule on the merits of the church autonomy de­fense, and they permit Defendants to continue asserting the defense.”); Tucker, 36 F.4th at 1047 (finding that the district court’s decision “clearly contemplates further factual proceedings” and is therefore inconclusive). The Tenth Circuit found a defendant’s ministerial exception defense to be conceptually distinct from the merits of the underlying employment discrimination claim. See Tucker, 36 F.4th at 1036. But the Second and Seventh Circuits held that defendants’ church-autonomy defenses were inseparable from the merits, as resolving the defenses would require answering factual questions that were also material to the plaintiffs’ underlying defamation and discrimination claims. See Garrick, 95 F.4th at 1115; Belya, 45 F.4th at 632. And because the church-autonomy defenses protected only against liability, their erroneous denial could be adequately vindicated after final judgment by reversing any damages award against the defendants. Garrick, 95 F.4th at 1115–16; Belya, 45 F.4th at 633; Tucker, 36 F.4th at 1036–37.

III. Treating Church Autonomy as an Immediately Appealable Immunity Contravenes Historical Understandings of Church–State Relations

In rejecting the argument that church autonomy is an immunity sub­ject to collateral-order review, the Second, Seventh, and Tenth Circuits largely failed to engage with historical views of the nature and extent of religious institutions’ independence from civil government. 98 See supra section II.C (describing the Second, Seventh, and Tenth Circuits’ reasoning). But the Supreme Court has instructed that the Religion Clauses’ protections “must be interpreted by ‘reference to historical practices and understandings.’” 99 Kennedy v. Bremerton Sch. Dist., 142 S. Ct. 2407, 2428 (2022) (quoting Town of Greece v. Galloway, 572 U.S. 565, 576 (2014)). This Part argues that, though missing from the circuit courts’ analyses, understandings of church–state relations in the early Republic support their conclusion that church autonomy is not an immunity from suit. Allowing interlocutory appeals of church-autonomy defenses would thus contra­vene historical practice. To comport with historical understandings, church autonomy should instead be treated as a deference doctrine that only protects against liability and does not give rise to collateral orders.

As Professors Sarah Barringer Gordon and Kellen Funk have both documented, “extensive legislative and judicial oversight of churches and other religious organizations” was commonplace in the early Republic. 100 Sarah Barringer Gordon, The First Disestablishment: Limits on Church Power and Property Before the Civil War, 162 U. Pa. L. Rev. 307, 311 (2014); see also Kellen Funk, Church Corporations and the Conflict of Laws in Antebellum America, 32 J.L. & Religion 263, 266–70 (2017) (describing eighteenth-century state legislation regulating the incorporation of churches). Most states made incorporation generally available to religious societies soon after disestablishing their official churches, starting with New York in 1784. 101 See Funk, supra note 100, at 266–68 (“[T]he major principles of [New York’s] sys­tem were eventually adopted in every American state during the antebellum era.”). Because of the immense benefits of the corporate form, which al­lowed a church to secure “a perpetual legal existence that could hold and transfer property and defend its rights in state courts,” churches incorpo­rated en masse. 102 Id. at 266; see also Gordon, supra note 100, at 316 (“In most American jurisdic­tions, religious corporations were ubiquitous by the early nineteenth century . . . .”). In return for these benefits, state legislatures placed strict limits on the amount of land and property churches could own. 103 See Gordon, supra note 100, at 323–24. Legislatures also “imposed regimes of lay governance” to democratize churches, placing “[c]ontrol of all property and funds . . . in the hands of congregants, not clergy.” 104 Id. at 324; see also Funk, supra note 100, at 269–70 (explaining that states’ general incorporation statutes “prevented the concentration of land, power, and allegiance in a sin­gle [church] corporation”). And states required that churches incorporate at the local congregational level, “functionally transform[ing] all denom­inations into congregational polities” regardless of their allegiance to a central religious authority. 105 Funk, supra note 100, at 269 (citing Mark deWolfe Howe, The Garden and the Wilderness: Religion and Government in American Constitutional History 44–47 (1965)).

The purpose of this extensive regulation was to prevent churches from operating like sovereign governments. Once states began to disestab­lish their official churches in the mid-1770s, commentators worried that newly independent churches would become rival sovereigns that could threaten the states. 106 Id. at 269–70; see also Gordon, supra note 100, at 315–16 (“To protect individual liberty, churches were constrained in their capacities to acquire wealth and broadly sub­jected to lay control.”). The risk of such a destabilizing “imperium in imperio,” or “state within the state,” was believed to stem from a “concen­tration of property and privilege that could insulate an elite [ecclesiastical] class from the will of the sovereign people.” 107 Funk, supra note 100, at 269–70; see also Alison L. LaCroix, The Ideological Origins of American Federalism 81 (2010) (“The specter of imperium in imperio would haunt colonial, and later national, political debate at least until 1789 . . . .”). States aimed to mitigate that risk by treating churches as ordinary corporations, rather than as autono­mous sovereigns, and by restricting church corporations’ material wealth and governance structures. 108 See Gordon, supra note 100, at 323–24 (“In return for the protections of the cor­porate form, these statutes paired such limits on wealth and land with other forms of regu­lation, especially mandatory forms for internal governance.”); see also Funk, supra note 100, at 269–70 (describing how legislative caps on “how much land or revenue a church corpo­ration could command” served to prevent the concentration of ecclesiastical power).

During the early Republic, courts were active participants in this reg­ulatory regime, enforcing limitations on religious corporations. “Frequent judicial interpretation of those laws meant that the internal workings of religious organizations were exposed to scrutiny and judgment in thou­sands of conflicts that pitted the faithful against each other and against their ministers and priests.” 109 Gordon, supra note 100, at 320; see also id. at 342–43 (“The amount of litigation is staggering. At every turn, quarrels devolved into arguments over church polity, the rights of congregants, the disposition of church property, and the standing of ministers.”). And “judges often inquired into religious doctrine to decide questions of church polity” or to interpret cor­porate charters in disputes involving church property. 110 Id. at 320; see also Funk, supra note 100, at 273 (noting that, after disestablishment, “state courts ever more invasively scrutinized church doctrine and personal belief”). As a practical matter, then, churches in the early Republic were hardly immune from litigation concerning their religious affairs.

The Supreme Court stamped its approval on the view of churches as private corporations in 1819. In Trustees of Dartmouth College v. Woodward, the Court confronted efforts by the New Hampshire legislature to alter the charter of Dartmouth College—a private religious institution—to exert greater state control over the college. 111 17 U.S. (4 Wheat.) 518, 539–49 (1819). The Court reasoned that New Hampshire’s acts could be constitutional only if Dartmouth functioned as a part of the civil government, as the state would then be “unrestrained by any limitation . . . imposed by the constitution.” 112 Id. at 629–30. But the Court rejected this view of chartered religious organizations due to implicit fears that churches might become rival sovereigns. “The unstated assumption,” Funk has explained, was that treating chartered religious organizations as arms of the state “would create the monstrous imperium in imperio” by ena­bling those religious organizations to exercise state governance powers. 113 Funk, supra note 100, at 272 (citing Trs. of Dartmouth Coll., 17 U.S. (4 Wheat.) at 630). The Court instead held that chartered religious organizations were ordinary private corporations, so New Hampshire’s efforts to alter Dartmouth’s charter violated the Contract Clause. 114 See Trs. of Dartmouth Coll., 17 U.S. (4 Wheat.) at 635–39, 654.

Given this historical understanding of religious institutions as regula­ble private corporations, treating church autonomy as an immunity from suit subject to collateral-order review is inappropriate. As discussed in Part I, such immunities typically (1) stem from an explicit statutory or consti­tutional guarantee or (2) serve the public interest in preserving the effec­tive functioning of or proper allocation of power within government. 115 See supra Part I. The Religion Clauses provide no express protection from the burdens of litigation. 116 See supra note 79. For church autonomy to constitute an immunity such that de­nial warrants immediate appeal, it must therefore fall in the second category. Indeed, that is why proponents of the immunity theory analogize church autonomy to qualified and sovereign immunity, which serve im­portant public interests in effective government, the separation of powers, and the dignitary interests of sovereign states. 117 See supra section II.B. But church autonomy protects only private religious institutions—it cannot serve public interests in the operation and structure of government unless those private institu­tions are seen as arms of the state. Put another way, treating church autonomy like qualified or sovereign immunity makes sense only if churches are themselves sovereign. That directly contravenes historical un­derstandings of church–state relations from the disestablishment period and vitiates the logic of Dartmouth College. 118 Recent scholarship arguing that the history of church–state relations in the early Republic compels treating church autonomy as an immunity subject to collateral-order re­view overlooks this essential point.
         Weinberger, for example, argues that the pervasive regulation of church corporations in the disestablishment period does not undermine the historical basis for treating church autonomy as an immunity because churches could avoid such regulations by choosing not to incorporate. See Lael Weinberger, The Origins of Church Autonomy: Religious Liberty After Disestablishment 34–37 (Feb. 4, 2025) (unpublished manuscript), https://ssrn.com/abstract=4933864 [https://perma.cc/W98W-G2WB]. As such, he con­tends, regulations of church corporations are not definitive evidence of an intent to deprive religious defendants of privileged status. Id. To the contrary, “[w]here the legislature had not spoken, the courts [of the early Republic] would try to give churches maximum freedom of self-government” and “leave them as untouched by civil law as possible.” Id. at 36. In Weinberger’s view, that “suggest[s] that religious liberty concerns require that churches be protected from the burdens of litigation” and, in turn, “that [church autonomy] should be treated akin to modern immunity doctrines” with procedural features like the availability of interlocutory review. Id. at 48–49. But the fact that courts privileged churches to some de­gree does not mean that churches enjoyed the kind of sweeping immunity afforded to sov­ereign states or their officers. Weinberger does not explain why treating church autonomy as a deference doctrine short of full immunity does not accord with early American courts’ conception of religious liberty. Nor, more importantly, does he account for the conflict be­tween extending sovereign-like immunity to churches and the “democratic desire to restrain the power of churches in the nineteenth century” to prevent an imperium in imperio, which he acknowledges motivated much of the regulation of churches by legislatures. See id. at 33 (acknowledging the restraint of ecclesiastical power as “one of the threads” animating nineteenth-century regulations of church governance). As Funk has documented, that mo­tivation was shared by courts. See Funk, supra note 100, at 271–72 (“The Court [in Dartmouth College] was . . . concerned that religious societies . . . already looked too much like states, performing public services and exercising governmental functions.”).
         Branton Nestor similarly argues that nineteenth-century courts understood church-autonomy principles to limit judicial inquiry into ecclesiastical matters. See Branton J. Nestor, Judicial Power and Church Autonomy, 100 Notre Dame L. Rev. (forthcoming) (man­uscript at 28–30), https://ssrn.com/abstract=4818612 [https://perma.cc/8ALH-9SVX]. As Nestor details, such limitations “were explained in both institutional-deference and judicial-competence terms.” Id. at 31. The former expressed “the view that civil courts must defer to competent religious institutions over matters in their domain” to prevent undue interfer­ence in churches’ internal affairs, while the latter expressed “the view that civil courts lack competence of religious questions.” Id. at 29–30 (describing federal courts’ approaches); see also id. at 35–36 (describing state courts’ approaches). Nestor contends that, like the interests protected by sovereign and official immunities, these historic interests in prevent­ing interference and guarding against judicial incompetence require allowing immediate appeal of interlocutory church-autonomy decisions. See id. at 64–70. But like Weinberger, he fails to reconcile his argument for affording churches a procedural protection usually reserved for sovereign states and their agents with the widespread fears of the imperium in imperio in the early Republic and legislatures’ and courts’ consequent reticence to extend sovereign-like protections to churches. To the contrary, despite repeatedly acknowledging these concerns, see id. at 16–18, 38, 60, 74, 76, 86, Nestor directly contradicts them, arguing for collateral-order review of church-autonomy decisions by insisting that churches should be treated like sovereigns. See id. at 65 (arguing that judicial inquiry into church autonomy may implicate a “‘sovereignty’ problem,” analogous to “subjecting a sovereign state to suit,” in which a reviewing court “violat[es] the sovereignty or authority of the entity by subjecting it to suit and review”).
         Further, it is not clear why a deference theory of church autonomy—which would not permit collateral-order appeals—cannot adequately serve the goals Nestor reads into the nineteenth century cases. Deference doctrines often serve to prevent undue judicial inter­ference. See, e.g., Cuker v. Mikalauskas, 692 A.2d 1042, 1046 (Pa. 1997) (explaining that “[t]he business judgment rule ‘expresses a sensible policy of judicial noninterference with business decisions’” (quoting Rosenfield v. Metals Selling Corp., 643 A.2d 1253, 1262 (Conn. 1994))). Deference doctrines are also said to account for a lack of judicial competence over certain areas. See Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 865 (1984) (explaining that deference to agency interpretations of statutes is warranted because “[j]udges are not experts in the field”), overruled by Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244 (2024); cf. Loper Bright, 144 S. Ct. at 2266 (explaining that deference to agencies is not appropriate, in part, because “agencies have no special competence in resolving stat­utory ambiguities” and “[c]ourts do”). And as Nestor himself notes, early articulations of church autonomy “sound[] in . . . deference . . . to modern ears.” Nestor, supra, at 23; see also id. at 30 (noting that Watson v. Jones “provid[es] robust support for the institutional-deference account of the church-autonomy doctrine’s procedural protections”).

The better approach is to treat church autonomy as a deference doc­trine that bars courts only from overruling ecclesiastical authorities’ answers to religious questions. While church autonomy compels courts to defer to churches’ religious determinations, it permits the judiciary to re­solve any remaining secular issues. 119 See Frederick Mark Gedicks, The Religious-Question Doctrine: Free-Exercise Right or Anti-Establishment Immunity? 8 (Robert Schuman Ctr. for Advanced Stud., Working Paper No. RSCAS 2016/10, 2016), https://ssrn.com/abstract=2746593 [https://perma.cc/425D-YTFQ] (explaining that “[t]he religious-question doctrine does not prohibit a court from adjudicating a case in which a religious question is presented, but only from adjudicating such a case by answering the religious question”); see also Matthew R. Goldammer, Protecting Church Autonomy in the Twenty-First Century: A Defense of the Compulsory Deference Approach for Church Property Litigation, 37 Notre Dame J.L., Ethics & Pub. Pol’y 11, 21–25, 29–30 (2023) (describing Watson and its progeny as creating a regime of “compulsory deference” to ecclesiastical authorities’ religious determinations). In other words, it does not immunize churches from the judicial process. Many of the Supreme Court’s discus­sions of church autonomy support this view. 120 In Watson v. Jones, the Court instructed that “[w]hen a civil right depends upon an ecclesiastical matter, it is the civil court and not the ecclesiastical which is to decide. But the civil tribunal tries the civil right, and no more, taking the ecclesiastical decisions out of which the civil right arises as it finds them.” 80 U.S. (13 Wall.) 679, 731 (1871) (internal quotation marks omitted) (quoting Harmon v. Dreher, 17 S.C. Eq. (Speers Eq.) 87, 87 (S.C. Ct. App. Eq. 1843)); see also id. at 734 (explaining that the lower court’s error was in “substituting its own judgment for that of the ecclesiastical court”). The Court has subsequently repeated that command for courts to “accept [ecclesiastical] decisions as binding on them, in their application to the religious issues of doctrine or polity before them.” Serbian E. Orthodox Diocese v. Milivojevich, 426 U.S. 696, 709 (1976); see also Presbyterian Church v. Mary Elizabeth Blue Hull Mem’l Presbyterian Church, 393 U.S. 440, 449 (1969) (describing the church autonomy doctrine as “command[ing] civil courts to decide church property dis­putes without resolving underlying controversies over religious doctrine”). And the Court’s decision in Jones v. Wolf explicitly framed the church autonomy doctrine in terms of defer­ence, explaining that it “requires that civil courts defer to the resolution of issues of religious doctrine or polity by the highest court of a hierarchical church organization.” 443 U.S. 595, 602 (1979) (emphasis added) (citing Milivojevich, 426 U.S. at 724–25); see also id. at 604 (“[T]he court must defer to the resolution of the doctrinal issue by the authoritative eccle­siastical body.” (citing Milivojevich, 426. U.S. at 709)). And because it parallels the business judgment rule, under which courts defer to corporate directors’ decisions, 121 See Dodge v. Ford Motor Co., 170 N.W. 668, 682 (Mich. 1919) (restating the “well-recognized principle of law” that corporate directors alone “have the power to declare a dividend of the earnings of the corporation” and courts cannot interfere unless there is evidence of fraud or misappropriation of corporate funds (quoting Hunter v. Roberts, Throp & Co., 47 N.W. 131, 134 (Mich. 1890))). it reflects the historical understanding of churches as private corporations. 122 Cf. Watson, 80 U.S. (13 Wall.) at 723–25 (explaining that “the general doctrine of courts . . . as to charities” is “equally applicable to ecclesiastical matters” and that “the rights of such bodies . . . must be determined by the ordinary principles which govern voluntary associations”); Funk, supra note 100, at 273 n.49 (remarking that an 1832 church-property case “reads like an early instance of the modern business-judgment rule”).
     Alternatively, as Professor Michael Helfand has argued, church autonomy can be analogized to the deference owed by courts to arbitrators. See Helfand, supra note 68, at 1918–51.

Whatever the best analogy for the church autonomy doctrine, con­struing its protections as an immunity from suit akin to sovereign or qualified immunity cannot be squared with historical understandings of church–state relations in the early Republic. Decisions denying church-autonomy defenses should thus not be immediately appealable as collateral orders.

Conclusion

The Second, Seventh, and Tenth Circuits were right to reject imme­diate appeals of interlocutory decisions denying church-autonomy defenses. In light of the Supreme Court’s warnings against the expansion of the collateral-order doctrine, only the clearest immunities from suit war­rant creating a new category of collateral orders. Church autonomy does not fit the bill. Historical practice in the early Republic, when churches were often subjected to extensive legislative and judicial interference in their religious affairs, repudiates the notion that churches were immun­ized against the burdens of litigation. And treating church autonomy as the kind of immunity eligible for collateral-order review tacitly treats reli­gious institutions as units of government, creating the potential for ecclesiastical states-within-the-state that cases like Dartmouth College aimed to prevent. The more historically grounded approach is to view church autonomy as a deference doctrine that cannot give rise to collateral-order appeals.