Issue 8

Presidential Adjudication

Over the last several decades, administrative law has recognized an expanding role for the President in controlling agency decision-making. Agency adjudication—and especially formal hearings conducted under the Administrative Procedure Act (“APA”)—have been viewed as properly insulated from this development. To protect due process, the APA established a regime for ensuring that competent, impartial Administrative Law Judges (“ALJs”) preside over formal hearings. The regime includes two apparent levels of for-cause removal protection for ALJs combined with robust agency head control over the policymaking aspects of formal adjudication. Today, the regime is in peril because it appears to be inconsistent with the Supreme Court’s unitary executive theory of administration.

This Article defends the constitutionality of the APA’s ALJ regime under the Supreme Court’s recent separation of powers cases. It argues that the APA’s robust preservation of agency head control satisfies Article II, while its for-cause protections for ALJs ensure due process and faithful execution of the law through adjudicatory hearings. The statute is, in short, well-designed to ensure properly presidential adjudication.

The Article further argues, however, that there is a deeper conceptual challenge lurking here. The APA and the administrative state were founded upon a New Deal-era conception of administrative power as quasi-legislative and quasi-judicial and fundamentally not executive. Modern administrative law has rejected this conception, embracing instead the view that administrative power necessarily entails the exercise of executive power. The current threat to the APA offers an opportunity to improve upon this conception by recognizing that administration is about both discretion and duty. Political control has its place. But the President must also be able to rely on subordinate officers that Congress has equipped with the legal and institutional support necessary to fairly and faithfully execute the law.

Introduction

The Administrative Procedure Act’s (“APA”) most central reform—its regime for ensuring competent, impartial presiding officers in adjudicatory hearings—is on a collision course with the Supreme Court’s recent separation of powers jurisprudence. In peril is the APA’s structure for empowering and protecting Administrative Law Judges (“ALJs”), who preside over administrative hearings and issue initial decisions that may become final in the absence of agency head review.1 1.See 5 U.S.C. §§ 556(b)–(c), 557(b).Show More The primary threat to the regime is Free Enterprise Fund v. PCAOB,2 2.561 U.S. 477 (2010).Show More a 2010 case in which the Supreme Court held that “multilevel protection from removal is contrary to Article II’s vesting of the executive power in the President.”3 3.Id. at 484.Show More Seemingly like the structure at issue in Free Enterprise Fund, the APA’s ALJ structure entails “multilevel protection from removal.”4 4.Id.Show More ALJs can be removed from office only for cause, which is determined by the Merit Systems Protection Board (“MSPB”),5 5.See 5 U.S.C. § 7521.Show More the members of which likewise can be removed only for cause.6 6.See id. § 1202(d).Show More The situation is further complicated when ALJs are employed by independent agencies such as the Securities and Exchange Commission (“SEC”), which are headed by multimember bodies whose members likely enjoy for-cause removal protection.7 7.In Free Enterprise Fund, the parties and the majority assumed that SEC commissioners can be removed only for cause, although the SEC’s organic statute contains no for-cause provision. See 561 U.S. at 487; see also id.at 545–46 (Breyer, J., dissenting, joined by Stevens, Ginsburg & Sotomayor, JJ.) (noting that the majority “assume[d] without deciding that . . . SEC Commissioners . . . are removable only ‘for cause’” (emphasis omitted)). SEC commissioners are, however, appointed for a term of years, see 15 U.S.C. § 78d(a), which perhaps should be interpreted as a protection against removal by the President for the duration of the term. See Jane Manners & Lev Menand, The Three Permissions: Presidential Removal and the Statutory Limits of Agency Independence, 121 Colum. L. Rev. 1, 5 (2021). But see Severino v. Biden, 71 F.4th 1038, 1045–46 (D.C. Cir. 2023) (rejecting this approach to interpreting a statutory term of years).Show More The principle of Free Enterprise Fund would pose no threat to the APA if ALJs were mere employees, but the Supreme Court foreclosed this possibility in 2018, when it held in Lucia v. SEC that SEC ALJs are “Officers of the United States.”8 8.U.S. Const. art. II, § 2, cl. 2; Lucia v. SEC, 138 S. Ct. 2044, 2049 (2018). Although Lucia dealt only with ALJs employed by the SEC, its holding likely reaches ALJs employed by other agencies because of the similarity of functions. See Guidance on Administrative Law Judges After Lucia v. SEC (S. Ct.), July 2018, 132 Harv. L. Rev. 1120, 1122–23 (2019).Show More In Jarkesy v. SEC,9 9.34 F.4th 446 (5th Cir. 2022), aff’d on other grounds, 144 S. Ct. 2117, 2127, 2139 (2024).Show More the U.S. Court of Appeals for the Fifth Circuit held that the multilevel removal protection provided by the APA’s ALJ structure is unconstitutional under Free Enterprise Fund.10 10.Id. at 464.Show More Although the Supreme Court granted certiorari on this question, it affirmed the Fifth Circuit on alternative grounds.11 11.See Jarkesy, 144 S. Ct. at 2127, 2139; see also Petition for a Writ of Certiorari at I, Jarkesy, 144 S. Ct. 2117 (No. 22-859) (presenting several questions, including “[w]hether Congress violated Article II by granting for-cause removal protection to administrative law judges in agencies whose heads enjoy for-cause removal protection”).Show More But the Supreme Court will have to address the removal question eventually. The APA’s day of reckoning has only been delayed.

Lurking beneath the surface of this controversy is a more fundamental conflict: in recent decades, the Supreme Court has developed a conception of administrative action fundamentally at odds with that which prevailed in the New Deal era and animated the APA. As I have argued in prior work, the APA is based on a conception of administrative action as exclusively quasi-legislative and quasi-judicial, and fundamentally not executive.12 12.See Emily S. Bremer, The Rediscovered Stages of Agency Adjudication, 99 Wash. U. L. Rev. 377, 436–47 (2021) [hereinafter Bremer, Rediscovered Stages].Show More Although ordinarily associated with the Supreme Court’s 1935 decision in Humphrey’s Executor v. United States,13 13.295 U.S. 602 (1935).Show More which involved the constitutionality of for-cause removal protection for commissioners of the FTC,14 14.Id. at 608.Show More this understanding of administrative power was not confined to independent regulatory commissions. To the contrary, this understanding applied to all statutory grants of quasi-legislative (rulemaking) or quasi-judicial (adjudication) power, whether made to an independent agency or a traditional executive department.15 15.See, e.g., Ariz. Grocery Co. v. Atchison, 284 U.S. 370, 389 (1932) (same but involving the Interstate Commerce Commission); Morgan v. United States, 298 U.S. 468, 477, 481–82 (1936) (applying the New Deal conception of “administrative” action to ratemaking conducted by the Department of Agriculture).Show More Indeed, the Attorney General’s Committee on Administrative Procedure—which conducted the extensive research that provided the APA’s “intellectual foundation”16 16.K.C. Davis, Walter Gellhorn & Paul Verkuil, Present at the Creation: Regulatory Reform Before 1946, 38 Admin. L. Rev. 511, 514 (1986). I have examined this research in detail in previous work. See Emily S. Bremer, The Undemocratic Roots of Agency Rulemaking, 108 Cornell L. Rev. 69, 90–93 (2022); Bremer, Rediscovered Stages, supra note 12, at 396–402. The relevant documents are available electronically in The Bremer-Kovacs Collection: Historic Documents Related to the Administrative Procedure Act of 1946, https://heinonline.‌org/HOL/Index?collection=bremer. See generally Emily S. Bremer & Kathryn E. Kovacs, Essay, Introduction to The Bremer-Kovacs Collection: Historic Documents Related to the Administrative Procedure Act of 1946 (HeinOnline 2021), 106 Minn. L. Rev. Headnotes 218 (2022) (offering a narrative introduction to the collection).Show More—employed this conception to scope its study.17 17.See Off. of the Att’y Gen., Final Report of the Attorney General’s Committee on Administrative Procedure 2–4 (1941) [hereinafter Final Report]. The Final Report is based on 27 monographs examining the procedures and practices of “administrative” agencies. See id. at 3–4. Purely “executive” agencies—such as the Government Printing Office, the Bureau of Standards, the Civil Service Commission (“CSC”), the Bureau of the Budget, and the General Accounting Office—were left out of the study. See id. at 5.Show More Only agencies that were “administrative” in the New Deal sense were included. This choice left an indelible mark on the APA, which regulates binding agency action according to the mutually exclusive categories of adjudication and rulemaking.18 18.See 5 U.S.C. § 551(4)–(7). “This particular line may be the APA’s most important innovation.” Kristin E. Hickman & Aaron L. Nielson, Narrowing Chevron’sDomain, 70 Duke L.J. 931, 942 (2021). It was inspired by—but “not on all fours with”—the pre-APA definitions of quasi-judicial and quasi-legislative action. Emily S. Bremer, Blame (or Thank) the Administrative Procedure Act for Florida East Coast Railway, 97 Chi.-Kent L. Rev. 79, 96–97 (2022) [hereinafter Bremer, Blame (or Thank)].Show More

When the APA was enacted in 1946, most administrative action was adjudication, and the statute’s primary aim was to address the constitutional challenges presented by this quasi-judicial form of agency action.19 19.The legislature was also influenced by concerns—made concrete by the World Wars and related political developments in Europe—about how to ensure effective administration without facilitating authoritarianism. See, e.g., Kathryn E. Kovacs, Avoiding Authoritarianism in the Administrative Procedure Act, 28 Geo. Mason L. Rev. 573, 596–600 (2021); Noah A. Rosenblum, The Antifascist Roots of Presidential Administration, 122 Colum. L. Rev. 1, 14, 48–49 (2022).Show More At the time, adjudication was understood as a staged or “phase[d]” process.20 20.See Final Report, supra note 17, at 5.Show More (Modern administrative law has forgotten this, although adjudication today retains its staged structure.21 21.See Bremer, Rediscovered Stages, supra note 12, at 433.Show More) The initial stage of adjudication involves myriad informal, non-hearing techniques such as investigations, inspections, examinations, conferences, negotiations, and settlements.22 22.See id. at 402–03.Show More In the relatively rare instances in which these techniques are insufficient to resolve a matter with the affected private party’s consent,23 23.See Final Report, supra note 17, at 35–38, 41–42.Show More a judicial-type hearing might be required to resolve the dispute. Congress often prefers that the needed hearing be conducted by the agency—rather than by a court on judicial review—and so includes a hearing requirement in the agency’s governing statute.24 24.Bremer, Rediscovered Stages, supra note 12, at 431.Show More This approach ensures the agency’s primary jurisdiction, but it presents significant constitutional challenges, threatening due process as a matter of both separation of powers and individual rights.25 25.For an originalist discussion of the relationship between due process and separation of powers, see Nathan S. Chapman & Michael W. McConnell, Due Process as Separation of Powers, 121 Yale L.J. 1672, 1677–78 (2012). The sovereign power and individual rights aspects of due process are also observable, for example, in personal jurisdiction doctrine. See, e.g., Thomas D. Rowe, Jr., Suzanna Sherry & Jay Tidmarsh, Civil Procedure 452 (5th ed. 2020) (“The Court has wavered about whether personal-jurisdiction doctrine rests on individual liberty or state sovereignty (or both).”).Show More The need to address these challenges was the driving force behind the APA. The hearing provisions enacted by Congress achieved that goal by establishing a default procedural regime intended to apply across all adjudicating agencies.26 26.The APA’s procedures are a default because they apply unless Congress affirmatively elects to displace them. See 5 U.S.C. §§ 556(b), 559.Show More

In the three-quarters of a century since the APA’s adoption, rulemaking has become central to administration, working an inevitable change on the dominant conception of “administrative” power. Beginning in the 1960s and ’70s, rulemaking began to displace adjudication as the preferred method of agency policymaking,27 27.See, e.g., Antonin Scalia, Vermont Yankee: The APA, the D.C. Circuit, and the Supreme Court, 1978 Sup. Ct. Rev. 345, 376 (describing “the constant and accelerating flight away from individualized, adjudicatory proceedings to generalized disposition through rulemaking”); see also Alan B. Morrison, The Administrative Procedure Act: A Living and Responsive Law, 72 Va. L. Rev. 253, 254–55 (1986); Ralph F. Fuchs, Development and Diversification in Administrative Rule Making, 72 Nw. U. L. Rev. 83, 89 (1977) (noting the “growing tendency” of many agencies to confront policy issues via rulemaking rather than adjudication).Show More and Congress created a host of new agencies with broad statutory mandates to protect public health and safety through rules. This shift in turn heralded the rise of “presidential administration”28 28.See generally Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245 (2001) (analyzing the President’s recent primacy in setting the direction of administrative process); cf. Michael A. Livermore, Political Parties and Presidential Oversight, 67 Ala. L. Rev. 45, 53–61 (2015) (describing how executive restructuring, presidential control in administrative law, and evolution in the operation of American political parties have all contributed to increase the President’s influence in federal administration). For an early warning that this development might disrupt the carefully mediated tension between law and politics in administrative law, see Peter L. Strauss, Presidential Rulemaking, 72 Chi.-Kent L. Rev. 965, 984–86 (1997).Show More by giving presidents a “grip” on agency policymaking that was elusive when agencies primarily made policy incrementally, through ad hoc adjudication.29 29.See Emily S. Bremer, Power Corrupts, 41 Yale J. on Regul. 426, 456–58 (2024) [hereinafter Bremer, Power Corrupts].Show More In response to these developments, the Supreme Court’s administrative law docket increasingly focused on policymaking undertaken pursuant to statutes that grant broad discretion and contemplate a central role for rulemaking.30 30.The Chevron doctrine—according to which courts would defer to reasonable agency interpretations of the statutes by which Congress has delegated power to them—was both cause and consequence of this change in focus. See Chevron v. Nat. Res. Def. Council, 467 U.S. 837, 865–66 (1984). This doctrine governed for forty years, but the Supreme Court recently overruled Chevron. Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244, 2273 (2024).Show More As the Court has decided these modern disputes, a profoundly different—fundamentally executive—conception of administrative action has emerged.31 31.See, e.g., City of Arlington v. FCC, 569 U.S. 290, 304 n.4 (2013) (explaining that administrative actions “take ‘legislative’ and ‘judicial’ forms, but they are exercises of—indeed under our constitutional structure they must be exercises of—the ‘executive Power.’”).Show More At the same time, support has grown for a more unitary theory of executive power that seeks to legitimize agency action through the President’s democratic accountability.32 32.“Presidential Administration intersects with (while being distinct from) . . . the unitary executive theory,” which itself has both stronger and weaker formulations. Elena Chachko, Administrative National Security, 108 Geo. L.J. 1063, 1115 n.331 (2020). The important point for purposes of this Article is that there has been a strong trend in administrative law, which has manifested doctrinally in the Court’s recent separation of powers cases, to embrace presidential control of administration and to look skeptically on legal impediments to such control. See Mark Tushnet, A Political Perspective on the Theory of the Unitary Executive, 12 U. Pa. J. Const. L. 313, 315, 325–29 (2010); Steven G. Calabresi & Saikrishna B. Prakash,The President’s Power to Execute the Laws, 104 Yale L.J. 541, 545 (1994) (“Recently, the weight of academic opinion has shifted . . . to the theory of the unitary Executive.”); Andrea Scoseria Katz & Noah A. Rosenblum, Removal Rehashed, 136 Harv. L. Rev. F. 404, 404 (2023) (“Over the last decade, [the Supreme Court’s] majority has increasingly embraced a unitary theory of Article II . . . .”); cf. Kathryn A. Watts, Proposing a Place for Politics in Arbitrary and Capricious Review, 119 Yale L.J. 2, 8 (2009) (arguing “that what count as ‘valid’ reasons under arbitrary and capricious review should be expanded to include certain political influences from the President, other executive officials, and members of Congress, so long as the political influences are openly and transparently disclosed in the agency’s rulemaking record”).Show More The result is a unitary executive conception of administration that fits most naturally with the type of agency action that spawned it: policymaking through the development and enforcement of general rules adopted pursuant to broad statutory delegations.

The Supreme Court now confronts the challenge of adapting its unitary executive conception of administration to formal adjudicatory hearings, a genuinely quasi-judicial form of agency action that implicates very different issues and values than those at stake in the rulemaking context. To date, adjudication generally has been viewed as an area of administration that is properly insulated from presidential control.33 33.Kagan, supra note 28, at 2306 (“The only mode of administrative action from which Clinton shrank was adjudication. At no time in his tenure did he attempt publicly to exercise the powers that a department head possesses over an agency’s on-the-record determinations.”); see also Wiener v. United States, 357 U.S. 349, 356 (1958) (holding that the adjudicative functions of the War Claims Commission “precluded the President from influencing the Commission in passing on a particular claim” and from removing a member of the Commission “for no reason other than that he preferred to have on that Commission men of his own choosing”); Myers v. United States, 272 U.S. 52, 135 (1926) (“Then there may be duties of a quasi-judicial character imposed on executive officers and members of executive tribunals whose decisions after hearing affect interests of individuals, the discharge of which the President can not in a particular case properly influence or control.”).Show More In her seminal article identifying the phenomenon of “presidential administration,” then-Professor Kagan recognized that adjudication “is fundamentally different” from other forms of agency policymaking such as rulemaking.34 34.Kagan, supra note 28, at 2362; see also Chachko, supra note 32, at 1122 (“Kagan herself did not argue for presidential administration of administrative adjudication.”).Show More In adjudication, “presidential participation in administration, of whatever form, would contravene procedural norms and inject an inappropriate influence into the resolution of controversies.”35 35.Kagan, supra note 28, at 2363; see also Adrian Vermeule, Conventions of Agency Independence, 113 Colum. L. Rev. 1163, 1211 (2013) (“Proponents of expansive presidential power to direct subordinates’ exercise of delegated discretion stop short of arguing for presidential directive power over adjudication, even where strictly executive agencies . . . are concerned.”).Show More Although presidential administration has made some inroads into the adjudication context, these developments have been limited.36 36.See, e.g., Jerry L. Mashaw & David Berke, Presidential Administration in a Regime of Separated Powers: An Analysis of Recent American Experience, 35 Yale J. on Regul. 549, 594–95 (2018) (noting, but “tak[ing] no position on,” the Obama Administration’s push for “more muscular use of adjudication and regulatory enforcement actions across agencies to further policy goals”); Memorandum from Dana Remus, Counsel to the President, to All White House Counsel Staff 1–2, 4–6, 10 (July 21, 2021), https://www.whitehouse.gov/wp-content/uploads/2021/07/White-House-Policy-for-Contacts-with-Agencies-and-Department‌s.pdf [https://perma.cc/T2FA-VMNN] (advising White House staff not to contact agencies about the adjudication of specific cases). Although the Trump Administration issued some directives to agencies regarding the conduct of adjudication, these efforts were general, procedurally focused, and recommendatory. See, e.g., Memorandum from Paul J. Ray, Adm’r, Off. of Info. & Regul. Affs., to the Deputy Sec’ys of Exec. Dep’t & Agencies 1–3 (Aug. 31, 2020), https://www.whitehouse.gov/wp-content/uploads/2020/08/M-20-31.pdf [https://perma‌.cc/9BNA-8WM9].Show More The Supreme Court nonetheless has seemed poised to extend its strong vision of the President’s executive power into the adjudicative space, an outcome that some commentators view as logical and appropriate.37 37.See, e.g., Vermeule, supra note 35, at 1212 (“If one believes that Presidents hold directive power over the delegated discretion of executive agencies, it is unclear why that power would not extend straightforwardly to adjudicative functions of agencies as well as rulemaking functions.”).Show More After all, the Constitution vests the executive power in the President, and “[a]gency adjudication, just as much as agency rulemaking, is an exercise of the ‘executive power’ under Article II.”38 38.Id.Show More

This Article argues that the APA’s ALJ regime is constitutional under the Supreme Court’s new approach because it provides the procedural and institutional structures necessary to ensure faithful execution of the law through administrative adjudication. At the level of legal doctrine, the important point is that the APA’s carefully constructed regime masterfully integrates procedural requirements, employment structures, and agency head control in a way that, taken together, promotes political accountability consistent with the demands of due process. Viewed in its totality, the APA’s regime erects only one—not two—effective levels of for-cause removal protection between the President and the ALJs.39 39.Although Free Enterprise Fund suggests that such a holistic, functional analysis is inappropriate, see 561 U.S. 477, 499–500 (2010), the Court’s more recent decision in United States v. Arthrex, Inc. embraces it, see 141 S. Ct. 1970, 1980–86 (2021).Show More The statute’s robust preservation of agency head control ensures proper presidential control over, and responsibility for, the policymaking aspects of formal adjudication.40 40.If there is an Article II problem to be found here, it is not in the APA’s regime but in the for-cause protection afforded to the principal officers who collectively form the head of agencies such as the SEC. The Court has so far been able to avoid squarely considering this question, but it should do so (in an appropriate case) instead of sacrificing the APA just to kick that can further down the road. See infra Section II.C.Show More Meanwhile, the for-cause removal protections and related employment structures enable the President to ensure impartial adjudication in hearings before the agencies and the MSPB, respectively. At the level of administrative theory, the analysis reveals that executive power in the adjudicatory context is more about discharging duties than exercising discretion. The President must be able to depend on inferior executive officers to fairly adjudicate (as required by due process) and faithfully execute the law (as required by Article II) through an incredible volume of formal adjudicatory hearings. Neither the President nor the heads of departments can review all of these adjudicatory decisions—they must be able to rely on delegation to inferior officers. The APA’s regime ensures that ALJs are sufficiently competent and impartial to meet this need,41 41.For example, ALJs must be lawyers, while non-ALJ adjudicators (often referred to as “administrative judges” or “AJs”) are not always subject to that requirement and often are not lawyers. See Kent Barnett, Against Administrative Judges, 49 U.C. Davis L. Rev. 1643, 1660, 1703 (2016). This may contribute to variable competence across adjudication programs and may also convey the impression that some kinds of agency adjudication are more important than others. Cf. Sara Sternberg Greene & Kristen M. Renberg, Judging Without a J.D., 122 Colum. L. Rev. 1287, 1291 (2022) (studying the use of lay judges in state courts and arguing that “allowing a system of nonlawyer judges perpetuates long-standing inequalities in how litigants experience courts”).Show More while agency heads have proper control over the adjudicatory programs for which they are responsible. From this perspective, it emerges that the restrictive aspects of the APA’s regime empower the President to ensure faithful execution in the unique, quasi-judicial context of formal administrative hearings.42 42.Cf. Robert L. Glicksman & Richard E. Levy, The New Separation of Powers Formalism and Administrative Adjudication, 90 Geo. Wash. L. Rev. 1088, 1096 (2022)(“[P]roperly understood, most administrative adjudication is fully consistent with separation of powers formalism because it involves the execution of the law by officials within the executive branch.”).Show More This insight in turn reveals a path toward reconciling the constitutional tensions between political accountability and impartiality protections in agency adjudication.43 43.See, e.g., Christopher J. Walker, Constitutional Tensions in Agency Adjudication, 104 Iowa L. Rev. 2679, 2680 (2019).Show More

This Article proceeds in three parts. Part I grounds this Article’s analysis in administrative history and reality. It explains the problems Congress sought to remedy by enacting the APA’s ALJ regime, examines that regime in detail, and explains the forces that threaten its continued viability. Part II argues that the APA’s hearing provisions are consistent with the Supreme Court’s recent separation of powers cases. The recent decision in United States v. Arthrex, Inc. is critical, for it establishes that the APA’s robust preservation of agency head control is sufficient to satisfy Article II. This in turn clarifies that if there is a constitutional infirmity in adjudication before independent agencies such as the SEC, it is to be found in for-cause protection for the agency’s principal officers. If the Supreme Court wants to address that issue, it should do so separately and directly. Part III goes deeper, arguing that the Supreme Court’s reconceptualization of administrative action over the last several decades presents deeper threats to administrative adjudication than has previously been recognized. It explores the challenges of embracing an executive theory of administrative adjudication, particularly in a time of presidential primacy. It argues that salvation can be found by embracing the substantial nondiscretionary aspects of formal adjudication and recognizing that proper restrictions on executive action are sometimes necessary to facilitate faithful execution of the law.

  1.  See 5 U.S.C. §§ 556(b)–(c), 557(b). ↑
  2.  561 U.S. 477 (2010). ↑
  3.  Id. at 484. ↑
  4.  Id. ↑
  5.  See 5 U.S.C. § 7521. ↑
  6.  See id. § 1202(d). ↑
  7.  In Free Enterprise Fund, the parties and the majority assumed that SEC commissioners can be removed only for cause, although the SEC’s organic statute contains no for-cause provision. See 561 U.S. at 487; see also id. at 545–46 (Breyer, J., dissenting, joined by Stevens, Ginsburg & Sotomayor, JJ.) (noting that the majority “assume[d] without deciding that . . . SEC Commissioners . . . are removable only ‘for cause’” (emphasis omitted)). SEC commissioners are, however, appointed for a term of years, see 15 U.S.C. § 78d(a), which perhaps should be interpreted as a protection against removal by the President for the duration of the term. See Jane Manners & Lev Menand, The Three Permissions: Presidential Removal and the Statutory Limits of Agency Independence, 121 Colum. L. Rev. 1, 5 (2021). But see Severino v. Biden, 71 F.4th 1038, 1045–46 (D.C. Cir. 2023) (rejecting this approach to interpreting a statutory term of years). ↑
  8.  U.S. Const. art. II, § 2, cl. 2; Lucia v. SEC, 138 S. Ct. 2044, 2049 (2018). Although Lucia dealt only with ALJs employed by the SEC, its holding likely reaches ALJs employed by other agencies because of the similarity of functions. See Guidance on Administrative Law Judges After Lucia v. SEC (S. Ct.), July 2018, 132 Harv. L. Rev. 1120, 1122–23 (2019). ↑
  9.  34 F.4th 446 (5th Cir. 2022), aff’d on other grounds, 144 S. Ct. 2117, 2127, 2139 (2024). ↑
  10.  Id. at 464. ↑
  11.  See Jarkesy, 144 S. Ct. at 2127, 2139; see also Petition for a Writ of Certiorari at I, Jarkesy, 144 S. Ct. 2117 (No. 22-859) (presenting several questions, including “[w]hether Congress violated Article II by granting for-cause removal protection to administrative law judges in agencies whose heads enjoy for-cause removal protection”). ↑
  12.  See Emily S. Bremer, The Rediscovered Stages of Agency Adjudication, 99 Wash. U. L. Rev. 377, 436–47 (2021) [hereinafter Bremer, Rediscovered Stages]. ↑
  13.  295 U.S. 602 (1935). ↑
  14.  Id. at 608. ↑
  15.  See, e.g., Ariz. Grocery Co. v. Atchison, 284 U.S. 370, 389 (1932) (same but involving the Interstate Commerce Commission); Morgan v. United States, 298 U.S. 468, 477, 481–82 (1936) (applying the New Deal conception of “administrative” action to ratemaking conducted by the Department of Agriculture). ↑
  16.  K.C. Davis, Walter Gellhorn & Paul Verkuil, Present at the Creation: Regulatory Reform Before 1946, 38 Admin. L. Rev. 511, 514 (1986). I have examined this research in detail in previous work. See Emily S. Bremer, The Undemocratic Roots of Agency Rulemaking, 108 Cornell L. Rev. 69, 90–93 (2022); Bremer, Rediscovered Stages, supra note 12, at 396–402. The relevant documents are available electronically in The Bremer-Kovacs Collection: Historic Documents Related to the Administrative Procedure Act of 1946, https://heinonline.‌org/HOL/Index?collection=bremer. See generally Emily S. Bremer & Kathryn E. Kovacs, Essay, Introduction to The Bremer-Kovacs Collection: Historic Documents Related to the Administrative Procedure Act of 1946 (HeinOnline 2021), 106 Minn. L. Rev. Headnotes 218 (2022) (offering a narrative introduction to the collection). ↑
  17.  See Off. of the Att’y Gen., Final Report of the Attorney General’s Committee on Administrative Procedure 2–4 (1941) [hereinafter Final Report]. The Final Report is based on 27 monographs examining the procedures and practices of “administrative” agencies. See id. at 3–4. Purely “executive” agencies—such as the Government Printing Office, the Bureau of Standards, the Civil Service Commission (“CSC”), the Bureau of the Budget, and the General Accounting Office—were left out of the study. See id. at 5. ↑
  18.  See 5 U.S.C. § 551(4)–(7). “This particular line may be the APA’s most important innovation.” Kristin E. Hickman & Aaron L. Nielson, Narrowing Chevron’s Domain, 70 Duke L.J. 931, 942 (2021). It was inspired by—but “not on all fours with”—the pre-APA definitions of quasi-judicial and quasi-legislative action. Emily S. Bremer, Blame (or Thank) the Administrative Procedure Act for Florida East Coast Railway, 97 Chi.-Kent L. Rev. 79, 96–97 (2022) [hereinafter Bremer, Blame (or Thank)]. ↑
  19.  The legislature was also influenced by concerns—made concrete by the World Wars and related political developments in Europe—about how to ensure effective administration without facilitating authoritarianism. See, e.g., Kathryn E. Kovacs, Avoiding Authoritarianism in the Administrative Procedure Act, 28 Geo. Mason L. Rev. 573, 596–600 (2021); Noah A. Rosenblum, The Antifascist Roots of Presidential Administration, 122 Colum. L. Rev. 1, 14, 48–49 (2022). ↑
  20.  See Final Report, supra note 17, at 5. ↑
  21.  See Bremer, Rediscovered Stages, supra note 12, at 433. ↑
  22.  See id. at 402–03. ↑
  23.  See Final Report, supra note 17, at 35–38, 41–42. ↑
  24.  Bremer, Rediscovered Stages, supra note 12, at 431. ↑
  25.  For an originalist discussion of the relationship between due process and separation of powers, see Nathan S. Chapman & Michael W. McConnell, Due Process as Separation of Powers, 121 Yale L.J. 1672, 1677–78 (2012). The sovereign power and individual rights aspects of due process are also observable, for example, in personal jurisdiction doctrine. See, e.g., Thomas D. Rowe, Jr., Suzanna Sherry & Jay Tidmarsh, Civil Procedure 452 (5th ed. 2020) (“The Court has wavered about whether personal-jurisdiction doctrine rests on individual liberty or state sovereignty (or both).”). ↑
  26.  The APA’s procedures are a default because they apply unless Congress affirmatively elects to displace them. See 5 U.S.C. §§ 556(b), 559. ↑
  27.  See, e.g., Antonin Scalia, Vermont Yankee: The APA, the D.C. Circuit, and the Supreme Court, 1978 Sup. Ct. Rev. 345, 376 (describing “the constant and accelerating flight away from individualized, adjudicatory proceedings to generalized disposition through rulemaking”); see also Alan B. Morrison, The Administrative Procedure Act: A Living and Responsive Law, 72 Va. L. Rev. 253, 254–55 (1986); Ralph F. Fuchs, Development and Diversification in Administrative Rule Making, 72 Nw. U. L. Rev. 83, 89 (1977) (noting the “growing tendency” of many agencies to confront policy issues via rulemaking rather than adjudication). ↑
  28.  See generally Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245 (2001) (analyzing the President’s recent primacy in setting the direction of administrative process); cf. Michael A. Livermore, Political Parties and Presidential Oversight, 67 Ala. L. Rev. 45, 53–61 (2015) (describing how executive restructuring, presidential control in administrative law, and evolution in the operation of American political parties have all contributed to increase the President’s influence in federal administration). For an early warning that this development might disrupt the carefully mediated tension between law and politics in administrative law, see Peter L. Strauss, Presidential Rulemaking, 72 Chi.-Kent L. Rev. 965, 984–86 (1997). ↑
  29.  See Emily S. Bremer, Power Corrupts, 41 Yale J. on Regul. 426, 456–58 (2024) [hereinafter Bremer, Power Corrupts]. ↑
  30.  The Chevron doctrine—according to which courts would defer to reasonable agency interpretations of the statutes by which Congress has delegated power to them—was both cause and consequence of this change in focus. See Chevron v. Nat. Res. Def. Council, 467 U.S. 837, 865–66 (1984). This doctrine governed for forty years, but the Supreme Court recently overruled Chevron. Loper Bright Enters. v. Raimondo, 144 S. Ct. 2244, 2273 (2024). ↑
  31.  See, e.g., City of Arlington v. FCC, 569 U.S. 290, 304 n.4 (2013) (explaining that administrative actions “take ‘legislative’ and ‘judicial’ forms, but they are exercises of—indeed under our constitutional structure they must be exercises of—the ‘executive Power.’”). ↑
  32.  “Presidential Administration intersects with (while being distinct from) . . . the unitary executive theory,” which itself has both stronger and weaker formulations. Elena Chachko, Administrative National Security, 108 Geo. L.J. 1063, 1115 n.331 (2020). The important point for purposes of this Article is that there has been a strong trend in administrative law, which has manifested doctrinally in the Court’s recent separation of powers cases, to embrace presidential control of administration and to look skeptically on legal impediments to such control. See Mark Tushnet, A Political Perspective on the Theory of the Unitary Executive, 12 U. Pa. J. Const. L. 313, 315, 325–29 (2010); Steven G. Calabresi & Saikrishna B. Prakash, The President’s Power to Execute the Laws, 104 Yale L.J. 541, 545 (1994) (“Recently, the weight of academic opinion has shifted . . . to the theory of the unitary Executive.”); Andrea Scoseria Katz & Noah A. Rosenblum, Removal Rehashed, 136 Harv. L. Rev. F. 404, 404 (2023) (“Over the last decade, [the Supreme Court’s] majority has increasingly embraced a unitary theory of Article II . . . .”); cf. Kathryn A. Watts, Proposing a Place for Politics in Arbitrary and Capricious Review, 119 Yale L.J. 2, 8 (2009) (arguing “that what count as ‘valid’ reasons under arbitrary and capricious review should be expanded to include certain political influences from the President, other executive officials, and members of Congress, so long as the political influences are openly and transparently disclosed in the agency’s rulemaking record”). ↑
  33.  Kagan, supra note 28, at 2306 (“The only mode of administrative action from which Clinton shrank was adjudication. At no time in his tenure did he attempt publicly to exercise the powers that a department head possesses over an agency’s on-the-record determinations.”); see also Wiener v. United States, 357 U.S. 349, 356 (1958) (holding that the adjudicative functions of the War Claims Commission “precluded the President from influencing the Commission in passing on a particular claim” and from removing a member of the Commission “for no reason other than that he preferred to have on that Commission men of his own choosing”); Myers v. United States, 272 U.S. 52, 135 (1926) (“Then there may be duties of a quasi-judicial character imposed on executive officers and members of executive tribunals whose decisions after hearing affect interests of individuals, the discharge of which the President can not in a particular case properly influence or control.”). ↑
  34.  Kagan, supra note 28, at 2362; see also Chachko, supra note 32, at 1122 (“Kagan herself did not argue for presidential administration of administrative adjudication.”). ↑
  35.  Kagan, supra note 28, at 2363; see also Adrian Vermeule, Conventions of Agency Independence, 113 Colum. L. Rev. 1163, 1211 (2013) (“Proponents of expansive presidential power to direct subordinates’ exercise of delegated discretion stop short of arguing for presidential directive power over adjudication, even where strictly executive agencies . . . are concerned.”). ↑
  36.  See, e.g., Jerry L. Mashaw & David Berke, Presidential Administration in a Regime of Separated Powers: An Analysis of Recent American Experience, 35 Yale J. on Regul. 549, 594–95 (2018) (noting, but “tak[ing] no position on,” the Obama Administration’s push for “more muscular use of adjudication and regulatory enforcement actions across agencies to further policy goals”); Memorandum from Dana Remus, Counsel to the President, to All White House Counsel Staff 1–2, 4–6, 10 (July 21, 2021), https://www.whitehouse.gov/wp-content/uploads/2021/07/White-House-Policy-for-Contacts-with-Agencies-and-Department‌s.pdf [https://perma.cc/T2FA-VMNN] (advising White House staff not to contact agencies about the adjudication of specific cases). Although the Trump Administration issued some directives to agencies regarding the conduct of adjudication, these efforts were general, procedurally focused, and recommendatory. See, e.g., Memorandum from Paul J. Ray, Adm’r, Off. of Info. & Regul. Affs., to the Deputy Sec’ys of Exec. Dep’t & Agencies 1–3 (Aug. 31, 2020), https://www.whitehouse.gov/wp-content/uploads/2020/08/M-20-31.pdf [https://perma‌.cc/9BNA-8WM9]. ↑
  37.  See, e.g., Vermeule, supra note 35, at 1212 (“If one believes that Presidents hold directive power over the delegated discretion of executive agencies, it is unclear why that power would not extend straightforwardly to adjudicative functions of agencies as well as rulemaking functions.”). ↑
  38.  Id. ↑
  39.  Although Free Enterprise Fund suggests that such a holistic, functional analysis is inappropriate, see 561 U.S. 477, 499–500 (2010), the Court’s more recent decision in United States v. Arthrex, Inc. embraces it, see 141 S. Ct. 1970, 1980–86 (2021). ↑
  40.  If there is an Article II problem to be found here, it is not in the APA’s regime but in the for-cause protection afforded to the principal officers who collectively form the head of agencies such as the SEC. The Court has so far been able to avoid squarely considering this question, but it should do so (in an appropriate case) instead of sacrificing the APA just to kick that can further down the road. See infra Section II.C. ↑
  41.  For example, ALJs must be lawyers, while non-ALJ adjudicators (often referred to as “administrative judges” or “AJs”) are not always subject to that requirement and often are not lawyers. See Kent Barnett, Against Administrative Judges, 49 U.C. Davis L. Rev. 1643, 1660, 1703 (2016). This may contribute to variable competence across adjudication programs and may also convey the impression that some kinds of agency adjudication are more important than others. Cf. Sara Sternberg Greene & Kristen M. Renberg, Judging Without a J.D., 122 Colum. L. Rev. 1287, 1291 (2022) (studying the use of lay judges in state courts and arguing that “allowing a system of nonlawyer judges perpetuates long-standing inequalities in how litigants experience courts”). ↑
  42.  Cf. Robert L. Glicksman & Richard E. Levy, The New Separation of Powers Formalism and Administrative Adjudication, 90 Geo. Wash. L. Rev. 1088, 1096 (2022) (“[P]roperly understood, most administrative adjudication is fully consistent with separation of powers formalism because it involves the execution of the law by officials within the executive branch.”). ↑

  43.  See, e.g., Christopher J. Walker, Constitutional Tensions in Agency Adjudication, 104 Iowa L. Rev. 2679, 2680 (2019). ↑

Sex Discrimination Formalism

Critics of antidiscrimination law have long lamented that the Supreme Court is devoted to a shallow, formal version of equality that fails to account for substantive inequities and stands in the way of affirmative efforts to remediate systemic injustice. But these criticisms are primarily focused on the Supreme Court’s interpretations of race discrimination law. The Court’s most recent foray into statutory sex discrimination law, Bostock v. Clayton County, employed formalistic reasoning to move the law in an expansive direction, interpreting Title VII’s sex discrimination provision to prohibit discrimination against lesbian, gay, and transgender employees. Examining post-Bostock developments, this Article asks whether formal equality might have more potential to advance civil rights than previously thought. It argues that “formal equality” is not a single legal inquiry; rather, in practice, it takes the form of at least three distinct tests. These tests lead to different results in different sex discrimination controversies, such as whether it is discrimination to treat someone adversely for being bisexual or nonbinary; to single out pregnancy, menstruation, breasts, or other aspects of reproductive biology for disparate treatment; to enforce sex-specific dress codes; to exclude transgender people from restrooms consistent with their gender identities; to ban gender-affirming health care; or to restrict who can change the sex designations on their identity documents. Although no formal test neatly maps onto prevailing normative theories and sociological insights about what discrimination is, in recent cases, courts have used formal tests to achieve results consistent with those theories. This account suggests that, rather than insisting that courts adopt substantive tests, civil rights scholars might reconsider the virtues of formalism.

“Equality, in the abstract, has no limits; it is forever demanding to be carried to its ultimate logical conclusions.”1 1.Kenneth L. Karst, The Supreme Court, 1976 Term—Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 Harv. L. Rev. 1, 39 (1977).Show More

Introduction

The law of race discrimination is mired in what critics call “formal equality”: an ahistorical, decontextualized vision of equality law that ignores the social, economic, and political realities of systemic racial inequality and treats affirmative action as the moral equivalent of 1950s-style segregation.2 2.See, e.g., Ian Haney-López, Intentional Blindness, 87 N.Y.U. L. Rev. 1779, 1784 (2012) (“[D]iscriminatory intent doctrine excludes evidence of continued discrimination against non-Whites rooted in history, contemporary practices, and social science . . . . Meanwhile, . . . colorblindness similarly closes courthouse doors to evidence showing that state actors sometimes use race to break down inequality and to foster integration.”); Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of Status-Enforcing State Action, 49 Stan. L. Rev. 1111, 1113 (1997) (criticizing “[c]ontemporary equal protection law” because it “is premised on a formal and historically static conception of ‘discrimination’” focused on “classification” or “discriminatory purpose—a concept the Court has defined as tantamount to malice”).Show More As a result, antidiscrimination scholars are almost uniformly scornful of formal equality, proposing that it be replaced with more substantive definitions of discrimination attuned to context;3 3.See, e.g., Haney-López, supranote 2, at1876 (proposing a “contextual intent” test).Show More social, historical, and cultural meanings;4 4.See, e.g., Issa Kohler-Hausmann, Eddie Murphy and the Dangers of Counterfactual Causal Thinking About Detecting Racial Discrimination, 113 Nw. U. L. Rev. 1163, 1166, 1172 (2019) (arguing for a definition that accounts for “the system of social meanings or practices” that constitute social categories such as race and sex); Charles R. Lawrence III, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, 39 Stan. L. Rev. 317, 355–56 (1987) (proposing a “cultural meaning” test that “would evaluate governmental conduct to see if it conveys a symbolic message to which the culture attaches racial significance” and “considering evidence regarding the historical and social context in which the decision was made and effectuated”).Show More systemic and accumulated group-based disadvantages;5 5.See, e.g., Richard Thompson Ford, Bias in the Air: Rethinking Employment Discrimination Law, 66 Stan. L. Rev. 1381, 1384 (2014) (“[T]he law should replace the conceptually elusive goal of eliminating discrimination with the more concrete goal of requiring employers, government officials, and other powerful actors to meet a duty of care to avoid unnecessarily perpetuating social segregation or hierarchy.”); Siegel, supranote 2, at1146 (suggesting that equal protection doctrine might require scrutiny for “facially neutral policies” that “perpetuate, or aggravate, historic patterns of race and gender inequality”).Show More or “costs and benefits of alternative proposals in each specific setting.”6 6.See, e.g., R. Richard Banks, Class and Culture: The Indeterminacy of Nondiscrimination, 5 Stan. J. C.R. & C.L. 1, 3 (2009) (“[W]e should approach race-related policy disputes in a pragmatic manner, weighing the costs and benefits of alternative proposals in each specific setting.”).Show More

By contrast to the atrophy of race discrimination law through formalism, the law of sex discrimination seems relatively vibrant. In its landmark decision in Bostock v. Clayton County, the Roberts Court ruled that discrimination on the basis of “sex” under Title VII of the Civil Rights Act includes discrimination against lesbian, gay, and transgender workers.7 7.140 S. Ct. 1731, 1737 (2020).Show More But that decision’s reasoning is not based in any sort of contextual or historically grounded understanding of gender-based subordination.8 8.Id.at 1750–51 (denying the relevance of history and pointing out that “applying protective laws to groups that were politically unpopular at the time of the law’s passage—whether prisoners in the 1990s or homosexual and transgender employees in the 1960s—often may be seen as unexpected”).Show More Rather, it relied on a formal, sterile, individualistic concept of “but-for” causation—“if changing the employee’s sex would have yielded a different choice by the employer—a statutory violation has occurred.”9 9.Id. at 1741.Show More Thus, if an employer would not fire a woman for being attracted to men, that employer may not fire a man for being attracted to men.10 10.Id. The same argument works for the transgender employees—for example, a transgender woman may not be penalized for having traits that would be acceptable in an employee who was assigned female at birth. Id.Show More Lower courts have extended Bostock to new contexts, holding, for example, that it requires that schools allow transgender children to use restrooms consistent with their gender identities,11 11.See, e.g., infraSubsection II.A.1 (discussing Grimm v. Gloucester County School Board, 972 F.3d 586, 616, 619 (4th Cir. 2020) (affirming summary judgment in favor of a transgender plaintiff on equal protection and Title IX claims), cert. denied, 141 S. Ct. 2878 (2021)).Show More forbids an employer from firing an employee because her tampon triggered a security scanner,12 12.SeeinfraSubsection II.A.2 (discussing Flores v. Virginia Department of Corrections, No. 20-cv-00087, 2021 WL 668802, at *6 (W.D. Va. Feb. 22, 2021) (denying summary judgment in a sex discrimination case in which an employee was fired when her tampon set off a security scanner triggering the false suspicion that she was smuggling contraband)).Show More and bars schools from imposing dress codes requiring girls to wear skirts.13 13.See, e.g., infraSubsection II.A.1 (discussing Peltier v. Charter Day School, Inc., 37 F.4th 104 (4th Cir. 2021) (en banc) (affirming grant of summary judgment to plaintiffs on § 1983 equal protection claim and reversing grant of summary judgment to school on Title IX claim challenging discriminatory dress code)).Show More

This Article argues that Bostock’s but-for test is an example of a broader phenomenon that it describes as “sex discrimination formalism”: attempts to define intentional sex discrimination according to formal, abstract, logical tests, minimizing consideration of social realities and normative values.14 14.I define discrimination formalism more precisely infra Section I.A. Cf.Morton J. Horwitz, The Transformation of American Law, 1780–1860, at 254 (1977) (discussing “legal formalism” as “an intellectual system which gave common law rules the appearance of being self-contained, apolitical, and inexorable, and which, by making ‘legal reasoning seem like mathematics,’ conveyed ‘an air . . . of . . . inevitability’ about legal decisions”). I do not suggest formal rules succeed at perfect abstraction or constraint; formalism is a matter of degree. See, e.g., Cass R. Sunstein, Must Formalism Be Defended Empirically?, 66 U. Chi. L. Rev. 636, 640 (1999) (“The real question is ‘what degree of formalism?’ rather than ‘formalist or not?’”).Show More It identifies and examines abstract tests used by courts to determine what types of reasons count as intentional sex discrimination in various constitutional and statutory contexts and assesses how those tests work in particular cases. Contrary to the consensus view among civil rights scholars that formalism is anathema to equality,15 15.See, e.g.,supranotes 2, 4, 6 and accompanying text. But cf.Mary Anne Case,“The Very Stereotype the Law Condemns”: Constitutional Sex Discrimination Law as a Quest for Perfect Proxies, 85 Cornell L. Rev. 1447, 1448–52 (2000) (characterizing equal protection cases on sex as standing for the formalistic rule that, when a law, on its face, treats men and women differently, it may not be based on a generalization that would be untrue for even a single individual man or woman, and arguing that, if courts took this rule seriously, it would lead “in interesting and radical directions” like marriage equality).Show More this Article argues that recent cases relying on formal tests have expanded the reach of sex discrimination law to forms of gender inequality overlooked in the past.

One contribution of this Article is to offer a typology of formal tests of disparate treatment. Much scholarship on discrimination law assumes that there are only two modes for thinking about equality: formal and substantive, and that all formal rules are the same.16 16.See, e.g.,Guy-Uriel E. Charles & Luis Fuentes-Rohwer, Pathological Racism, Chronic Racism & Targeted Universalism, 109 Calif. L. Rev. 1107, 1111 (2021) (discussing the “standard doctrinal account,” which lumps together concerns about formal equality and anticlassification in equal protection law);cf.Aziz Z. Huq, What Is Discriminatory Intent?, 103 Cornell L. Rev. 1211, 1223–24 (2018) (“Questions of how discriminatory intent is defined and proved tend to be ancillary and subordinate to a larger critique of the ideological orientation of the doctrine.”).Show More This Article argues there are at least three distinct types of formal rules when it comes to intentional sex discrimination: (1) but-for causation, which asks whether mistreatment would have befallen an individual if their sex were different; (2) anticlassification rules, also referred to as “blindness,”17 17.This is a problematic metaphor, for, among other reasons, the fact that blind people do see race. See generally Osagie Obasogie, Blinded by Sight: Seeing Race Through the Eyes of the Blind (2013).Show More which ask whether a decision-maker acted pursuant to an explicit or implicit policy that considers sex; and (3) “similarly situated” rules, which forbid decision-makers from treating individuals of different sexes who are alike in all relevant respects differently. Importantly, these heuristics for determining discriminatory intent do not require proof of the specific motives of discriminators.18 18.Cf.Pers. Adm’r v. Feeney, 442 U.S. 256, 279 (1979) (“‘Discriminatory purpose’ . . . implies that the decisionmaker . . . selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group.”).Show More While ostensibly aimed at discerning the same core phenomenon—discriminatory intent—these tests have taken on lives of their own in the case law as independent legal “theories” or “claims.” They most often point to the same result, but in a subset of difficult cases, the choice of formal rule can change the outcome. For example, one district judge, a Republican appointee, concluded that Bostock’s but-for test would not count discrimination on the basis of bisexuality as sex discrimination, but an anticlassification inquiry that requires decisions that are “blind” to sex would.19 19.Bear Creek Bible Church v. EEOC, 571 F. Supp. 3d 571, 622 (N.D. Tex. 2021) (concluding that an employer who discriminates on the basis of bisexuality is not discriminating on the basis of sex under “[t]he traditional but-for ‘favoritism’ analyses,” but is failing to act in a way that is “‘blind’ to sex”), vacated sub nom. Braidwood Mgmt., Inc. v. EEOC, 70 F.4th 914, 940 (5th Cir. 2023). Ideology is unlikely to be the explanation for this twist in reasoning. The district court judge, Reed O’Connor, was appointed by President George W. Bush, and is known for striking down the policies of the Biden and Obama administrations. Tierney Sneed, Judge Notorious for Anti-Obamacare Rulings Has Another Crack, CNN (Jan. 28, 2022, 7:56 AM), https://www.cnn.com/2022/01/28/politics/obamacare-reed-oconnor-biden-doj-health/index.html [https://perma.cc/H3G3-TMDH].Show More

Another contribution of this Article is to offer an assessment of the reach of these various formal tests, relevant to next-generation sex discrimination disputes. While scholars have debated the theoretical potential of Bostock’s but-for inquiry,20 20.CompareKatie Eyer, The But-For Theory of Anti-Discrimination Law, 107 Va. L. Rev. 1621, 1624–25 (2021) (applauding the but-for theory on the ground that it clarifies disparate treatment law and avoids the intent requirement), withRobin Dembroff & Issa Kohler-Hausmann, Supreme Confusion About Causality at the Supreme Court, 25 CUNY L. Rev. 57, 58 (2022) (arguing thatBostock’s but-for test is incoherent and “threaten[s] to limit the reach of antidiscrimination law”), Benjamin Eidelson, Dimensional Disparate Treatment, 95 S. Cal. L. Rev. 785, 794 (2022) (arguing that rather than clarifying disparate treatment law, the but-for theory compounds confusion, is not justified by statutory text, and leads to “untenable results”), and Guha Krishnamurthi, Not the Standard You’re Looking For: But-For Causation in Anti-Discrimination Law, 108 Va. L. Rev. Online 1, 4, 11 (2022) (arguing “that the Court’s simple but-for causation test, by its own lights, does not advance anti-discrimination law”).Show More they have not examined how judges are applying it in new contexts. Nor have they compared the but-for rule to other formal rules on the ground. In just over three years, Bostock has been cited by almost a thousand cases.21 21.According to the Westlaw database, Bostock had been cited by 962 federal and state cases as of October 1, 2023.Show More This Article discusses more than fifty cases decided since Bostock that are related to arguably novel or potentially controversial applications of sex discrimination doctrine.22 22.This Article reviews cases through October 1, 2023.Show More It examines these decisions from the inside out,23 23.While this is a work of legal scholarship, I draw loose inspiration from anthropological methods. Cf.Annelise Riles, The Network Inside Out 6, 16, 19 (2000) (describing an ethnographic method that attempts to gain access to modern knowledge practices from within, beginning by rendering familiar and mundane artifacts visible for analysis, in contexts in which “thick description” is challenging “because the phenomena are dispersed and the cultures are many”); Annelise Riles, A New Agenda for the Cultural Study of Law: Taking on the Technicalities, 53 Buff. L. Rev. 973, 1029–30 (2005) (urging “that the cultural study of legal technology make a methodological commitment not to reduce technology to the politics, culture, history, or personalities surrounding it—that we take the agency of technological form seriously, as a subject on its own terms, as the legal engineers among us do”).Show More endeavoring to see how their reasoning works, to take it seriously, and to hypothesize about where it might go.

This analysis reveals that courts extending sex discrimination law are foregrounding formal rules as the reasons for their decisions, not sociological arguments about the nature of discrimination or feminist or other such normative theories of the harms of discrimination.24 24.Bostock itself is an example. Cary Franklin, Living Textualism, 2020 Sup. Ct. Rev. 119, 143 n.106 (pointing out that Bostock could have been justified based on “antisubordinationist and anti-stereotyping arguments,” but these arguments “necessitate more analytical work than the simple anticlassificationist argument, and conservatives generally reject them”).Show More Formal rules can sometimes circumvent roadblocks to antidiscrimination projects, such as judgments that traits that are unique to men or women cannot be the bases for discrimination,25 25.See infraSubsection II.A.2.Show More that certain groups and individuals are too blameworthy to deserve protection,26 26.See supranote 8 (quoting Bostock v. Clayton County, 140 S. Ct. 1731, 1750 (2020)).Show More or that discriminatory practices are justified by tradition or convention.27 27.See infra Subsection II.A.1.Show More The results are not always what would be expected based on crude measures of judicial ideology.28 28.See, e.g., supranote 19. I note the political affiliations of judges throughout this Article.Show More But a close look at post-Bostock cases reveals that rather than applying formal tests with the rigor of a philosopher, judges apply them with some plasticity, reaching situations that strike them as substantively unfair. Moreover, while courts extending sex discrimination law to new contexts often gesture to Bostock’s but-for inquiry, they are more likely to rely on anticlassification and similarly situated rules. A similarly situated inquiry, which asks whether people are alike in relevant respects, has been particularly prominent in transgender rights litigation.29 29.See infra Subsection II.A.1.Show More

But formalism also has well-known drawbacks. Abstract tests of discrimination suffer from the flaws of all formalistic legal reasoning: they are, to varying degrees, indeterminate, requiring that judges rely on normative and empirical premises to apply them, but deny that they are doing so,30 30.This is a standard criticism of legal formalism. See, e.g., Felix Cohen, Transcendental Nonsense and the Functional Approach, 35 Colum. L. Rev. 809, 820 (1935) (“In every field of law we should find peculiar concepts which are not defined either in terms of empirical fact or in terms of ethics but which are used to answer empirical and ethical questions alike, and thus bar the way to intelligent investigation of social fact and social policy.”).Show More and they are both over- and underinclusive.31 31.See, e.g., Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685, 1701 (1976) (“The more general and the more formally realizable the rule, the greater the equitable pull of extreme cases of over- or underinclusion.”); Frederick Schauer, Formalism, 97 Yale L.J. 509, 510, 535 (1988) (describing formalism as “the concept of decisionmaking according to rule,” and pointing out that “it is exactly a rule’s rigidity, even in the face of applications that would ill serve its purpose, that renders it a rule”).Show More This Article does not make any broad claims about the causal role of formal legal reasoning in judicial decision-making—causation is complex and context specific. It is also not a brief in support of discrimination formalism as a tool of progressive politics—what tools movement lawyers of any political persuasion ought to use will depend on the circumstances. Nor does it argue that sex discrimination formalism achieves rule of law aspirations such as determinacy, predictability, or judicial constraint—particularly not in legal disputes that implicate acute ideological conflicts. Rather, this Article attempts, to the extent possible, to offer a thick description32 32.See supranote 23.Show More of how thin legal rules33 33.Cf.Toni M. Massaro, Gay Rights, Thick and Thin, 49 Stan. L. Rev. 45, 46–47 (1996) (contrasting “thin” doctrinal arguments that appeal to “principles of neutrality” with “thick” arguments that ask judges to “define, or appear to endorse,” particular sexual orientations).Show More operate in a discrete set of cases. It contributes to scholarly criticism of formalism in discrimination law by arguing that, like unhappy families, each of the various formal tests is problematic in its own way. It departs from those criticisms in disputing that a wholesale move toward more substantive inquiries of the sort favored by most progressive scholars would achieve those scholars’ ultimate aspirations for the law. This Article does not endeavor to advance any one single theory of discrimination law, which is a “ramshackle institution, full of compromise and contradiction.”34 34.Robert Post, Prejudicial Appearances: The Logic of American Antidiscrimination Law, 88 Calif. L. Rev. 1, 16 (2000).Show More Rather, it adds to the evidence that a unified theory is not normatively desirable.35 35.See, e.g., Banks, supranote 6, at19 (“The effort to arrive at a unitary conception of discrimination would be misguided even if an authoritative single decision maker—say, the United States Supreme Court—propounded the definition. Any single definition would fail to account for the distinctive features of the various settings where claims of racial discrimination might arise.”); Huq, supranote 16, at 1240 (explaining that discriminatory intent is “unavoidabl[y]” “protean and plural”); George Rutherglen, Disaggregated Discrimination and the Rise of Identity Politics, 26 Wm. & Mary J. Race, Gender & Soc. Just. 391, 394–95 (2020) (arguing that the multiplicity of plausible philosophical theories of the wrong of discrimination and “discrepancies” in legal doctrines “counsel against the quest for uniformity based on the essential nature of discrimination”).Show More

Questions about the meaning of sex discrimination are timely as courts resolve issues involving the scope of LGBTQ+ rights after Bostock and the constitutionality of legal restrictions on abortion after Dobbs.36 36.Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022).Show More Bostock did not address whether its holding would apply to dress codes, restrooms, health care, and many other topics—controversies now being resolved by federal courts.37 37.Bostock v. Clayton County, 140 S. Ct. 1731, 1753 (2020); id.at 1778–83 (Alito, J., dissenting) (noting these among “some of the potential consequences” of Bostock).Show More While transgender litigants racked up an impressive win rate through 2021,38 38.Katie Eyer, Transgender Constitutional Law, 171 U. Pa. L. Rev. 1405, 1405, 1408 (2023) (surveying constitutional transgender rights cases from 2017–2021 and concluding that “recent transgender rights litigation has resulted in important and consistent victories for transgender constitutionalism in the lower and state courts”).Show More results since have been mixed.39 39.There have been significant recent losses. See, e.g., Williams ex rel. L.W. v. Skrmetti, No. 23-5600, 2023 WL 6321688, at *23 (6th Cir. Sept. 28, 2023) (reversing grants of preliminary injunctions against Kentucky and Tennessee laws barring gender-affirming health care for transgender minors); Eknes-Tucker v. Governor of Ala., No. 22-11707, 2023 WL 5344981, at *1 (11th Cir. Aug. 21, 2023) (vacating district court’s preliminary injunction of Alabama law prohibiting gender-affirming health care for transgender minors); Kasper ex rel. Adams v. Sch. Bd., 57 F.4th 791, 799–800 (11th Cir. 2022) (en banc) (reversing district court’s conclusions, following a bench trial, that school policy barring a transgender boy from the boys’ restroom violated the Equal Protection Clause and Title IX); Fowler v. Stitt, No. 22-cv-00115, 2023 WL 4010694, at *24 (N.D. Okla. June 8, 2023) (granting motion to dismiss challenge to state policy prohibiting transgender individuals from changing the sex designations on their birth certificates), appeal docketed, No. 23-5080 (10th Cir. July 7, 2023); Gore v. Lee, No. 19-cv-00328, 2023 WL 4141665, at *37 (M.D. Tenn. June 22, 2023) (similar), appeal docketed, No. 23-5669 (6th Cir. July 26, 2023); B.P.J. v. W. Va. State Bd. of Educ., No. 21-cv-00316, 2023 WL 111875, at *10 (S.D. W. Va. Jan. 5, 2023) (denying transgender litigant’s motion for summary judgment in case challenging law forbidding transgender girls from playing girls’ sports in school), argued, No. 23-1078 (4th Cir. Oct. 27, 2023); A.H. ex rel. D.H. v. Williamson Cnty. Bd. of Educ., 638 F. Supp. 3d 821, 837 (M.D. Tenn. Nov. 2, 2022) (denying preliminary injunction in case challenging Tennessee state law barring transgender schoolchildren from using restrooms consistent with their gender identities).Show More Most notably, a 2022 en banc decision by the Eleventh Circuit rejected a “cornucopia” of formal theories advanced by a transgender student in a case over restroom access.40 40.Adams, 57 F.4th at 846 n.13 (Jill Pryor, J., dissenting) (describing six distinct theories that the majority rejected).Show More While Dobbs addressed equal protection issues, its statements on that question are dicta.41 41.Reva B. Siegel, Serena Mayeri & Melissa Murray, Equal Protection in Dobbs and Beyond: How States Protect Life Inside and Outside of the Abortion Context, 43 Colum. J. Gender & L. 67, 68, 93 (2022) (noting that the parties had not asserted an equal protection claim in Dobbs and observing that “Justice Alito’s attempt to block an equal protection claim that was not even before the Court in Dobbs is evidence of equality’s power, not its weakness”).Show More State courts are now resolving equal protection challenges to abortion bans under their own state constitutions.42 42.SeePlanned Parenthood of the Great Nw. v. State, 522 P.3d 1132, 1198–200 (Idaho 2023) (rejecting equal protection challenges to Idaho law restricting abortion); Siegel et al., supranote 41, at 95–96 (discussing state court decisions on the right to abortion as a matter of gender equality).Show More Yet few scholars are focused “on questions of equal protection and pregnancy.”43 43.Siegel et al., supranote 41, at 73–74 (explaining that “[t]his is because, for decades, the question has been buried under the substantive due process doctrines regulating abortion . . . , and under federal statutes that prohibit pregnancy discrimination, including by government actors”).Show More

This Article proceeds in four Parts. Part I defines discrimination formalism, explains its importance, and offers a typology of formal theories of disparate treatment. Part II argues that courts are relying on formalistic tests to expand sex discrimination law in several contested contexts, including debates over discrimination based on bisexuality, nonbinary gender, menstruation, genitalia, and other aspects of reproductive biology, and sex-segregated restrooms, dress codes, and other such policies. It asks whether various formal tests have potential to further expand sex discrimination law on these issues, and explains the reasons for the appeal of formal over substantive inquiries. Part III probes the limits of sex discrimination formalism and addresses potential criticisms of formal rules. Part IV draws out lessons from this account for debates over formal equality and the future of civil rights law.

  1.  Kenneth L. Karst, The Supreme Court, 1976 Term—Foreword: Equal Citizenship Under the Fourteenth Amendment, 91 Harv. L. Rev. 1, 39 (1977). ↑
  2.  See, e.g., Ian Haney-López, Intentional Blindness, 87 N.Y.U. L. Rev. 1779, 1784 (2012) (“[D]iscriminatory intent doctrine excludes evidence of continued discrimination against non-Whites rooted in history, contemporary practices, and social science . . . . Meanwhile, . . . colorblindness similarly closes courthouse doors to evidence showing that state actors sometimes use race to break down inequality and to foster integration.”); Reva Siegel, Why Equal Protection No Longer Protects: The Evolving Forms of Status-Enforcing State Action, 49 Stan. L. Rev. 1111, 1113 (1997) (criticizing “[c]ontemporary equal protection law” because it “is premised on a formal and historically static conception of ‘discrimination’” focused on “classification” or “discriminatory purpose—a concept the Court has defined as tantamount to malice”). ↑
  3.  See, e.g., Haney-López, supra note 2, at 1876 (proposing a “contextual intent” test). ↑
  4.  See, e.g., Issa Kohler-Hausmann, Eddie Murphy and the Dangers of Counterfactual Causal Thinking About Detecting Racial Discrimination, 113 Nw. U. L. Rev. 1163, 1166, 1172 (2019) (arguing for a definition that accounts for “the system of social meanings or practices” that constitute social categories such as race and sex); Charles R. Lawrence III, The Id, the Ego, and Equal Protection: Reckoning with Unconscious Racism, 39 Stan. L. Rev. 317, 355–56 (1987) (proposing a “cultural meaning” test that “would evaluate governmental conduct to see if it conveys a symbolic message to which the culture attaches racial significance” and “considering evidence regarding the historical and social context in which the decision was made and effectuated”). ↑
  5.  See, e.g., Richard Thompson Ford, Bias in the Air: Rethinking Employment Discrimination Law, 66 Stan. L. Rev. 1381, 1384 (2014) (“[T]he law should replace the conceptually elusive goal of eliminating discrimination with the more concrete goal of requiring employers, government officials, and other powerful actors to meet a duty of care to avoid unnecessarily perpetuating social segregation or hierarchy.”); Siegel, supra note 2, at 1146 (suggesting that equal protection doctrine might require scrutiny for “facially neutral policies” that “perpetuate, or aggravate, historic patterns of race and gender inequality”). ↑
  6.  See, e.g., R. Richard Banks, Class and Culture: The Indeterminacy of Nondiscrimination, 5 Stan. J. C.R. & C.L. 1, 3 (2009) (“[W]e should approach race-related policy disputes in a pragmatic manner, weighing the costs and benefits of alternative proposals in each specific setting.”). ↑
  7.  140 S. Ct. 1731, 1737 (2020). ↑
  8.  Id. at 1750–51 (denying the relevance of history and pointing out that “applying protective laws to groups that were politically unpopular at the time of the law’s passage—whether prisoners in the 1990s or homosexual and transgender employees in the 1960s—often may be seen as unexpected”). ↑
  9.  Id. at 1741. ↑
  10.  Id. The same argument works for the transgender employees—for example, a transgender woman may not be penalized for having traits that would be acceptable in an employee who was assigned female at birth. Id. ↑
  11.  See, e.g., infra Subsection II.A.1 (discussing Grimm v. Gloucester County School Board, 972 F.3d 586, 616, 619 (4th Cir. 2020) (affirming summary judgment in favor of a transgender plaintiff on equal protection and Title IX claims), cert. denied, 141 S. Ct. 2878 (2021)). ↑
  12.  See infra Subsection II.A.2 (discussing Flores v. Virginia Department of Corrections, No. 20-cv-00087, 2021 WL 668802, at *6 (W.D. Va. Feb. 22, 2021) (denying summary judgment in a sex discrimination case in which an employee was fired when her tampon set off a security scanner triggering the false suspicion that she was smuggling contraband)). ↑
  13.  See, e.g., infra Subsection II.A.1 (discussing Peltier v. Charter Day School, Inc., 37 F.4th 104 (4th Cir. 2021) (en banc) (affirming grant of summary judgment to plaintiffs on § 1983 equal protection claim and reversing grant of summary judgment to school on Title IX claim challenging discriminatory dress code)). ↑
  14.  I define discrimination formalism more precisely infra Section I.A. Cf. Morton J. Horwitz, The Transformation of American Law, 1780–1860, at 254 (1977) (discussing “legal formalism” as “an intellectual system which gave common law rules the appearance of being self-contained, apolitical, and inexorable, and which, by making ‘legal reasoning seem like mathematics,’ conveyed ‘an air . . . of . . . inevitability’ about legal decisions”). I do not suggest formal rules succeed at perfect abstraction or constraint; formalism is a matter of degree. See, e.g., Cass R. Sunstein, Must Formalism Be Defended Empirically?, 66 U. Chi. L. Rev. 636, 640 (1999) (“The real question is ‘what degree of formalism?’ rather than ‘formalist or not?’”). ↑
  15.  See, e.g., supra notes 2, 4, 6 and accompanying text. But cf. Mary Anne Case, “The Very Stereotype the Law Condemns”: Constitutional Sex Discrimination Law as a Quest for Perfect Proxies, 85 Cornell L. Rev. 1447, 1448–52 (2000) (characterizing equal protection cases on sex as standing for the formalistic rule that, when a law, on its face, treats men and women differently, it may not be based on a generalization that would be untrue for even a single individual man or woman, and arguing that, if courts took this rule seriously, it would lead “in interesting and radical directions” like marriage equality). ↑
  16.  See, e.g., Guy-Uriel E. Charles & Luis Fuentes-Rohwer, Pathological Racism, Chronic Racism & Targeted Universalism, 109 Calif. L. Rev. 1107, 1111 (2021) (discussing the “standard doctrinal account,” which lumps together concerns about formal equality and anticlassification in equal protection law); cf. Aziz Z. Huq, What Is Discriminatory Intent?, 103 Cornell L. Rev. 1211, 1223–24 (2018) (“Questions of how discriminatory intent is defined and proved tend to be ancillary and subordinate to a larger critique of the ideological orientation of the doctrine.”). ↑
  17.  This is a problematic metaphor, for, among other reasons, the fact that blind people do see race. See generally Osagie Obasogie, Blinded by Sight: Seeing Race Through the Eyes of the Blind (2013). ↑
  18.  Cf. Pers. Adm’r v. Feeney, 442 U.S. 256, 279 (1979) (“‘Discriminatory purpose’ . . . implies that the decisionmaker . . . selected or reaffirmed a particular course of action at least in part ‘because of,’ not merely ‘in spite of,’ its adverse effects upon an identifiable group.”). ↑
  19.  Bear Creek Bible Church v. EEOC, 571 F. Supp. 3d 571, 622 (N.D. Tex. 2021) (concluding that an employer who discriminates on the basis of bisexuality is not discriminating on the basis of sex under “[t]he traditional but-for ‘favoritism’ analyses,” but is failing to act in a way that is “‘blind’ to sex”), vacated sub nom. Braidwood Mgmt., Inc. v. EEOC, 70 F.4th 914, 940 (5th Cir. 2023). Ideology is unlikely to be the explanation for this twist in reasoning. The district court judge, Reed O’Connor, was appointed by President George W. Bush, and is known for striking down the policies of the Biden and Obama administrations. Tierney Sneed, Judge Notorious for Anti-Obamacare Rulings Has Another Crack, CNN (Jan. 28, 2022, 7:56 AM), https://www.cnn.com/2022/01/28/politics/obamacare-reed-oconnor-biden-doj-health/index.html [https://perma.cc/H3G3-TMDH]. ↑
  20.  Compare Katie Eyer, The But-For Theory of Anti-Discrimination Law, 107 Va. L. Rev. 1621, 1624–25 (2021) (applauding the but-for theory on the ground that it clarifies disparate treatment law and avoids the intent requirement), with Robin Dembroff & Issa Kohler-Hausmann, Supreme Confusion About Causality at the Supreme Court, 25 CUNY L. Rev. 57, 58 (2022) (arguing that Bostock’s but-for test is incoherent and “threaten[s] to limit the reach of antidiscrimination law”), Benjamin Eidelson, Dimensional Disparate Treatment, 95 S. Cal. L. Rev. 785, 794 (2022) (arguing that rather than clarifying disparate treatment law, the but-for theory compounds confusion, is not justified by statutory text, and leads to “untenable results”), and Guha Krishnamurthi, Not the Standard You’re Looking For: But-For Causation in Anti-Discrimination Law, 108 Va. L. Rev. Online 1, 4, 11 (2022) (arguing “that the Court’s simple but-for causation test, by its own lights, does not advance anti-discrimination law”). ↑
  21.  According to the Westlaw database, Bostock had been cited by 962 federal and state cases as of October 1, 2023. ↑
  22.  This Article reviews cases through October 1, 2023. ↑
  23.  While this is a work of legal scholarship, I draw loose inspiration from anthropological methods. Cf. Annelise Riles, The Network Inside Out 6, 16, 19 (2000) (describing an ethnographic method that attempts to gain access to modern knowledge practices from within, beginning by rendering familiar and mundane artifacts visible for analysis, in contexts in which “thick description” is challenging “because the phenomena are dispersed and the cultures are many”); Annelise Riles, A New Agenda for the Cultural Study of Law: Taking on the Technicalities, 53 Buff. L. Rev. 973, 1029–30 (2005) (urging “that the cultural study of legal technology make a methodological commitment not to reduce technology to the politics, culture, history, or personalities surrounding it—that we take the agency of technological form seriously, as a subject on its own terms, as the legal engineers among us do”). ↑
  24.  Bostock itself is an example. Cary Franklin, Living Textualism, 2020 Sup. Ct. Rev. 119, 143 n.106 (pointing out that Bostock could have been justified based on “antisubordinationist and anti-stereotyping arguments,” but these arguments “necessitate more analytical work than the simple anticlassificationist argument, and conservatives generally reject them”). ↑
  25.  See infra Subsection II.A.2. ↑
  26.  See supra note 8 (quoting Bostock v. Clayton County, 140 S. Ct. 1731, 1750 (2020)). ↑
  27.  See infra Subsection II.A.1. ↑
  28.  See, e.g., supra note 19. I note the political affiliations of judges throughout this Article. ↑
  29.  See infra Subsection II.A.1. ↑
  30.  This is a standard criticism of legal formalism. See, e.g., Felix Cohen, Transcendental Nonsense and the Functional Approach, 35 Colum. L. Rev. 809, 820 (1935) (“In every field of law we should find peculiar concepts which are not defined either in terms of empirical fact or in terms of ethics but which are used to answer empirical and ethical questions alike, and thus bar the way to intelligent investigation of social fact and social policy.”). ↑
  31.  See, e.g., Duncan Kennedy, Form and Substance in Private Law Adjudication, 89 Harv. L. Rev. 1685, 1701 (1976) (“The more general and the more formally realizable the rule, the greater the equitable pull of extreme cases of over- or underinclusion.”); Frederick Schauer, Formalism, 97 Yale L.J. 509, 510, 535 (1988) (describing formalism as “the concept of decisionmaking according to rule,” and pointing out that “it is exactly a rule’s rigidity, even in the face of applications that would ill serve its purpose, that renders it a rule”). ↑
  32.  See supra note 23. ↑
  33.  Cf. Toni M. Massaro, Gay Rights, Thick and Thin, 49 Stan. L. Rev. 45, 46–47 (1996) (contrasting “thin” doctrinal arguments that appeal to “principles of neutrality” with “thick” arguments that ask judges to “define, or appear to endorse,” particular sexual orientations). ↑
  34.  Robert Post, Prejudicial Appearances: The Logic of American Antidiscrimination Law, 88 Calif. L. Rev. 1, 16 (2000). ↑
  35.  See, e.g., Banks, supra note 6, at 19 (“The effort to arrive at a unitary conception of discrimination would be misguided even if an authoritative single decision maker—say, the United States Supreme Court—propounded the definition. Any single definition would fail to account for the distinctive features of the various settings where claims of racial discrimination might arise.”); Huq, supra note 16, at 1240 (explaining that discriminatory intent is “unavoidabl[y]” “protean and plural”); George Rutherglen, Disaggregated Discrimination and the Rise of Identity Politics, 26 Wm. & Mary J. Race, Gender & Soc. Just. 391, 394–95 (2020) (arguing that the multiplicity of plausible philosophical theories of the wrong of discrimination and “discrepancies” in legal doctrines “counsel against the quest for uniformity based on the essential nature of discrimination”). ↑
  36.  Dobbs v. Jackson Women’s Health Org., 142 S. Ct. 2228 (2022). ↑
  37.  Bostock v. Clayton County, 140 S. Ct. 1731, 1753 (2020); id. at 1778–83 (Alito, J., dissenting) (noting these among “some of the potential consequences” of Bostock). ↑
  38.  Katie Eyer, Transgender Constitutional Law, 171 U. Pa. L. Rev. 1405, 1405, 1408 (2023) (surveying constitutional transgender rights cases from 2017–2021 and concluding that “recent transgender rights litigation has resulted in important and consistent victories for transgender constitutionalism in the lower and state courts”). ↑
  39.  There have been significant recent losses. See, e.g., Williams ex rel. L.W. v. Skrmetti, No. 23-5600, 2023 WL 6321688, at *23 (6th Cir. Sept. 28, 2023) (reversing grants of preliminary injunctions against Kentucky and Tennessee laws barring gender-affirming health care for transgender minors); Eknes-Tucker v. Governor of Ala., No. 22-11707, 2023 WL 5344981, at *1 (11th Cir. Aug. 21, 2023) (vacating district court’s preliminary injunction of Alabama law prohibiting gender-affirming health care for transgender minors); Kasper ex rel. Adams v. Sch. Bd., 57 F.4th 791, 799–800 (11th Cir. 2022) (en banc) (reversing district court’s conclusions, following a bench trial, that school policy barring a transgender boy from the boys’ restroom violated the Equal Protection Clause and Title IX); Fowler v. Stitt, No. 22-cv-00115, 2023 WL 4010694, at *24 (N.D. Okla. June 8, 2023) (granting motion to dismiss challenge to state policy prohibiting transgender individuals from changing the sex designations on their birth certificates), appeal docketed, No. 23-5080 (10th Cir. July 7, 2023); Gore v. Lee, No. 19-cv-00328, 2023 WL 4141665, at *37 (M.D. Tenn. June 22, 2023) (similar), appeal docketed, No. 23-5669 (6th Cir. July 26, 2023); B.P.J. v. W. Va. State Bd. of Educ., No. 21-cv-00316, 2023 WL 111875, at *10 (S.D. W. Va. Jan. 5, 2023) (denying transgender litigant’s motion for summary judgment in case challenging law forbidding transgender girls from playing girls’ sports in school), argued, No. 23-1078 (4th Cir. Oct. 27, 2023); A.H. ex rel. D.H. v. Williamson Cnty. Bd. of Educ., 638 F. Supp. 3d 821, 837 (M.D. Tenn. Nov. 2, 2022) (denying preliminary injunction in case challenging Tennessee state law barring transgender schoolchildren from using restrooms consistent with their gender identities). ↑
  40.  Adams, 57 F.4th at 846 n.13 (Jill Pryor, J., dissenting) (describing six distinct theories that the majority rejected). ↑
  41.  Reva B. Siegel, Serena Mayeri & Melissa Murray, Equal Protection in Dobbs and Beyond: How States Protect Life Inside and Outside of the Abortion Context, 43 Colum. J. Gender & L. 67, 68, 93 (2022) (noting that the parties had not asserted an equal protection claim in Dobbs and observing that “Justice Alito’s attempt to block an equal protection claim that was not even before the Court in Dobbs is evidence of equality’s power, not its weakness”). ↑
  42.  See Planned Parenthood of the Great Nw. v. State, 522 P.3d 1132, 1198–200 (Idaho 2023) (rejecting equal protection challenges to Idaho law restricting abortion); Siegel et al., supra note 41, at 95–96 (discussing state court decisions on the right to abortion as a matter of gender equality). ↑

  43.  Siegel et al., supra note 41, at 73–74 (explaining that “[t]his is because, for decades, the question has been buried under the substantive due process doctrines regulating abortion . . . , and under federal statutes that prohibit pregnancy discrimination, including by government actors”). ↑

Multi-Textual Constitutions

We have long been taught that constitutions are either “written” or “unwritten.” But this binary classification is wrong. All constitutions are in some way written, and all constitutions contain unwritten rules. This false distinction moreover overlooks the most important formal difference among the constitutions of the world: some constitutions consist of a single, supreme document of higher law while others consist of multiple documents, each enacted separately with shared supremacy under law. Ubiquitous but so far unnoticed, these constitutions comprising multiple texts are a unique constitutional form that has yet to be studied and theorized. I call them multi-textual constitutions.

This Article is the first on multi-textuality as a constitutional form. I draw from current and historical constitutions in Africa, the Americas, Asia, Europe, and Oceania to explain, illustrate, and theorize the design and operation of multi-textual constitutions. I examine their origins, compare how they perform relative to the alternative uni-textual constitutional form, and outline a research agenda for further study. What results is a reordering of our basic constitutional categories, a deep analytical dive into a distinct constitutional form, and a disruptive revelation about the United States Constitution, the world’s paradigmatic model of a uni-textual constitution.

Introduction: Beyond Written and Unwritten Constitutions

For generations, the study of constitutional law has begun with a standard distinction: some constitutions are “written” while others are “unwritten.”1 1.See Michael Foley, The Silence of Constitutions: Gaps, ‘Abeyances’ and Political Temperament in the Maintenance of Government 3 (1989) (“One of the most traditional points of departure in the study of constitutions has been to classify them according to whether they are ‘written’ or ‘unwritten.’”); Andrew Heywood, Politics 293 (2d ed. 2002) (“Traditionally, considerable emphasis has been placed on the distinction between written and unwritten constitutions.”); Herbert W. Horwill, The Usages of the American Constitution 1 (1925) (“Once upon a time some unknown humorist divided constitutions into written and unwritten, and since then text-book after text-book has taken his classification seriously. The American Constitution, we are told, is an example of the former class and the English of the latter.”); see also A.V. Dicey, Introduction to the Study of the Law of the Constitution 3–6 (3d ed. 1889) (distinguishing written and unwritten constitutions on several grounds, including how to locate them and how to identify their constitutive rules); Paul Craig, Written and Unwritten Constitutions: The Modality of Change, in Pragmatism, Principle, and Power inCommon Law Constitutional Systems 263, 263 (Sam Bookman, Edward Willis, Hanna Wilberg & Max Harris eds., 2022) (describing written constitutions as “the norm” and unwritten constitutions as “the rare exception”); James Allan, Against Written Constitutionalism, 14 Otago L. Rev. 191, 191–93 (2015) (observing that “[m]ost of the democratic world has some sort of written constitution” while at most three democracies have an “unwritten constitution,” namely Israel, New Zealand, and the United Kingdom).Show More According to this traditional contrast, written constitutions exist on parchment in documentary form, while unwritten constitutions are intangible sets of invisible rules consisting of norms, principles, and practices that sustain the constitutional order without entrenchment in written word.2 2.See, e.g., W.J. Cocker, The Government of the United States 55 (1889) (“Constitutions are either written or unwritten. A written constitution is a body of laws, contained in a written document, under which the government is conducted. The Constitution of the United States is an example. . . . An unwritten constitution is one having no definite form. The English constitution is an example.”); Lucius Hudson Holt, The Elementary Principles of Modern Government 26 (1923) (“A constitution may be written or unwritten. It may be a single document, like the constitution of the United States, or it may be a combination of legal precedent, individual bills and grants, and immemorial customs, like the constitution of England.”); John Alexander Jameson, A Treatise on Constitutional Conventions; Their History, Powers, and Modes of Proceeding 77 (4th ed. 1887) (“An unwritten Constitution is made up largely of customs and judicial decisions, the former more or less evanescent and intangible . . . . Not so with written Constitutions.”); Emlin McClain, Constitutional Law in the United States 11 (1905) (“If the body of rules and principles is not reduced to definite and authoritatively written form, the constitution is said to be unwritten, as in the familiar case of Great Britain.”).Show More This foundational distinction has been the basic building block in constitutional studies. But it is both incorrect and misleading.

The distinction between written and unwritten constitutions is incorrect because all constitutions are in some way written. Even parts of the paradigmatically “unwritten” Constitution of the United Kingdom are written somewhere, namely in statutes that are endowed with constitutional status,3 3.SeeThoburn v. Sunderland City Council [2002] EWHC (Admin) 195 [62], [2003] QB 151 (Eng.) (enumerating statutes that have constitutional status).Show More for instance the Magna Carta,4 4.Magna Carta 1297, 25 Edw. 1 c. 9 (Eng.).Show More the Bill of Rights,5 5.Bill of Rights 1688, 1 W. & M. c. 2 (Eng.).Show More and the Human Rights Act.6 6.Human Rights Act 1998, c. 42 (UK).Show More It is more accurate to describe an “unwritten” constitution as partly codified and partly uncodified, since many of its constitutional norms appear in official texts.

The familiar distinction between written and unwritten constitutions is moreover misleading because all constitutions contain unwritten rules. No constitution is ever fully written, and one might well wonder whether it is possible for a constitution to be set out entirely in documentary form.7 7.See John Gardner, Can There Be a Written Constitution?, in 1 Oxford Studies in Philosophy of Law 162, 188–92 (Leslie Green & Brian Leiter eds., 2011).Show More Even the United States Constitution—the archetypical “written” constitution—consists of “a constitution outside the constitution,”8 8.See Ernest A. Young, The Constitution Outside the Constitution, 117 Yale L.J. 408, 410–14 (2007).Show More a common reference to the extra-canonical norms, practices, relationships, and institutions that form part of the constitution beyond its text. Scholars have properly recognized that the U.S. Constitution is comprised of various “invisible”9 9.Laurence H. Tribe, The Invisible Constitution 25–27 (2008). Similar themes appear in relation to works on the “unwritten” Constitution of the United States. SeeAkhil Reed Amar, America’s Unwritten Constitution: The Precedents and Principles We Live By, at ix–x (2012); Don K. Price, America’s Unwritten Constitution: Science, Religion, and Political Responsibility 9 (1983).Show More elements, and they have even inquired whether and how it might be possible to amend America’s unwritten constitution.10 10.See Richard Albert, Ryan C. Williams & Yaniv Roznai, Introduction: A Return to Constitutional Basics: Amendment, Constitution, and Writtenness, in Amending America’s Unwritten Constitution 1, 14–16 (Richard Albert, Ryan C. Williams & Yaniv Roznai eds., 2022).Show More

This false distinction between written and unwritten constitutions comes at a great cost. It overlooks and obscures the most important formal distinction among the constitutions of the world: some constitutions consist of a single, supreme document of higher law while others consist of multiple documents, each enacted separately with shared supremacy under law. In jurisdictions governed by more than one document of higher law, the constitution is composed of more than one self-standing text of equal legal force, and together those texts are regarded jointly as the supreme law of the land.11 11.In this Article, I focus only on multi-textual national constitutions, but multi-textual constitutions exist also at the subnational and supranational levels.Show More The documents comprising these constitutions are enacted separately in a variety of forms, for instance, as founding constitutional texts, organic laws endowed with constitution-level status, and constitutional amendments promulgated as separate documents.

Ubiquitous yet so far unidentified, this constitutional form defies our conventional understanding of “written” constitutions. Rather than one official text, there are many, and no single text prevails over another because all are considered equal. These constitutions are therefore unlike uni-textual constitutions whose written elements appear in a single document that is treated as the only supreme law of the land. I call them multi-textual constitutions. Multi-textual constitutions differ from single-text constitutions on the major markers of constitutional life: their initial design, their ongoing evolution, their authoritative interpretation, and their formal amendment. Multi-textual constitutions moreover raise intriguing possibilities for governance in relation to democracy, human rights, and the rule of law that set them apart from what is regarded as the world’s dominant model of uni-textual constitutions.12 12.SeeDenis J. Galligan & Mila Versteeg, Theoretical Perspectives on the Social and Political Foundations of Constitutions, in Social and Political Foundations of Constitutions 3, 6 (Denis J. Galligan & Mila Versteeg eds., 2013) (observing that “the standard practice across the nations of the world, with just a few exceptions, is to have a single written constitutional document”).Show More

There are advantages to introducing “multi-textual constitutions” as a term and category in the field of public law generally and constitutional studies specifically. Using “multi-textual” as a new term to identify this unique classification of constitutions brings a much-needed correction to the mistaken identification of constitutions as “unwritten.” In addition, using “multi-textual” as a new category for constitutions distinguishes them in both form and function from the alternative uni-textual model.

Scholars have yet to identify, explain, and theorize multi-textuality as a distinctive constitutional form despite its prevalence in every region of the world, across all legal traditions, and in all types of constitutional states no matter their age. My purpose in this Article is to introduce, illustrate, and theorize multi-textuality with reference to current and historical constitutions, and to show how this ubiquitous constitutional form disrupts much of what we know about constitutions, including the U.S. Constitution. The great revelation is that the U.S. Constitution consists of multiple documents of higher law, each equally supreme in the constitutional order. Yet, as I will show, although the U.S. Constitution satisfies the trio of legal criteria to be defined in form and operation as multi-textual, it fails the sociological test of public recognition as multi-textual because it is perceived in law and society as uni-textual.

I begin, in Part I, by showing the remarkable omnipresence of multi-textual constitutions in the world. I draw from many constitutional traditions to show the prevalence of multi-textuality in countries rooted in civil and common law traditions, with parliamentary and presidential systems, and in the Global North and South. I furthermore show that multi-textual constitutions are created in one of two ways: either by express design or by unplanned evolution. In Part II, I identify problems created by multi-textual constitutions in connection with three basic questions that are not ordinarily asked of uni-textual constitutions: (1) what is the constitution?, (2) where is the constitution?, and (3) when does a set of legal rules becomes constitutional? Part III then turns to the potential promise of multi-textuality. I highlight three areas of strength for multi-textual constitutions: (1) they make possible incremental constitutional development as a constitutional state begins the transition from one regime to another; (2) they make available multiple options for constitutional reform and may therefore offer more flexibility in managing changes to higher law; and (3) they may help forestall the rise of a popular obsession with the constitution, what scholars have diagnosed as “veneration,” a problematic phenomenon traceable to James Madison,13 13.See The Federalist No. 49, at 340 (James Madison) (Jacob E. Cooke ed., 1961).Show More one of the authors of the U.S. Constitution. I close with a research agenda for future study to enhance our understanding of both uni-textual and multi-textual constitutions.

  1.  See Michael Foley, The Silence of Constitutions: Gaps, ‘Abeyances’ and Political Temperament in the Maintenance of Government 3 (1989) (“One of the most traditional points of departure in the study of constitutions has been to classify them according to whether they are ‘written’ or ‘unwritten.’”); Andrew Heywood, Politics 293 (2d ed. 2002) (“Traditionally, considerable emphasis has been placed on the distinction between written and unwritten constitutions.”); Herbert W. Horwill, The Usages of the American Constitution 1 (1925) (“Once upon a time some unknown humorist divided constitutions into written and unwritten, and since then text-book after text-book has taken his classification seriously. The American Constitution, we are told, is an example of the former class and the English of the latter.”); see also A.V. Dicey, Introduction to the Study of the Law of the Constitution 3–6 (3d ed. 1889) (distinguishing written and unwritten constitutions on several grounds, including how to locate them and how to identify their constitutive rules); Paul Craig, Written and Unwritten Constitutions: The Modality of Change, in Pragmatism, Principle, and Power in Common Law Constitutional Systems 263, 263 (Sam Bookman, Edward Willis, Hanna Wilberg & Max Harris eds., 2022) (describing written constitutions as “the norm” and unwritten constitutions as “the rare exception”); James Allan, Against Written Constitutionalism, 14 Otago L. Rev. 191, 191–93 (2015) (observing that “[m]ost of the democratic world has some sort of written constitution” while at most three democracies have an “unwritten constitution,” namely Israel, New Zealand, and the United Kingdom). ↑
  2.  See, e.g., W.J. Cocker, The Government of the United States 55 (1889) (“Constitutions are either written or unwritten. A written constitution is a body of laws, contained in a written document, under which the government is conducted. The Constitution of the United States is an example. . . . An unwritten constitution is one having no definite form. The English constitution is an example.”); Lucius Hudson Holt, The Elementary Principles of Modern Government 26 (1923) (“A constitution may be written or unwritten. It may be a single document, like the constitution of the United States, or it may be a combination of legal precedent, individual bills and grants, and immemorial customs, like the constitution of England.”); John Alexander Jameson, A Treatise on Constitutional Conventions; Their History, Powers, and Modes of Proceeding 77 (4th ed. 1887) (“An unwritten Constitution is made up largely of customs and judicial decisions, the former more or less evanescent and intangible . . . . Not so with written Constitutions.”); Emlin McClain, Constitutional Law in the United States 11 (1905) (“If the body of rules and principles is not reduced to definite and authoritatively written form, the constitution is said to be unwritten, as in the familiar case of Great Britain.”). ↑
  3.  See Thoburn v. Sunderland City Council [2002] EWHC (Admin) 195 [62], [2003] QB 151 (Eng.) (enumerating statutes that have constitutional status). ↑
  4.  Magna Carta 1297, 25 Edw. 1 c. 9 (Eng.). ↑
  5.  Bill of Rights 1688, 1 W. & M. c. 2 (Eng.). ↑
  6.  Human Rights Act 1998, c. 42 (UK). ↑
  7.  See John Gardner, Can There Be a Written Constitution?, in 1 Oxford Studies in Philosophy of Law 162, 188–92 (Leslie Green & Brian Leiter eds., 2011). ↑
  8.  See Ernest A. Young, The Constitution Outside the Constitution, 117 Yale L.J. 408, 410–14 (2007). ↑
  9.  Laurence H. Tribe, The Invisible Constitution 25–27 (2008). Similar themes appear in relation to works on the “unwritten” Constitution of the United States. See Akhil Reed Amar, America’s Unwritten Constitution: The Precedents and Principles We Live By, at ix–x (2012); Don K. Price, America’s Unwritten Constitution: Science, Religion, and Political Responsibility 9 (1983). ↑
  10.  See Richard Albert, Ryan C. Williams & Yaniv Roznai, Introduction: A Return to Constitutional Basics: Amendment, Constitution, and Writtenness, in Amending America’s Unwritten Constitution 1, 14–16 (Richard Albert, Ryan C. Williams & Yaniv Roznai eds., 2022). ↑
  11.  In this Article, I focus only on multi-textual national constitutions, but multi-textual constitutions exist also at the subnational and supranational levels. ↑
  12.  See Denis J. Galligan & Mila Versteeg, Theoretical Perspectives on the Social and Political Foundations of Constitutions, in Social and Political Foundations of Constitutions
    3, 6 (

    Denis J. Galligan & Mila Versteeg eds.,

    2013) (

    observing that “the standard practice across the nations of the world, with just a few exceptions, is to have a single written constitutional document”). ↑

  13.  See The Federalist No. 49, at 340 (James Madison) (Jacob E. Cooke ed., 1961). ↑

©Virginia Law Review Association. All rights reserved.