Separation of Structures

*Assistant Professor of Law, Emory University School of Law. J.D., Yale Law School; Ph.D., Yale University. I thank Tony Kronman and Nick Parrillo for their encouragement and wise guidance during the initial stages of this project, Will Baude, Erin Delaney, Bill Eskridge, Daniel Farber, Abbe Gluck, James Russell Gordley, Paul Gowder, Andrew Hammond, Christine Jolls, Heidi Kitrosser, Gary Lawson, Lawrence Liu, Sandra Mayson, Christopher Peterson, James Pfander, Robert Post, Daniel Rodriguez, Nadav Shoked, Jed Shugerman, Nick Vincent, John Witt, and the audience at the Power in the Administrate State Workshop for their helpful feedback and criticism.Show More

In a series of decisions—Free Enterprise Fund v. Public Company Accounting Oversight Board, Seila Law v. Consumer Financial Protection Bureau, and Collins v. Yellen—the Supreme Court struck down for-cause removal restrictions over agency heads. These rulings fault structural elements of the respective agency—double-layer protections or single directorships—for violating separation of powers because they insulate the agency from presidential review and oversight. But while the Court increasingly relies on agency structures to adjudicate constitutionality, separation of powers scholarship has focused on the division of powers into legislative, executive, and judicial functions.

This Article supplies the missing account of separation of structures, and in the process defends the legitimacy of the administrative state against its critics. It argues that an emphasis on an agency’s institutional structure in adjudicating constitutionality is deeply rooted in constitutional design and the Founders’ reception of ancient Greek and Roman political philosophy. By introducing the link between institutional design and the Constitution, separation of structures sketches a doctrinal terrain of how judicial adjudications of agency structure could proceed beyond the formalist approach latent in the Court’s recent decisions. By shifting the doctrinal focus from the nature of political functions to the design of accountability mechanisms in governance structures, this Article provides strong support for the constitutionality of congressional delegation of legislative powers to agencies. This more capacious understanding of structural separation of powers accords with constitutional design and better accommodates the dynamic needs of modern regulation.

Introduction

Modern separation of powers doctrine is in disarray. While the Supreme Court routinely decides questions of interbranch conflict, agency structure, and delegation,¹ 1.E.g., Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 582 (1952); Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020); Stern v. Marshall, 564 U.S. 462, 469 (2011); Gundy v. United States, 139 S. Ct. 2116, 2129 (2019).Show More both its approaches and the cases’ outcomes feature sharp disagreement and immense unpredictability. Much of the contemporary jurisprudence on the President’s power to remove agency officials, for example, derives from two contrasting precedents. In Myers v. United States, the Court held that the Decision of 1789 gave the President constitutional entitlement to remove executive branch officials for any reason.² 2.272 U.S. 52, 119 (1926).Show More A mere nine years later, in Humphrey’s Executor v. United States, the Court empowered Congress to specify for-cause removal conditions for independent agencies with quasi-legislative and quasi-judicial functions.³ 3.Humphrey’s Ex’r v. United States, 295 U.S. 602, 629 (1935); see also Aditya Bamzai, Taft, Frankfurter, and the First Presidential For-Cause Removal, 52 U. Rich. L. Rev. 691, 699–701 (2018) (recounting Justice Sutherland’s attempts to distinguish Humphrey’s Executor from Myers based on the term of years established in the FTC’s organic statute and the quasi-legislative, quasi-judicial character of the FTC).Show More Today’s debate tracks this disagreement: after Morrison v. Olson articulated an open-textured inquiry of whether removal restrictions impede the President’s ability to execute his Take Care duties, the Court reversed course by adopting, in Free Enterprise Fund v. Public Company Accounting Oversight Board (“PCAOB”), a bright-line rule that dual-layered for-cause restrictions are unconstitutional.⁴ 4.Morrison v. Olson, 487 U.S. 654, 691–92 (1988); Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); see Jack Goldsmith & John F. Manning, The Protean Take Care Clause, 164 U. Pa. L. Rev. 1835, 1841–43 (2016).Show More Such incongruity extends to other spheres of doctrinal engagement. With respect to congressional grants of adjudicative authority to non-Article III tribunals, the Court has applied a pragmatic test and concluded that an agency’s jurisdiction over common law counterclaims is constitutional; the Court has also taken a more formalist approach and held that such jurisdiction contravenes separation of powers.⁵ 5.Commodity Futures Trading Comm’n v. Schor, 478 U.S. 833, 851–52 (1986); Stern, 564 U.S. at 482–83; see William Baude, Adjudication Outside Article III, 133 Harv. L. Rev. 1511, 1519–21 (2020) (reconciling non-Article-III adjudication with functional separation of powers); infra Figure 2.Show More When determining which officials are “inferior Officers” for the purposes of the Appointments Clause, the Court has characterized an independent counsel—not subordinate to any executive branch officers—as an inferior officer, while defining, a decade later, inferior officers as those supervised by principal officers.⁶ 6.Morrison, 487 U.S. at 671; Edmond v. United States, 520 U.S. 651, 663 (1997); see United States v. Arthrex, Inc., 141 S. Ct. 1970, 1985–86 (2021).Show More The latest victims of this doctrinal quagmire are the Consumer Financial Protection Bureau (“CFPB”) and the Federal Housing Finance Agency (“FHFA”): the Court—splintered along ideological lines—invalidated for-cause removal restrictions on those agencies’ directors.⁷ 7.Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020); Collins v. Yellen, 141 S. Ct. 1761, 1787 (2021).Show More

Underpinning these doctrinal puzzles are patterns that only muddy the waters. The Court has announced, with some consistency, the purpose of its separation of powers doctrine: to erect “structural protections against abuse of power [that are] critical to preserving liberty.”⁸ 8.Seila Law, 140 S. Ct. at 2202 (quoting Bowsher v. Synar, 478 U.S. 714, 730 (1986)).Show More But precisely how (or why) policing the confines of government bodies’ distinct powers contributes to individual freedom is unclear,⁹ 9.Daryl J. Levinson, The Supreme Court 2015 Term—Foreword: Looking for Power in Public Law, 130 Harv L. Rev. 31, 37 (2016).Show More and the mechanisms of effectuating that goal are unpredictable. The Court has considered a combination of three factors: function, power, and design. It has asked whether the function, or the type of authority, exercised by the government body is of the kind constitutionally assigned to it by its Vesting Clause: for example, whether the Commodities Futures Trading Commission’s jurisdiction over common law claims represents an exercise of the judicial function.¹⁰ 10.Schor, 478 U.S. at 851.Show More It has asked whether the magnitude of one actor’s authority impedes the ability of another to fulfill its constitutional responsibilities: for example, whether the CFPB has “potent enforcement powers” and “extensive adjudicatory authority.”¹¹ 11.Seila Law, 140 S. Ct. at 2193.Show More It has also considered issues of institutional design: for example, whether congressionally mandated for-cause removal conditions create a double layer of protection for executive personnel.¹² 12.Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); see also Lisa Shultz Bressman, What Seila Law Says About Chief Justice Roberts’ View of the Administrative State, U. Chi. L. Rev. Online (Aug. 27, 2020), https://lawreviewblog.uchicago.edu/2020/08/27/seila-bress‌man [https://perma.cc/97HT-CNRH] (arguing that Seila Law changes existing jurisprudence by “let[ting] the structure of the agency determine the degree of presidential control over its principal officers”).Show More But precisely which factor the Court will emphasize (and the interaction among them) remains a puzzle. In particular, it is unclear whether considerations of design constitute an independent analysis or are merely parasitic upon issues of power and function. For these reasons, scholars have characterized separation of powers doctrine as a “hoary non sequitur”¹³ 13.Stephen L. Carter, The Independent Counsel Mess, 102 Harv. L. Rev. 105, 105 (1988).Show More and criticized it for its “[l]ack of progress.”¹⁴ 14.M. Elizabeth Magill, The Real Separation in Separation of Powers Law, 86 Va. L. Rev. 1127, 1129 (2000); see also Jerry L. Mashaw, Of Angels, Pins, and For-Cause Removal: A Requiem for the Passive Virtues, U. Chi. L. Rev. Online (Aug. 27, 2020), https://lawreview‌blog.uchicago.edu/2020/08/27/seila-mashaw [https://perma.cc/BMT9-NK6C] (characterizing the doctrine on for-cause removal as a “jurisprudential train wreck”).Show More

Academic commentary has not successfully explained the doctrinal variation.¹⁵ 15.See infra Part I.Show More Scholars have developed complex models to ground the Court’s separation of powers jurisprudence. But those models only underscore disagreement over the fundamental building blocks of their theories. Relying on the Vesting Clauses, scholars have argued that the three constitutional branches of government are each assigned distinct functions.¹⁶ 16.See infra notes 32–36 and accompanying text.Show More These separation models, however, suffer from inconsistency with contemporary practice, not the least from the rise of the powerful administrative state.¹⁷ 17.See infra notes 40–48 and accompanying text. Strict separation-model theorists question the legitimacy of the modern administrative state, prominently by appeals to the nondelegation doctrine. E.g., Gary Lawson, Delegation and Original Meaning, 88 Va. L. Rev. 327, 332 (2002) (arguing that the “civics-book model of legislators legislating, executives executing, and judges judging has enormous intuitive—and legitimating—power,” which explains the modern obsession with the nondelegation doctrine despite its disuse); Ilan Wurman, Nondelegation at the Founding, 130 Yale L.J. 1490, 1493–94 (2021) (arguing there is significant evidence the Founding generation adhered to a robust nondelegation doctrine that was keenly faithful to traditional separation of powers); see also infra Subsection IV.B.3.Show More Other scholars have committed to a more fluid balance among the branches and proposed judicial intervention as a means to restore accountability and good governance.¹⁸ 18.See infra notes 49–55 and accompanying text.Show More But these balance models offer little doctrinal determinacy and threaten nonjusticiability.¹⁹ 19.See infra notes 56–57 and accompanying text.Show More Most attempts to combine the two main approaches are limited to specialized arenas and have not generated consensus.²⁰ 20.See infra Section I.C.Show More The most recent scholarly strands have suggested exogenous approaches that abandon existing doctrinal molds altogether.²¹ 21.See infra notes 73–78.Show More

This Article argues that, in contemporary discourse about separation of powers, an important piece of the puzzle is missing. The Article articulates a theory of separation of structures, which in its simplest version posits that political authority should depend not only on the power being exercised but also on the institutional structure of the government entity that exercises the power. Previous theories—separation and balance models alike—have focused exclusively on the nature or the magnitude of the contested functions: for example, whether an agency in the executive branch has performed actions that are adjudicative in nature (and therefore encroached on the judiciary), or whether Congress has assigned to itself so extensive an authority as to disrupt the distribution of powers among the constitutional branches.²² 22.See infra Sections I.A–B.Show More But an account of separation of powers is incomplete without considering the structural design of the entity performing the contested functions: for example, whether an agency’s unitary structure concentrates power and heightens the need for accountability, or whether a multimember body facilitates deliberation and expertise necessary for technical decision-making. The case law of the past decade has unmistakably established the relevance of institutional design.²³ 23.E.g., Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2191–92 (2020); Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); Collins v. Yellen, 141 S. Ct. 1761, 1770–71 (2021).Show More This Article supplies this missing account of institutional structure in separation of powers.

Importantly, separation of structures originated in ancient Greek and Roman political theory, indelibly shaped the Founding generation’s understanding, and formed an integral part of the constitutional design. Separation of powers—the structural and the functional strands—finds its genesis in Aristotle’s typology of regimes, which divides constitutions into six types based on the numerosity of the governing class and constitution’s compliance with (or deviation from) the normative ends of government.²⁴ 24.See infra Section II.A.Show More Polybius, a second-century Greek historian, transforms this typology into a theory of mixed government.²⁵ 25.See infra Section II.B.Show More None of the basic Aristotelian constitutional forms (monarchy, aristocracy, and democracy), individually considered, instantiates desirable political design. The perfect constitution incorporates each regime type. Separation of structures remained highly influential in the early-modern period: Montesquieu subscribed to a version of the model,²⁶ 26.See infra Section II.D.Show More and the English political theorists adapted it to justify the constitutional setup of England.²⁷ 27.See infra Section II.C.Show More The Founding generation, well-versed in classical philosophy and ancient history, saw separation of structures and mixed government as background assumptions of any successful constitutional design.²⁸ 28.See infra Part III.Show More Although the Founders ultimately abandoned the British (what I call the sociological) notion of mixed government, the structural provisions of the Constitution, with its institution of representation, evinced a return to Aristotelian separation of structures. The absence of separation of structures in contemporary discussion accounts in part for the doctrinal disarray and the scholarly disagreement.²⁹ 29.Most scholars give only cursory treatment to the Aristotelian origins of separation of powers. See, e.g., Steven G. Calabresi, Mark E. Berghausen & Skylar Albertson, The Rise and Fall of the Separation of Powers, 106 Nw. U. L. Rev. 527, 529–36 (2012). Serious assessments of classical political philosophy’s contribution to modern jurisprudence are outdated and cannot account for the dramatic rise of the administrative state. See John A. Fairlie, The Separation of Powers, 21 Mich. L. Rev. 393, 393–94 (1923); Arthur S. Miller, Separation of Powers: An Ancient Doctrine Under Modern Challenge, 28 Admin. L. Rev. 299, 300 (1976); Malcolm P. Sharp, The Classical American Doctrine of the “Separation of Powers,” 2 U. Chi. L. Rev. 385, 386–87 (1935) (identifying Aristotle’s Politics as containing the “original statement of the doctrine” of mixed regimes “closely related in classic American political writing to the separation of powers”). While Gordon S. Wood, The Creation of the American Republic, 1776–1787, at 152–53 (1998) rejects the mixed-regime view of separation of powers, John Hart Ely, The Apparent Inevitability of Mixed Government, 16 Const. Comment. 283, 292 (1999) acknowledges its inevitability. Part III argues that while Wood rightly points out the demise of the British theory, the Founders’ rejection of the sociological version of mixed government in fact signaled a return to the basic Aristotelian model.Show More

This Article makes three main contributions. First, it fleshes out the theory of separation of structures as distinct from contemporary scholarly approaches. Second, it writes the intellectual history of separation of structures, which has been an integral part of the separation of powers enterprise since its inception, including at the Founding. Third, it explores the scholarly and doctrinal implications of structural separation of powers. In particular, adjudicating the constitutionality of agency structures requires methodological pluralism that incorporates the normative values underlying the structural design. That is, under separation of structures, current doctrine should evolve beyond the formalism heavily criticized by scholars. This structural framework thus provides a limiting principle to the doctrine of Free Enterprise Fund, Seila Law, and Collins v. Yellen. Further, congressional delegation to agencies cannot be conceptualized as a violation of separation of powers on the sole ground that delegation allows executive branch agencies to exercise legislative power. Instead, advocates of a muscular nondelegation doctrine often fail to recognize that agency structure can mitigate potential violations of functional separation of powers. Both implications are urgent in today’s doctrinal milieu. Not only does the Court continue to entrench its agency-structure jurisprudence—it appears poised to extend the nondelegation doctrine.³⁰ 30.See infra Subsection IV.B.3.Show More

The remainder of the Article proceeds as follows. Part I situates separation of structures within the existing scholarly models. Part II turns to the classical and early-modern origins of separation of structures. Part III examines the adoption of separation of structures as part of Founding-era constitutional design. Part IV discusses doctrinal and scholarly implications.

1.  E.g., Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 582 (1952); Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020); Stern v. Marshall, 564 U.S. 462, 469 (2011); Gundy v. United States, 139 S. Ct. 2116, 2129 (2019). ↑
2.  272 U.S. 52, 119 (1926). ↑
3.  Humphrey’s Ex’r v. United States, 295 U.S. 602, 629 (1935); see also Aditya Bamzai, Taft, Frankfurter, and the First Presidential For-Cause Removal, 52 U. Rich. L. Rev. 691, 699–701 (2018) (recounting Justice Sutherland’s attempts to distinguish Humphrey’s Executor from Myers based on the term of years established in the FTC’s organic statute and the quasi-legislative, quasi-judicial character of the FTC). ↑
4.  Morrison v. Olson, 487 U.S. 654, 691–92 (1988); Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); see Jack Goldsmith & John F. Manning, The Protean Take Care Clause, 164 U. Pa. L. Rev. 1835, 1841–43 (2016). ↑
5.  Commodity Futures Trading Comm’n v. Schor, 478 U.S. 833, 851–52 (1986); Stern, 564 U.S. at 482–83; see William Baude, Adjudication Outside Article III, 133 Harv. L. Rev. 1511, 1519–21 (2020) (reconciling non-Article-III adjudication with functional separation of powers); infra Figure 2. ↑
6.  Morrison, 487 U.S. at 671; Edmond v. United States, 520 U.S. 651, 663 (1997); see United States v. Arthrex, Inc., 141 S. Ct. 1970, 1985–86 (2021). ↑
7.  Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2192 (2020); Collins v. Yellen, 141 S. Ct. 1761, 1787 (2021). ↑
8.  Seila Law, 140 S. Ct. at 2202 (quoting Bowsher v. Synar, 478 U.S. 714, 730 (1986)). ↑
9.  Daryl J. Levinson, The Supreme Court 2015 Term—Foreword: Looking for Power in Public Law, 130 Harv L. Rev. 31, 37 (2016). ↑
10.  Schor, 478 U.S. at 851. ↑
11.  Seila Law, 140 S. Ct. at 2193. ↑
12.  Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); see also Lisa Shultz Bressman, What Seila Law Says About Chief Justice Roberts’ View of the Administrative State, U. Chi. L. Rev. Online (Aug. 27, 2020), https://lawreviewblog.uchicago.edu/2020/08/27/seila-bress‌man [https://perma.cc/97HT-CNRH] (arguing that Seila Law changes existing jurisprudence by “let[ting] the structure of the agency determine the degree of presidential control over its principal officers”). ↑
13.  Stephen L. Carter, The Independent Counsel Mess, 102 Harv. L. Rev. 105, 105 (1988). ↑
14.  M. Elizabeth Magill, The Real Separation in Separation of Powers Law, 86 Va. L. Rev. 1127, 1129 (2000); see also Jerry L. Mashaw, Of Angels, Pins, and For-Cause Removal: A Requiem for the Passive Virtues, U. Chi. L. Rev. Online (Aug. 27, 2020), https://lawreview‌blog.uchicago.edu/2020/08/27/seila-mashaw [https://perma.cc/BMT9-NK6C] (characterizing the doctrine on for-cause removal as a “jurisprudential train wreck”). ↑
15.  See infra Part I. ↑
16.  See infra notes 32–36 and accompanying text. ↑
17.  See infra notes 40–48 and accompanying text. Strict separation-model theorists question the legitimacy of the modern administrative state, prominently by appeals to the nondelegation doctrine. E.g., Gary Lawson, Delegation and Original Meaning, 88 Va. L. Rev. 327, 332 (2002) (arguing that the “civics-book model of legislators legislating, executives executing, and judges judging has enormous intuitive—and legitimating—power,” which explains the modern obsession with the nondelegation doctrine despite its disuse); Ilan Wurman, Nondelegation at the Founding, 130 Yale L.J. 1490, 1493–94 (2021) (arguing there is significant evidence the Founding generation adhered to a robust nondelegation doctrine that was keenly faithful to traditional separation of powers); see also infra Subsection IV.B.3. ↑
18.  See infra notes 49–55 and accompanying text. ↑
19.  See infra notes 56–57 and accompanying text. ↑
20.  See infra Section I.C. ↑
21.  See infra notes 73–78. ↑
22.  See infra Sections I.A–B. ↑
23.  E.g., Seila Law LLC v. CFPB, 140 S. Ct. 2183, 2191–92 (2020); Free Enter. Fund v. PCAOB, 561 U.S. 477, 492 (2010); Collins v. Yellen, 141 S. Ct. 1761, 1770–71 (2021). ↑
24.  See infra Section II.A. ↑
25.  See infra Section II.B. ↑
26.  See infra Section II.D. ↑
27.  See infra Section II.C. ↑
28.  See infra Part III. ↑
29.  Most scholars give only cursory treatment to the Aristotelian origins of separation of powers. See, e.g., Steven G. Calabresi, Mark E. Berghausen & Skylar Albertson, The Rise and Fall of the Separation of Powers, 106 Nw. U. L. Rev. 527, 529–36 (2012). Serious assessments of classical political philosophy’s contribution to modern jurisprudence are outdated and cannot account for the dramatic rise of the administrative state. See John A. Fairlie, The Separation of Powers, 21 Mich. L. Rev. 393, 393–94 (1923); Arthur S. Miller, Separation of Powers: An Ancient Doctrine Under Modern Challenge, 28 Admin. L. Rev. 299, 300 (1976); Malcolm P. Sharp, The Classical American Doctrine of the “Separation of Powers,” 2 U. Chi. L. Rev. 385, 386–87 (1935) (identifying Aristotle’s Politics as containing the “original statement of the doctrine” of mixed regimes “closely related in classic American political writing to the separation of powers”). While Gordon S. Wood, The Creation of the American Republic, 1776–1787, at 152–53 (1998) rejects the mixed-regime view of separation of powers, John Hart Ely, The Apparent Inevitability of Mixed Government, 16 Const. Comment. 283, 292 (1999) acknowledges its inevitability. Part III argues that while Wood rightly points out the demise of the British theory, the Founders’ rejection of the sociological version of mixed government in fact signaled a return to the basic Aristotelian model. ↑

30.  See infra Subsection IV.B.3. ↑