abstract.The Supreme Court’s approach to the Constitution’s separation of powers is a puzzle. Although the Justices appear to agree on the doctrine’s goals, in almost every important line of cases the Court oscillates between hard-edged rules and open-textured standards. The Court’s seemingly erratic doctrinal shifts cannot be wholly explained by changes in the bench’s personnel or methodological fads. This Article isolates and analyzes pervasive doctrinal cycling between rules and standards as a distinctive element of separation-of-powers jurisprudence. We break from previous scholarship critical of the Court’s zigzagging, and instead consider whether purposeful cycling between rules and standards might be justified as a judicial strategy for implementing the separation of powers. We further develop a new theoretical account of the separation of powers where doctrinal cycling might be justified on two key assumptions: First, the separation of powers promotes a plurality of normative ends, and second, it does so in the context of a more heterogeneous institutional environment than a singular focus on the interplay of the three great branches would suggest. Doctrinal cycling between rules and standards could be used, at least in theory, to manage normative pluralism and police this “thick political surround” when simpler, more straightforward regulatory strategies would fail. This rational reconstruction of the feasible judicial role in the separation-of-powers context provides a benchmark for evaluating observed doctrinal oscillations, and, more generally, for determining whether courts possess the necessary institutional resources to promote separation-of-powers values.

author. Aziz Z. Huq is the Frank and Bernice J. Greenberg Professor of Law, University of Chicago. Jon D. Michaels is Professor of Law, UCLA School of Law. The former’s work was supported by the Frank Cicero, Jr. Faculty Fund. For helpful comments, we thank Josh Chafetz, David Fontana, Tara Leigh Grove, Randy Kozel, David Pozen, and Sai Prakash. Steve Lindsay and a terrific team of Yale Law Journal editors provided thoughtful and penetrating suggestions and edits, and we are grateful for their careful and diligent work throughout the editing process.

Introduction

The Supreme Court’s separation-of-powers jurisprudence is a puzzle. The Court endorses James Madison’s conviction that institutional separation is a “sacred” element of the Constitution’s grand design.1 It also accepts the conventional understanding that separation of powers “make[s] Government accountable” and “secure[s] individual liberty.”2 Yet beyond those broad strokes, the Court seems unmoored and unprincipled when it translates the separation of powers into legal doctrine. In several lines of cases, the Court oscillates between using rules and using standards, pivoting with a surprising alacrity that cannot be explained by changes in the bench’s personnel, macro-level shifts in the relative power of the political branches, or the ebb and flow of jurisprudential fads.

Consider three recent illustrations:

Presidential removal power:In Morrison v. Olson, the Court employed an open-textured standard to uphold a congressional limitation on the President’s Article II authority to fire an executive official.3 But in the next major challenge to such congressional limits on the President’s removal power, Free Enterprise Fund v. Public Co. Accounting Oversight Board, the Court refused to apply Morrison and instead imposed a hard-edged rule.4

Limits on Article I tribunals: In Stern v. Marshall, the Court adopted a rule to reject the authority of a non-Article III bankruptcy court to issue a final judgment on a particular state-law counterclaim.5 Only four years later, though, the Court in Wellness International Network, Ltd. v. Sharif rejected “formalistic and unbending rules” of the kind applied in Stern in favor of a “practical effect” standard.6

Congressional regulation of presidential foreign relations powers: When analyzing the constitutionality of legislative constraints on the President’s wartime actions, courts have relied heavily on Justice Jackson’s concurrence in Youngstown Sheet & Tube Co. v. Sawyer.7 Yet such almost reflexive reliance obscures considerable oscillation in the application. Specifically, in applying Justice Jackson’s framework, the Court alternatively reads statutes as narrow rules (thereby authorizing only limited presidential engagements) or as open-textured standards (effectuating delegations of broad authority to the President).8 The result is a jurisprudence that cycles between pro-presidential and pro-congressional positions.

Seeming inconsistencies within and between the separation-of-powers doctrines taunt and frustrate commentators. Many simply wash their hands of what they see as an “incoherent muddle.”9

This Article is the first to consider the Court’s mixed approach as a potential solution to the particularly thorny problems posed by the separation of powers.10 It makes both a positive and a normative contribution. As a positive matter, we show that the Court’s separation-of-powers case law can be understood as a form of cycling between rules and standards.11 Normatively, we develop a potentially justificatory account for such rules-standards cycling. This account starts with two structural premises of the separation of powers. Reasoning from those foundational principles, we argue that a court might sensibly resort to rules-standards cycling as a way to promote institutional contestation over conflicting normative values, encouraging salutary forms of confrontation, compromise, and cooperation within judicially imposed boundaries. This theoretical result, we emphasize, is a “proof of concept”: it provides a benchmark to evaluate existing precedent and then to analyze the viability of judicial enforcement of the separation of powers.

Central to our normative account are two key, if often overlooked, assumptions of the Constitution’s separation of powers: normative pluralism and institutional heterogeneity. Normative pluralism, our first assumption, recognizes and endorses a multiplicity of constitutional values infusing our federal government’s structure. These values include, but are not limited to, liberty, effective administration, democratic accountability, the rule of law, and the prevention of tyranny. These constitutional values, moreover, cannot be easily aggregated, ordered, or reconciled. Consider, for example, the way in which the separation of powers promotes efficiency by eliciting institutional specialization among the branches12 and prevents tyranny by diffusing power between different branches;13 such aims are not necessarily or inevitably commensurable. Indeed, they regularly conflict. Given normative pluralism, the separation of powers does not (and, in truth, cannot) require the maximization of a single value. It calls instead for the harmonization—or, at the very least, the cycling through—of competing, conflicting values.14

As for institutional heterogeneity, the separation of powers is properly conceived as something more complex than the standard “three-branch problem.”15 The three branches of the federal government do not stand in splendid isolation. Nor do they operate as monoliths. Rather, they are enveloped and infused by a teeming ecosystem of institutional, organizational, and individual actors within as well as outside of government. Within the federal government, congressional committees, a cadre of civil servants, and an assortment of independent agencies and other species of bureaucratic faction represent just a fraction of the denizens of this fertile ecosystem. Outside government are a jostling array of lobbyists, political party structures, media actors, and domestic and foreign interest groups. All told, these internal and external actors create a thick political surround that shapes and channels action by the three branches, sometimes facilitating and sometimes frustrating the realization of the multiple separation-of-powers values.16

A well-grounded separation-of-powers jurisprudence must account for both normative pluralism and the thick political surround. On our reckoning, such a jurisprudence—sensitive to the multiplicity of normative values and mindful of the various, thick patterns of institutional contestation inside and around the three branches—might well take the seemingly incoherent form of oscillating rules and standards. Such oscillations pose no shortage of difficulties when observed in judicial doctrine. Generally speaking, we celebrate doctrinal stability and consistency. We make sense of various lines of constitutional jurisprudence in terms of either rules orstandards—but not both. First-year law students have been taught certain rules (such as the First Amendment ban on prior restraints and the federalism-inspired prohibition on commandeering) and certain standards (such as those used for the Eighth Amendment and the dormant Commerce Clause doctrine) for generations. Not so with interbranch relations.

Neither standards nor rules alone are likely to vindicate the separation of powers. In important part, this is because, depending on the context, an oscillation between rules and standards has the potential to promote the sort of pluralistic political dynamics that the separation of powers is intended to foster. Standards invite flexibility, experimentation, negotiation, and contestation.17 They tend to enable a wider array of actors, championing a broad range of normative values, to enter the political arena and make their presence known. A jurisprudential turn to rules may, and frequently does, become necessary, to discipline some domineering actors, put a decisive end to unhealthy or abusive forms of engagement, and to clear paths for even greater democratic contestation down the road. Rules and standards are thus both needed to open and close the floodgates as institutional and political dynamics and demographics change. The willingness to toggle between the two could encourage an organic, dynamic form of normative pluralism in a thick political surround and could deter (and, if necessary, correct) forms of contestation that become corrupted or decayed. The resulting doctrinal movements echo patterns observed in other areas of the law, but rarely identified or analyzed in constitutional law.18

To be clear, we do not assert that the Court consciously styles itself as a regulator of the thick political surround in the fashion we describe. Our theoretical account of the separation of powers has not been articulated by any Justice, nor do we have reason to think that it has in fact animated any member of the Court on a conscious level. Our claim and thus our ambition is a more modest one. What we offer is a “rational reconstruction.”19 We take bodies of precedent that “may seem confused and disorderly, partly or potentially conflicting, gappy in places,” and then “put them back together, to reconstruct them in a way that makes them comprehensible because they are now shown as parts of a well ordered though complex whole.”20

Our rational reconstruction provides a much-needed normative baseline against which the Court’s actual interventions can be judged. As a result, our main contribution is to clarify what counts as jurisprudential success (although we also offer tentative thoughts on several lines of separation-of-powers cases). We emphasize that even with our criteria in mind, reasonable people can still disagree about whether courts are institutionally competent to play the role we describe, or whether particular strands of precedent make sense. Courts, like the other branches, are buffeted by an external ecosystem of interest groups, and have only limited epistemic and political resources. We do not here aim to settle the hard and contested question whether separation-of-powers challenges ought to be justiciable given these constraints. But we hope that the terrain upon which that disagreement arises will be henceforth perceived with greater perspicuity.

Because we recognize the central, albeit neglected, role of normative pluralism and the thick political surround, we are compelled to part ways from settled practice in another respect. Leading scholars have critiqued oscillations in separation-of-powers jurisprudence as incoherent in large part because they divide the doctrinal world between formalism and functionalism.21 John Manning distills what has long been conventional wisdom. “[L]egal academics have . . . discerned two basic approaches to separation of powers doctrine”: a “functionalist approach” and a “formalist approach.”22 Manning describes this conventional wisdom as “accurate[],” a view shared by most leading commentators on the separation of powers.23 Even recent efforts to transcend the dichotomy are framed largely as variations of that debate.24 Although functionalism and formalism dominate discussions of the separation of powers, we understand standards and rules as part of a more helpful, flexible, and forgiving vocabulary to use when describing and making sense of abrupt, oscillating interventions.

We think that a rules-versus-standards framing is more useful for two reasons. First, as a practical matter the formalism-versus-functionalism characterization tends to imply an overarching jurisprudential worldview. Generally speaking, judges are either formalists or functionalists, but not both—and must, among other things, vindicate the separation of powers in a manner consistent with their chosen worldview. As a result, any observed cycling is reflexively criticized as apostate. Such exclusivity arises in part because formalism tends to be associated with both textualist and originalist theories of constitutional interpretation, whereas functionalism is more often associated with pragmatic, dynamic, and hermeneutical approaches.25 This association remains generally true despite recent moves by some originalists to criticize exclusive adherence to rule-based structures.26 Still, we worry that shoehorning the vindication of separation of powers into a judge’s chosen worldview gets things backward, a case of the tail wagging the dog as it were. This would never happen so long as we thought in terms of rules and standards, tools which are equally useful but are far less freighted. No judge would insist she be called a “standards” judge, come hell or high water.

Second, the formalism-versus-functionalism framing tends to characterize the central challenge of separation-of-powers jurisprudence as a dichotomous choice between two distinct, extreme positions. Not only is there no toggling between the two, there is also no occupying (or seeking to occupy) a middle ground. While functionalists tend to endorse loose standards that promote a flexible balancing of powers,27 formalists derive “readily ascertainable and enforceable rules” from the textual accounts of separation.28 But neither polar position allows for proper recognition of a key function of the separation of powers: the presence of multiple normative values within a thick political surround. Classically conceived formalism dictates a needlessly rigid focus on the three constitutional branches to the exclusion of the intrabranch and external actors that influence their interplay. True functionalism, by contrast, is too permissive. It seemingly permits any actor within the thick political surround to promote any and all conceivable values. Equally damning, classical formalists and functionalists inhabit a Manichean world, allowing few, if any, opportunities to oscillate between the two poles (or gravitate to a median position) as circumstances warrant. Rules and standards have at least the theoretical potential to enable a happier equilibrium, more capacious than a narrow focus on the three branches permits and yet still capable of imposing order and discipline upon an otherwise chaotic thick political surround.

At the same time, we recognize and credit contemporary efforts to conflate rules with formalism and standards with functionalism as important contributions to legal and scholarly conversations and debates, including on the separation of powers.29 Still, we think there is good reason to keep the two sets of terms distinct. For at certain moments, preservation of a stringently polarized formalism-functionalism divide undoubtedly remains conceptually and lexicographically useful.

***

Our analysis proceeds in three Parts. In Part I, we explain how several lines of existing separation-of-powers jurisprudence can fairly be characterized in terms of judicial cycling between rules and standards. In Part II, we introduce the foundational concepts of normative pluralism and institutional heterogeneity, and further show that both can be discerned, albeit imperfectly and partially, as animating and inflecting the trajectory of those several lines of case law. In Part III, we explore how cycling between rules and standards might, at least as a theoretical matter, be ranked as a sensible response to the ever-changing demands of normative pluralism and institutional heterogeneity. Here we link our theoretical predicates of the separation of powers to the concrete challenges of creating workable doctrines and jurisprudential principles. In concluding, we show how our analysis can be used to explore whether courts are appropriately positioned to advance separation-of-powers values given the nature of that task.

i. the ubiquity of rules-standards cycling in separation-of-powers jurisprudence

We set the foundation for our argument by demonstrating in this Part that the Supreme Court’s jurisprudence on the separation of powers routinely cycles between rules and standards. To that end, we employ now-canonical definitions of rules as legal norms given content before regulated subjects act, and standards as legal norms given content after regulated subjects act.30 Rules (both in general and in the separation-of-powers context) tend to be more restrictive than standards; standards, by contrast, leave the law more open to novel and unanticipated considerations. The correlation between rules and inflexibility, however, is not inevitable. As we will show, in some instances rules can leave regulated actors with great flexibility, while standards can impose heavy burdens of compliance.

To show the pervasiveness of rules-standards cycling, we analyze five discrete lines of authority. In Section I.A, we discuss the legislative delegation of regulatory authority. In Sections I.B and I.C, we consider the design of administrative agencies, including, particularly, the removal of agency officials and the level of judicial deference afforded to agency interpretations of statutes. In Section I.D, we discuss the adjudication of disputes by non-Article III actors. And, in Section I.E, we address the management of foreign affairs. By considering multiple, distinct lines of precedent, we broaden and strengthen our descriptive claim that cycling cannot simply be chalked up to personnel changes on the Court or to the waxing or waning of jurisprudential philosophies or fads. Previous analyses that zeroed in on only one or two of these lines, to the exclusion of the others, understandably failed to appreciate the frequency and transsubstantive reach of rules-standards cycling.

Before proceeding, two cautions should be sounded. First, we concede that our account of cycling is limited rather than universal. Not every line of separation-of-powers jurisprudence exhibits patterns of cycling. We consider those non-cycling lines at the very end of this Part. Second, we appreciate that not every instance of cycling examined in this Part is necessarily reasonable or principled. Given our immediate task of simply mapping the doctrinal landscape, we do not grapple here with the logic or prudence of cycling. We defer that exercise to Parts II and III.

A. Legislative Delegations

The nondelegation doctrine seeks to restrict Congress from delegating legislative power in a manner inconsistent with its Article I duties.31 The doctrine is difficult to apply because it is not a simple task to determine what constitutes legislative power. In the modern era, the Court has required Congress to provide an “intelligible principle” when delegating lawmaking authority to administrative agencies. This requirement, first articulated in the 1928 case J.W. Hampton, Jr., & Co. v. United States,32 is meant to ensure that Congress guides, even if it does not entirely specify, the terms of agency action. J.W. Hampton remains good law today.33 But its practical effect has fluctuated. Periods of dutiful adherence to J.W. Hampton’s standard-like formulation have been punctuated by carve-outs of categorical, rule-like exceptions to the standard’s applicability.

On only two occasions has the Court struck down legislative delegations on constitutional grounds. Both cases involved provisions of the National Industrial Recovery Act (NIRA). In Panama Refining Co. v. Ryan,34the Court rejected section 9(c) of the NIRA, underscoring Congress’s failure to furnish an intelligible principle directing the President’s prohibition of the interstate transportation of excess petroleum and petroleum products. Simply stated, the Court found that “the Congress has declared no policy” to guide the President.35 Then, in A.L.A. Schechter Poultry Corp. v. United States,36 the Court held that section 3 of the NIRA, authorizing the President to approve of privately arrived-at codes of fair competition for various industries, also lacked an intelligible principle. The provision therefore left recipients of lawmaking power free to “roam at will.”37

Panama Refining and Schechter are outliers. Before 1935, the intelligible-principle requirement was glossed as an open-textured standard. Panama Refining and Schechter read it as a rule demanding a high degree of legislative specificity. But this rule-like understanding did not last. After 1935, the Court returned to treating the intelligible-principle imperative as a standard, giving Congress considerable flexibility in its delegations.38 Explaining this return to a standard-like formulation, the Court underscored its “practical understanding that in our increasingly complex society, replete with ever changing and more technical problems, Congress simply cannot do its job absent an ability to delegate power under broad general directives.”39

Schechter did more than just construe the intelligible-principle imperative as a rule. It also layered on an additional, equally categorical rulebarring delegations of state authority to private parties. This “private delegation” carve-out barred all delegations of rulemaking authority to private parties. It has never morphed into a standard and remains hard-edged today, foreclosing such delegations irrespective of how carefully and thoroughly Congress specifies an intelligible principle.40

Over time, two further categorical rules—more carve-outs from the “intelligible principle” standard—surfaced. Like Schechter’s private-delegation prong, both turn on the recipient of the delegation, not the scope or specificity of the power delegated. First, INS v. Chadha41 involved a challenge to Congress’s one-house veto over administrative decisions that the legislature delegated to the Attorney General. In effect, the veto empowered either the House or Senate to unilaterally reverse the Attorney General’s decision to suspend deportation proceedings against an undocumented person. The Court held this legislative veto unconstitutional: Congress may not delegate to itself a decision-making role outside of the one prescribed in Article I, Section 7 of the Constitution, regardless whether the single house of Congress exercising the veto is constrained by an intelligible principle furnished in the authorizing legislation. In dissent, Justice White expressed frustration at the abrupt imposition of a rule-like limitation on congressional delegations. All congressional delegations, Justice White insisted, should be evaluated pursuant to the Court’s general “intelligible principle” standard. Justice White posited the legislative veto as simply a reflection of a limited delegation of lawmaking power and a concession to practical administrative needs: “If the effective functioning of a complex modern government requires the delegation of vast authority which, by virtue of its breadth, is legislative or ‘quasi-legislative’ in character, I cannot accept that Art[icle] I . . . should forbid Congress from qualifying that grant with a legislative veto.”42

Second, in the Line Item Veto Act of 1996 (LIVA),43 Congress delegated to the President authority to cancel or nullify certain provisions of appropriations bills within five days of his signing those bills into law. Wielding that authority, President Clinton canceled a provision of the Balanced Budget Act of 1997 and a provision of the Taxpayers Relief Act of 1997. In Clinton v. City of New York, the Court invalidated LIVA, holding that Congress lacked authority to delegate cancellation power to the President.44 As in Chadha, the Clinton ruling did not turn on whether the delegation contained an intelligible principle, which the Act indeed seemed to provide.45 Instead, the Court imposed a bright-line rule: delegations to the executive, however precise, may not involve formal revisions to acts of Congress.46

The Clinton Court’s imposition of a bright-line rule invoked a sharp dissent from Justice Scalia. From Justice Scalia’s perspective, LIVA’s delegation was indistinguishable from an ordinary delegation of lawmaking authority. He conceded that it was possible to read the Constitution to admit no delegations of lawmaking power but cautioned that the Court had never hewed to such a reading of the Constitution. Instead, Justice Scalia homed in on the bright-line distinction, implicit in the Clinton majority’s ruling, between “cancelling” legislative provisions and furthering legislative provisions through rulemaking. He insisted that the Constitution “no more categorically prohibits the Executive reduction of congressional dispositions in the course of implementing statutes that authorize such reduction, than it categorically prohibits the Executive augmentation of congressional dispositions in the course of implementing statutes that authorize such augmentation—generally known as substantive rulemaking.”47

We appreciate that the Court in neither Clinton nor Chadha would classify its holding as a modification of the nondelegation doctrine. Only the dissenting Justices, who emphasized the doctrine’s openness to unforeseen pragmatic considerations, would. But that is precisely our point. The nondelegation doctrine is more dynamic than generally appreciated. Clinton and Chadha alike conjured new and unexpected limitations. They drew bright-line rules carving out narrow exceptions limiting the application of the still-capacious intelligible principle standard that applies liberally when the delegation is to a bona fide administrative agency.48 Whether Clinton and Chadha’s results are justified or not, our core point here is that they, like Schechter’s private delegation doctrine, illustrate how a broad standard that seems to cover the waterfront of delegation jurisprudence can prove amenable over time to rule-based limitations.

B. The Removal Power

Our second example concerns the separation-of-powers controversy over who controls top-ranking agency officials. The key question in these cases is whether the President must have complete and unfettered authority to remove agency officials, consistent with her duty to take care that the laws are faithfully executed.49 Evidence from the Founding period is mixed. The Constitution vests in the President the power to appoint principal officers of the United States,50 but is silent on the question of subsequent forms of control. Madison suggested that the power to unilaterally remove officers was an inherent element of Article II’s grant of executive power and the President’s corresponding duty to take care that the laws be faithfully executed.51 The Court in Marbury v. Madison seemingly acknowledged as much, at the same time conceding that Congress may at times restrict that unilateral presidential power.52

In the twentieth century, with the rise of the administrative state, questions of control over executive personnel took on greater importance. The modern jurisprudence on agency control is complex, characterized by oscillations akin to those observed in the nondelegation domain. We start with the bright-line rule announced in Myers v. United States.53 In Myers, the Court rejected congressionally imposed limitations on the removal of postal officials and announced what seemed like a comprehensive and categorical rule. Explaining that the power to remove officials “is an incident of the power to appoint them,”54 the Court insisted that complete presidential control was necessary for the Chief Executive to “properly supervise and guide [officials’] construction of the statutes under which they act in order to secure that unitary and uniform execution of the laws which Article II of the Constitution evidently contemplated in vesting general executive power in the President alone.”55

The seeming absolutism of Myers yielded in short order to a softer-edged standard. Just a decade after deciding the postal removal case, the Court in Humphrey’s Executor v. United States56rejected Myers’s rule as overbroad. In its place, the Court substituted an open-textured standard for identifying classes of officials Congress could insulate from summary presidential removal. Humphrey had been one of the Federal Trade Commission’s (FTC) Commissioners, a carryover from the Coolidge and Hoover administrations that President Franklin D. Roosevelt was eager to dismiss.57 After Humphrey’s death, his executor challenged the termination as inconsistent with the statutory protections Congress afforded commissioners, which limited the grounds for removal to “inefficiency, neglect of duty, or malfeasance in office.”58 The Court held that some agency officials need not fall under the President’s unfettered control. If those officials’ responsibilities are “quasi-legislative” or “quasi-judicial,” the Court opined, Congress may restrict the President’s removal powers by requiring a showing of good cause.59 Hence, a presidential power framed in Myers as categorical and rule-like was transformed into a standard that would require a searching and fact-sensitive ex post inquiry. Demonstrating the difficulty of applying this standard, later courts questioned whether the Humphrey’s Executor Court was even right on its facts given the FTC’s extensive executive responsibilities.60

Cases challenging the constitutionality of removal provisions occur infrequently. A subsequent pair in 1958 and 1986 followed Humphrey’s Executor and involved the Court investigating whether a terminated official’s responsibilities were primarily executive, judicial, or legislative.61 In 1988, however, the Court again changed tack. Rather than sliding back to a rule, it instead pushed even further in the direction of an open-ended standard. In Morrison v. Olson,62 a case concerning whether Congress could insulate a special prosecutor from removal by the Attorney General absent good cause, the Court disavowed Humphrey’s Executor. Morrison insisted that the Humphrey’s Executor standard was too unreliable and, perhaps, too rigid.63 Instead the Court fashioned a new, arguably muddier standard that obligated courts to determine “whether the removal restrictions are of such a nature that they impede the President’s ability to perform his constitutional duty.”64

Dissenting in Morrison, Justice Scalia lamented the long-abandoned bright-line rule of Myers.65 “The Court,” Justice Scalia bemoaned, “has . . . replaced the clear constitutional prescription that the executive power belongs to the President with a ‘balancing test.’ What are the standards to determine how the balance is to be struck, that is, how much removal of presidential power is too much?”66

The Court would effectively repudiate one element of Morrison concerning appointments in 1997,67 but Justice Scalia would have to wait twenty-odd years for the return of a bright-line rule to the Court’s removal jurisprudence. In Free Enterprise Fund v. Public Co. Accounting Oversight Board68 (PCAOB), the Court confronted an administrative regime where a double layer of political insulation immunized the titular Board from direct presidential control. Commissioners of the Securities and Exchange Commission (SEC) were understood to be subject to presidential removal only for cause. Similarly, PCAOB members could be removed by the SEC Commissioners only by a showing even greater than the typical cause.69 The Court could have applied the Morrison standard and assessed whether this dual limitation on full presidential control impeded the President’s ability to perform his constitutional duties, finding that this particular set of limitations represented too great a restriction. Instead, the Court without warning pivoted to a new bright-line rule. Repeatedly insisting that two layers of for-cause insulation from presidential control present concerns qualitatively different from one layer of insulation, the Court held that two layers were per seunconstitutional70—adopting what seems for the moment to be a rule-like carve-out from the Morrison standard, at least in special cases of double insulation. Whether Free Enterprise Fund’s new rule will be extended remains to be seen.

C. Deference to Agency Legal Interpretations

A third line of cases implicating separation-of-powers questions concerns the deference afforded to administrative agencies’ statutory interpretations. Lest there be any doubt, the degree to which courts cede interpretative authority is very much a matter of interbranch relations. The Court has construed deference in terms of Congress delegating interpretive legal authority to agencies as opposed to judges.71 Because the scope of such interpretive authority helps determine agencies’ policymaking power, it has become a heavily litigated battleground for separation-of-powers disputes.

We start with Skidmore v. Swift & Co.,72 a 1944 dispute about whether private firefighters at a meatpacking plant qualified for overtime pay while on call to battle potential fires. Skidmore turned partially on how much deference the Court would give to the Labor Department officials’ interpretation of the relevant overtime provisions of the Fair Labor Standards Act. The Court held that the Department’s interpretation was “entitled to respect.”73 The Court explained that the weight given to an agency’s judgment “in a particular case will depend upon the thoroughness evident in [the agency’s] consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade.”74 In other words, the Court used a standard encompassing a heterogeneous array of pragmatic considerations to measure the deference owed to agency interpretations.

Forty years later, the Court expressly reoriented its deference jurisprudence in a more rule-like direction. In its landmark decision in Chevron v. National Resources Defense Council, the Court held that whenever statutes authorizing agency action are vague, agency interpretations of those statutes are valid provided they constitute a permissible construction of the statutory text.75 The permissibility of Chevron’s formulation might suggest a continued devotion to standards. But Chevron abandoned the plethora of factors used by a generation of lawyers, judges, and policymakers working under Skidmore. Under Chevron, the interpretive deference given to agencies no longer depended on a searching, case-specific analysis. Instead, only one fact mattered: whether the relevant statute is ambiguous. If so, agencies are automatically entitled to deference.

Chevron is thus fairly characterized as a rule in our terminology. Nevertheless, it enlarged the space (in a categorical fashion) for discretionary action by regulated actors. At the same time, Chevron also implicitly assigned to the courts considerable discretion in deciding how to interpret whether a statute is indeed ambiguous.76 The Court did not impose any stable “ranking” of canons or other interpretive presumptions, leaving later benches to select between diverse approaches to statutory interpretation.77 As a result of this diversity of interpretive approaches at what is often called Step One, Chevron’s one-type-of-deference rule (which, again, applies only once courts find a statute to be ambiguous) may “not necessarily yield greater predictability and law-like behavior among judges than context-saturated standards.”78 Hence, notwithstanding its rule-like formulation, Chevron still enables a measure of cycling between different approaches to statutory interpretation.

Despite this elasticity, the sharper, one-type-of-deference rule installed in Chevron soon morphed into a compound, rule-standard analysis. The leading case here is the 2001 decision United States v. Mead Corp.79 Mead held that when agencies acted with the “force of law,” the Court would accord them Chevron deference.80 Otherwise, agencies likely merit only Skidmore deference.81 Mead thus contains a complex intermingling of rule-like and standard-like features. After Mead, agencies engaging in notice-and-comment rulemaking or in formal adjudications would be presumed to have acted with the force of law and would continue to receive Chevron deference.82 Those that employed more informal decisional processes are now apt to receive only the lesser, more open-ended Skidmore deference, wherein the measure of such deference turns on some gestalt-like computation of the plural Skidmore factors.83 Adding complexity and uncertainty to the deference regime, those factors have since multiplied, as later cases, notably Barnhart v. Walton84 have added to the list of considerations relevant to the application of the Mead/Skidmore standard. In effect, the oscillation from Skidmore to Chevron to the Mead mid-point suggests that we now find ourselves in a hybrid rule-standard world—with courts employing rules or standards based on the types of procedures agencies use.

D. Adjudication by Non-Article III Judges: The Bankruptcy Example

Notwithstanding Congress’s wide power to create federal courts, the national legislature has created any number of alternative adjudicative offices lacking the lifetime tenure and salary protections of traditional Article III appointments.85 These alternative tribunals include territorial courts, military courts, bankruptcy courts, tax courts, and a bewildering array of different agency adjudicators, including immigration and Social Security courts.86 Among this varied contingent of non-Article III adjudicators are more than 1,500 administrative law judges (ALJs) working across more than twenty-five agencies.87 ALJs outnumber Article III judges almost two-to-one,88 and decide more than 250,000 cases a year.89 The 372 bankruptcy judges90 are another important group to whom Congress grants front-line adjudicative responsibilities, albeit under the supervision of federal district court judges.91 ALJs and bankruptcy judges alike decide matters of large financial and personal significance.

Non-Article III adjudication has generated two related, but not wholly consistent, lines of precedent. One is characterized by cycling between rules and standards, while the other evinces marked stability. In brief, the Article III treatment of administrative agency adjudication has remained remarkably stable and conciliatory for more than 125 years. By contrast, the Court’s treatment of bankruptcy courts has oscillated wildly between restrictive rules and enabling standards. Scholars have puzzled over the divergent treatment of agency and bankruptcy adjudication.92 But that concern is secondary to our point here: that cycling is an important feature of bankruptcy jurisprudence. We focus on the latter here, holding off on the jurisprudence of agency adjudication until Section I.F, the Section devoted to doctrinal lines that do not exhibit rules-standards cycling.

Bankruptcy courts might seem to be a minor separation-of-powers problem. They operate under the supervision of Article III judges, not within executive departments. And they have a long historical pedigree: English bankruptcy practice, with which the Framers were well acquainted, allocated front-line adjudicative responsibilities to non-judicial commissioners.93 Nevertheless, the Court’s treatment of bankruptcy judges has cycled twice during the twentieth century, with a majority of Justices seemingly pivoting erratically from loose standard to rigid rule, reinstalling the rule after a period of inattention by the Supreme Court, and then finally relaxing somewhat, settling on another standard to assess constitutionality. The ensuing pattern, such as it is, does not lend itself to easy explanation, such as judicial turnover at One First Street.

The first separation-of-powers challenge to non-Article III bankruptcy judges arose under an 1898 Act of Congress.94 This statutory scheme drew a fuzzy distinction between “summary” jurisdiction over matters related to the estate (which bankruptcy judges actually or constructively possessed) and “plenary” jurisdiction over other matters (which they lacked).95 The distinction’s fuzziness meant that it fell to the federal courts to define bankruptcy courts’ powers.96 The Court initially applied a flexible standard that accommodated a range of policy interests, including ones seemingly unforeseen by the enacting Congress. For example, in the 1966 case of Katchen v. Landy, the Court considered bankruptcy judges’ authority to designate creditors’ claims against an estate as voidable preferences, payments made in anticipation of bankruptcy to avoid its distribution rule.97 Voidable preferences might be understood to be unrelated to the state given that they by definition involve a pre-filing transfer to a third party. At least at first blush, they thus seem to fall outside a bankruptcy judge’s summary jurisdiction. Moreover, while the Katchen petitioners did not argue that Article III had been violated, they did press a Seventh Amendment right to a jury trial, in effect claiming that Congress would have assigned the matter to the wrong branch if voidable preferences could be resolved by a non-Article III official.98 Resisting that inflexible and rule-like logic, the Court emphasized cost and administrability considerations in concluding that bankruptcy judges could decide preference claims even though the petitioners lacked actual or constructive possession of the property in question.99 The Court thus treated the 1898 statute’s apparent limit on bankruptcy judges’ authority as a standard rather than a rule, invoking consequentialist considerations to stretch that authority as circumstances warranted.

Yet when later reconsidering the constitutional bounds of bankruptcy courts’ power in light of new legislation, the Court applied a rigid rule with no allowance for the forward-looking, practical concerns Katchen endorsed. In Northern Pipeline Construction Co. v. Marathon Pipe Line Co.,a plurality of the Court invalidated provisions of the 1978 Bankruptcy Reform Act allocating bankruptcy judges power to decide all matters “related to” a bankruptcy case.100 Writing for four Justices, Justice Brennan insisted that Article III adjudication could be ousted in only three “historically and constitutionally . . . exceptional” pockets: territorial courts, military courts, and the adjudication of “public rights” cases (i.e., suits between the government and its citizens).101 Justice Brennan distinguished “the restructuring of debtor-creditor relations, which is at the core of the federal bankruptcy power,” from the “adjudication of state-created private rights,” which fall outside that power.102 The Court used this rule-like distinction to invalidate the 1978 jurisdictional provisions in toto.103

Thirty years later, the Court once again revisited the separation-of-powers question in the bankruptcy court context. The Justices were confronted with Congress’s answer to Northern Pipeline. This answer, part of the Bankruptcy Amendments and Federal Judgeship Act of 1984, was structured around a list of sixteen “core” matters within bankruptcy judges’ reach.104 In the interim, the relevant Article III regime for bankruptcy remained rule-like, even as the Court continued to issue standard-based decisions regarding agency adjudication.105 One might have read these latter agency adjudication cases to presage an impending shift from rules to standards in the bankruptcy context too. After all, as noted above, both sets of cases involve the extension of adjudicatory responsibilities to non-Article III tribunals. Alternatively, we might characterize this era as one where the Northern Pipeline rule was, in practice, ignored, as bankruptcy judges went unmolested as they continued creatively interpreting the sixteen expansive new statutory fonts of power in light of practical considerations and policy imperatives.106

When the Court did return to the separation-of-powers question raised by bankruptcy courts, it reaffirmed its commitment to rigid rules foreclosing experimentation or new considerations. Thus, in Stern v. Marshall,107 the Court held that the statutory category assigned to bankruptcy judges of “counterclaims by the estate against persons filing claims against the estate”108 violated Article III.109 Reaching back to the category of “public rights” employed in Northern Pipeline, the Stern Court defined the permissible scope of bankruptcy judges’ power as reaching only issues “integral to the restructuring of the debtor-creditor relationship.”110 On its face, this test might be read as a standard, inviting consideration of new factors, but the Stern Court seemingly rejected any such reading as it deployed the test as a rule to narrow bankruptcy judges’ power. Stern hence retained, and even doubled down on, the Northern Pipeline approach.

Perhaps more surprisingly, Stern’s ruledid not endure. Just three years later, and in the absence of any change in the Court’s personnel, the same nine Justices who decided Stern cycled back to a more standard-like articulation. In Executive Benefits Insurance Agency v. Arkison, a unanimous Court held that any constitutional troubles regarding the assignment of adjudicatory responsibilities over bankruptcy could be “cured” if the bankruptcy judge’s ruling was treated as proposed findings of facts and legal conclusions, to be evaluated de novo by a district court.111 A year later, a six-Justice majority in Wellness International Network, Ltd. v. Sharif found litigant consent sufficient to vest the bankruptcy courts with power to enter a final judgment.112 Without abjuring Stern, the Court in Wellness International reached across domains, citing and centrally relying on precedent from the agency adjudication context that, as noted above, employed a much more flexible standard.113 Rather than applying an unbending rule, the Wellness International Court engaged in open-ended consideration of the “practical” consequences of its decision.114 The separation-of-powers regime in bankruptcy, in short, is well on its way back to the standard that produced Katchen in 1966.

As suggested above, the shifts documented above do not obviously correlate with ideological changes in the Court nor with clear shifts in jurisprudential methodologies. To the contrary, it is striking that Justice Brennan, a jurist rarely seen as an arch-formalist, penned the leading rule-like decision in Northern Pipeline. At this moment, moreover, Article III jurisprudence seems to be transitioning from rules to standards and drawing connections between agency adjudicators and bankruptcy judges despite the stability of the Court’s personnel and the historically separate treatment of those two doctrinal lines.

E. Foreign Affairs and National Security

The final separation-of-powers domain that cycles between rules and standards concerns foreign affairs. In the four contexts canvassed so far, it is the Court’s rule of decision that is the cycling pivot. In the foreign affairs domain, by contrast, the crucial rule of decision has remained largely stable. Echoing dynamics observed in the Chevron context, cycling instead occurs in the application of that rule. In these cases, the modal question—defined most famously by Justice Robert Jackson’s path-making opinion in Youngstown Sheet & Tube Co. v. Sawyer115—is whether Congress prohibits or allows the presidential actions in question. The balance of interbranch powers, therefore, largely depends on how the Court measures the scope of congressional permission.

Patterns of rules-standards cycling surface here in the context of statutory interpretation, rather than in the crafting of constitutional rules. Specifically, the Court sometimes reads an act of Congress as a sharp-edged rule, marking a clear delineation of what forms of executive initiative the legislature does and does not support. At other times, the Court glosses relevant statutory text as a malleable standard capable of accommodating novel and capacious considerations of congressional intent. When the Court treats the relevant statute as a standard, the executive is more likely to find supportive legal authority. The result, as in the lines of cases canvassed above, is a series of wide fluctuations in the stringency of separation-of-powers limitations.

Under Justice Jackson’s influential three-part typology, Presidents have “maximum” authority when acting “pursuant to an express or implied authorization of Congress”;116 uncertain authority when “ac[ting] in absence of either a congressional grant or denial of authority”;117 and their “lowest ebb” of authority when “tak[ing] measures incompatible with the expressed or implied will of Congress.”118 This framework might be understood as susceptible to a range of more or less pro-legislative readings. Subsequent courts, however, have not explicitly read Youngstown to invite context-specific judgments about interbranch balance, but have rather treated the Jackson approach as a general framework for the analysis of separation-of-powers questions.119 More conventionally, what is perhaps the Court’s most famous concurrence represents a theory of constitutionalism channeled through interpretative construction of congressional enactments. But like the Chevron opinion discussed above, it is silent on how courts interpret statutes.120 For its part, Congress has proved (perhaps unsurprisingly) incapable of consistently writing statutes that are resistant to diverse, even inconsistent, readings.

Applying the Youngstown framework to a range of ambiguous statutes, the Court has alternated between rule-like and standard-like readings of the relevant statutes. Evaluating President Truman’s seizure of the steel mills in the Youngstown case itself, a plurality of the Court read legislation concerning the resolution of labor disputes narrowly as a set of specific, rule-like permissions for presidential intervention that did not include the mass seizure of facilities at issue in that case. Indeed, Justice Black’s plurality opinion emphasized a negative inference drawn from the Congress’s failure to enact authorization of the sort that could justify such a seizure.121

The Court’s willingness to read enabling statutes as rules, however, did not endure. Three decades later, when the Court next confronted an arguably unilateral presidential intervention in the wake of the Iran hostage crisis, the Court read the relevant statutes loosely. Then-Justice Rehnquist’s opinion in Dames & Moore v. Regan acknowledged the absence of precise statutory authority for such executive conduct,122 which had damned President Truman some thirty years before. In a novel interpretive move, however, the Court then eschewed attention to any single statute and refused to draw any negative inference from the absence of express statutory authority. It instead directed attention to “the general tenor of Congress’ legislation in this area.”123 That is, the Court adopted an interpretive strategy that a majority of Justices in Youngstown had rejected. The more general body of relevant statutes passed by Congress granting the President emergency economic powers, coupled with Congress’s inability to “anticipate and legislate with regard to every possible action” the President might take, resulted in the executive having “broad discretion” to determine what steps were necessary to address novel international situations.124 More prosaically stated, the Court read the relevant statutes as a standard.125 More than twenty years later, a plurality of the Court in Hamdi v. Rumsfeld would use the latitudinarian approach of Dames & Moore to find detention authority in a tersely worded authorization of military force that made no mention of anything approximating detention.126

Yet the pivot to standards was neither stable nor consistent. In a pair of cases decided soon after Hamdi, the Court read enabling legislation narrowly as rule-like authorizations. These cases are instructive because of the ideologically divergent coalitions of Justices in the majorities of each. In the first of the two cases, Hamdan v. Rumsfeld, the Court invalidated the military commissions established by President George W. Bush at Guantanamo Bay because they failed to comply with, among other things, Article 36(b) of the congressionally enacted Uniform Code of Military Justice (UCMJ).127 Article 36(b) demanded that military commission and court-martial procedures “be uniform insofar as practicable.”128 The Hamdan judgment rested on the fact that the President had determined that Article III trials were not practical, but had not officially made the same determination about courts-martial.129 Under the circumstances of the case, it was tolerably clear that such a judgment was at least implicit in the President’s order. The Court’s demand for specific compliance instead reflected its view of the UCMJ as a precise rule, setting forth ex ante particularized forms of compliance rather than inviting ex post application of a general standard.

Two years later in Medellín v. Texas, a different coalition of Justices similarly read the United States’ agreement to treaties establishing the International Court of Justice as precise and exhaustive of presidential authority.130 Chief Justice Roberts’s opinion for the Court in Medellín, to be sure, rested on a general rule that treaties are presumptively non-self-executing.131 This presumption, however, treats treaty text in a precise, rule-like fashion, rather than as a more open-ended standard. The majority framed the question presented in Medellín in terms of whether “explicit” textual authority existed.132 The dissent, in contrast, would have allowed a more latitudinarian approach to the text.133 Hamdan and Medellín are typically viewed as cases at ideological poles.134 But they share methodological common ground in their reliance on rules over standards vis-à-vis statutory construction.

These contexts are not the only ones where we see cycling in foreign affairs law. In cases concerning the preemptive effects of federal immigration law on state regulation, for example, the Court has also switched between standard-like field preemption135 and narrower, rule-like conflict preemption.136 More generally, along a longer timeline the Court has alternated between exclusive, rule-like and broad, standard-like readings of the President’s foreign affairs powers. In a pair of cases challenging President Roosevelt’s 1933 agreement with the Soviet Union, the Court thus relied on a “powerful presumption in favor of federal executive action” even absent either statutory authority or clear justification in the constitutional text.137 A decade later, in Youngstown itself, a plurality of the Court declined to infer additional presidential authority beyond Article II’s enumeration.138 Subsequently, the Court proved more willing to infer nonstatutory presidential authority to oust state law that impinged on the President’s foreign policy efforts.139 Most recently, the Court in Zivotofsky ex rel. Zivotofsky v. Kerry, determined that it is the “exclusive prerogative of the Executive” to “recognize a foreign state and its territorial bounds,” a prerogative that “resides in the President alone.”140

Again, ideology provides no organizing principle for the observed movement between standards and rules in foreign affairs cases. So-called liberal Justices rely on both rules (Hamdan, Youngstown) and standards (Zivotofsky). Similarly, holdings labeled conservative are just as likely to be grounded on standards (Dames & Moore) as rules (Medellín). What is consistent is not the alignment between one ideological side of the Court with one sort of legal norm, but rather the fluid movement back and forth between the two kinds of norms.

F. The Cycles of Separation-of-Powers Law: A Summary

Rules-standards cycling occurs across a varied separation-of-powers terrain. The doctrinal changes identified here are diverse. Some involve a sudden, conscious shift from rules to standards (or vice versa) that expressly rejects a prior methodological approach. Such about-faces are surprisingly rare, but include Northern Pipeline.141 More commonly the Court creates a carve-out by nesting a rule within a standard or layering a standard on top of a rule. Nondelegation cases, including Schechter, Chadha, and Clinton, thus impose rules that limit the relevance of the broad “intelligible principle” standard,142 while Free Enterprise Fund, a removal case, carves out a rule limiting the applicability of the broad standard announced in Morrison.143 Alternatively, the Court recalibratesby replacing one standard with another, or one rule for another. The move from Humphrey’s Executor to Morrison144 or from Northern Pipeline to Stern145 arguably falls into this category.

We hasten to add two caveats to our descriptive account. First, not all lines of separation-of-powers doctrine oscillate between rules and standards. In contrast to the cycling described above, the Court’s pronouncements in Appointments Clause cases have been quite stable.146 As Justice Kennedy suggested in his concurrence in Public Citizen v. United States Department of Justice,147 this may be in part because the Constitution’s text on appointments provides more guidance than its language does on removal authority. In Public Citizen, Justice Kennedy distinguished cases in which “the power at issue was not explicitly assigned by the text of the Constitution to be within the sole province of the President” from those “where the Constitution by explicit text commits the power at issue to the exclusive control of the President . . . .”148 He emphasized that there was no need to engage in any balancing “[w]here a power has been [textually] committed to a particular [b]ranch”—that “balance already has been struck by the Constitution itself.”149 We recognize this is only a partial answer, in part because the interpretive question of textual commitment itself is also vulnerable to cycling between rules and standards. Still, we do not think that textual specificity is without relevance.

Another domain in which stability prevails without textual specificity is one previewed above: Article III’s application to agency adjudication, which again stands in sharp contrast to the intense cycling found in jurisprudence exploring the Article III question in bankruptcy cases. We suspect that this phenomenon is best explained not by textual specificity but rather by judicial aversion to certain kinds of litigation. Since the first inklings of the modern administrative state, the federal judiciary has resisted efforts to assign itself a large ministerial role over the day-to-day conduct of administrative agency adjudication.150 As Thomas Merrill has explained, the early twentieth-century judiciary’s “fear of contamination” by involving itself in agency administration was conducive to a constrained judicial role in the retail operation of the administrative state.151 Although Merrill documents hesitation on judges’ part throughout the Progressive Era, his story ends with the judiciary ultimately accepting the validity of congressional delegations of adjudicatory power to agencies.152 By 1932, the year of the landmark Crowell v. Benson decision,153 the Court had permitted agency adjudication with Article III review of fact-finding based solely upon the administrative record.154 The Court has not since wavered from this basic position. Recent constitutional challenges to agency adjudication have been rejected under a loose standard that permits the Court to account for a wide variety of variables related to the operation of different kinds of adjudicative mechanisms within the regulatory state.155 A recent warning shot from Justice Thomas, calling for a reconsideration of agency adjudication’s status,156 thus signals a potentially destabilizing willingness, at least on the part of some, to reconsider an unusually fixed element of our separation-of-powers doctrine.

Our second caveat is that the lines of cases analyzed here even when combined with the few that do not exhibit much cycling do not exhaust the institutional landscape. Notwithstanding reports of its demise,157 the political question doctrine—and the paucity of justiciable controversies—mean that many constitutional questions about the design and operation of the federal government remain beyond judicial purview.158 Litigation about the separation of powers thus occurs against a backdrop of institutional change and development that proceeds largely (albeit not entirely) independently of what the courts do.159 Secular trends such as the twentieth-century rise of bureaucratic power do not evince the same cycling dynamic as the case law.160 This Part, in other words, has mapped the law’s trajectory—a path that should not be mistaken for the larger institutional dynamics of the federal government.

Yet the doctrinal cycling between standards and rules—and back again—still poses a puzzle. Whywould rational judges engage in such jurisprudential oscillation? Our aim in what follows is not to explain it as a historical matter, but rather to determine whether cycling might be justified by the foundational dynamics of the separation of powers.

ii. the separation of powers’ foundations: normative pluralism in the thick political surround

This Part returns to first principles to understand better how normative values and institutional forces shape interbranch dynamics, and thus create the background conditions for judicial intervention. The central normative claim we introduce here and develop further in Part III is that given the two background constraints—normative pluralism and institutional heterogeneity—rules-standards cycling may serve as a sensible mechanism for judicial vindication of the separation of powers. This Part introduces these two predicate assumptions.

The first assumption is that the separation of powers promotes a plurality of values, not just a single one. In recent cases, the Court has rooted the separation of powers in ideals of liberty, efficiency, democratic accountability, and the often-elusive rule of law. But the Court (and commentators) generally fails to note that these ideals cannot all be realized simultaneously. They are in irreconcilable tension. Consequently, the separation of powers must enable the dynamic contestation of those values. This basic fact of normative pluralism resonates with the longstanding resistance, starting with Madison, to absolutist solutions in American constitutional law.161 Power, on this view, is never to be wholly concentrated in one government institution nor given over to one type of authority, be it republican, populist, or mandarin. This familiar Madisonian resistance to tyranny (as reflected in the separation of powers) and the corresponding commitment to pluralism (as reflected in the diversification of powers) should be reconceived to reflect not just concern about literal, corporeal tyranny, but also about the tyranny of a single norm.

Whereas normative pluralism can be traced back to the separation of powers’ intellectual origins, the second predicate assumption of our argument is less clearly marked or celebrated.162 Yet as a factual and normative matter, it is of vital importance. We contend that diverse separation-of-powers values are in practice contested and ultimately realized in a multitude of venues populated by a broad array of actors acting within and around the three branches identified in the Constitution. These venues and the actors populating them—including political parties, activists, congressional committee staffers, state and local government officials, civil servants, foreign agents, and members of the general public (including special-interest groups and lobbyists)—constitute the thick political surround.

Though most of these actors go unmentioned in the Constitution (and those few who do are acknowledged only peripherally), they play pivotal roles in advancing or undermining the sundry separation-of-powers values. For this reason, institutional heterogeneity, reflected in the thick political surround, must therefore be accounted for in any serious analysis of the separation of powers. In what follows, we take up the normative pluralism and institutional heterogeneity predicates in turn and then briefly revisit the jurisprudence introduced in Part I to gauge whether and how it reflects judicial sensitivity to these two predicates.

A. Normative Pluralism

The Constitution’s separation of powers is not merely a heuristic for assigning responsibilities and resolving disputes among the competing branches. Nor is it just a reflection of the intrinsic value of the three-branch structure.163 The Constitution’s chief institutions have instrumental justifications. The three branches serve as devices through which a larger, pluralistic normative vision can be channeled and, ultimately, vindicated. The key term here is pluralistic: the federal government’s basic design is intended to simultaneously advance and harmonize diverse and conflicting normative ends. As a correlative, American separation-of-powers thinking since Madison has long registered antipathy to arrangements that concentrate power in one branch or, worse, a single element of a branch.164 It follows a fortiori that the tradition evinces resistance to prioritizing one separation-of-powers value over all others. Instead, Americans have been historically committed to maintaining institutional arrangements that enable normative pluralism to flourish.

This commitment is one that the Court seemingly embraces: the Court’s separation-of-powers opinions are shot through with normative pluralism. Here we emphasize four prominent norms—liberty, efficiency, democratic accountability, and the rule of law—and discuss their centrality to modern separation-of-powers jurisprudence.

1. Liberty

In recent cases, the Court has placed perhaps the greatest weight on the most libertarian of the separation of powers’ aspirations. In 2011, for example, a unanimous Court stated that the separation of powers “protect[s] each branch of government from incursion by the others,” but as importantly “protect[s] the individual as well” from an overreaching, possibly tyrannous State.165 The liberty principle is hammered home in cases where the threat to individual liberty is abundantly clear.166 But even where the connection to individual rights is not immediately obvious, the Justices are quick to remind us they are working to promote our liberty. They do so, for instance, in cases concerning recess appointments,167 administrative agency design,168 line-item vetoes,169 removal,170 and non-Article III bankruptcy court adjudication.171

2. Effective Administration

Beyond liberty, the separation of powers is thought to promote effective government by matching tasks to the comparative advantage of specific government institutions. The Court has credited this goal at some moments, but elsewhere resisted it. To see efficiency’s persisting allure, consider an unlikely parallelism between two leading cases, one involving agency adjudication and the other foreign affairs.172

First, in Commodity Futures Trading Commission v. Schor, the Court permitted agency adjudication of federal regulatory violations and state-law counterclaims on the ground that Congress’s “primary focus was on making effective a specific and limited federal regulatory scheme, not on allocating jurisdiction among federal tribunals.”173 The Schor Court rested its ruling centrally on efficiency concerns such as the value of “prompt, continuous, expert and inexpensive” dispute resolution by the specialized Commission.174 Second, efficiency concerns also emerge in the foreign affairs and national security contexts. Dissenting in Hamdan v. Rumsfeld, Justice Thomas, joined by Justice Scalia, identified “structural advantages attendant to the Executive Branch—namely, the decisiveness, activity, secrecy, and dispatch that flow from the Executive’s unity” as dispositive reasons for assigning power to that branch.175

The majority in Schor and Justice Thomas’s dissent in Hamdan might seem poles apart in subject matter and ideological orientation. Yet both rest upon the logic of administrative efficiency. And both prefer the same institutional settlement: reallocation of adjudicative authority traditionally possessed by Article III courts to bodies lacking federal judges’ accouterments of independence.

It would be misleading, of course, to imply that the Court’s treatment of comparative efficiencies is uniformly positive. After all, Justices Thomas and Scalia were dissenting in Hamdan. Another powerful jurisprudential strand is decidedly wary of efficiency arguments. In Myers v. United States, Justice Brandeis famously inveighed against efficiency justifications in the separation of powers. Although he wrote in dissent, his warning that the purpose of the separation of powers was “not to promote efficiency but to preclude the exercise of arbitrary power”176 has gained approving citation by majorities in a range of other contexts.177 For instance in Chadha, Chief Justice Burger channeled Justice Brandeis, insisting that “[c]onvenience and efficiency are not the primary objectives—or the hallmarks—of democratic government.”178 In Clinton, Justice Stevens assumes the Brandeis-Burger mantle, railing as he does against the line-item veto. Undoubtedly efficient, the line-item veto unacceptably short-circuits the “single, finely wrought and exhaustively considered, procedure” for enacting or revising federal laws.179

This wavering prioritization of efficiency reflects a basic fact about the relation between the values of liberty and efficiency in the separation of powers. As Justice Brandeis’s Myers dissent suggested, those goals need not—and generally do not—align.180 Whether a decision that permits efficient governmental action will be conducive to greater individual liberties depends on the uses to which governmental powers are placed. At times, efficiency and liberty go hand-in-hand. But quite often, they are at odds with one another: efficient government may be less inclusive and deliberative, admitting few opportunities for dissent and contestation. And liberty-prioritizing government may well be slow and cumbersome, if for no other reason than the democratic and juridical safeguards of liberty are time intensive and susceptible to manipulation and foot-dragging.

3. Democratic Accountability

A third normative value routinely ascribed to the separation of powers is democratic accountability. The relation between voters and elected representatives is a complex and contested one.181 The separation of powers is thought to promote one quite specific form of ex post democratic accountability by preserving clear lines of responsibility for distinct policy decisions.182 The clarity of responsibility enabled by crisp institutional separation facilitates voters’ retrospective assignment of liability at the ballot box.183

The Court’s Free Enterprise Fund decision invalidating so-called dual for-cause removal regimes incorporated such ex post democratic accountability directly into the jurisprudence. Chief Justice Roberts’s opinion in that case started from the premise that “[o]ur Constitution was adopted to enable the people to govern themselves, through their elected leaders.”184 The Chief Justice then drew upon Madison’s language in Federalist No. 51 to identify “dependence on the people” as the “primary controul on the government,” and asserted that such dependence is uniquely enabled (and preserved) by presidential control of the bureaucracy.185 Even in this case, democratic accountability did not prove wholly dispositive. The Court recognized the nonabsolutism of the separation of powers and tacitly let stand one layer of for-cause insulation from the President.186 It also declined to opine on the status of civil servants within independent agencies or on the status of ALJs.187 As a result, these effectively tenured civil servants and politically insulated adjudicators retain their independence, at considerable cost to democratic accountability as otherwise preserved through firm presidential direction and discipline.

Judicial invocations of democratic accountability do not have as long a pedigree as the liberty and efficiency strands of the separation of powers. There is also considerable empirical dispute over whether the public will in fact treat a given policy success or failure as a referendum on the policy architect herself.188 And democratic accountability is not costless, so much so that it sometimes seems to butt up against constitutional liberties and what we may call the rule of law.189

4. The Rule of Law

The rule of law is a complex and contested concept. It has both a “thin” formal and a “thick” substantive version.190 In its thinner, formal version, the rule of law requires (among other things) that rules be clear and relatively stable, and particularized determinations should be guided by the general rules.191 Separation-of-powers jurisprudence is alive to this kind of formal rule-of-law concern to the extent that judges limit the power of elected officials, and empower bureaucratic staff who are more likely to maintain stable and predictable policies. Similarly, some of the Court’s moves from standards to rules might be understood as efforts to promote rule-of-law values within a given domain. In the bankruptcy court context, for example, the Northern Pipeline Court’s attempt to regularize the kinds of issues a bankruptcy judge could decide might be understood as an effort to promote stability and predictability within a given doctrinal domain.192

This rule-of-law strand within the separation of powers has recently been recapitulated by Jeremy Waldron with characteristic eloquence. At the heart of the separation of powers, as Waldron conceives it, is a commitment to “articulated governance,” in which the process of democratic rule is “divide[d] conceptually into three main functions . . . .”193 By requiring the State to “slow[] down” its decisional process into “an orderly succession of phases” when making important decisions, Waldron suggests, the separation of powers promotes regularity and stability and enables broad participation in lawmaking.194 Further, he suggests, the distinct functions parsed out into separate institutions by the separation of powers correspond to “concerns about liberty, dignity, and respect that the [thick] rule of law represents.”195 When the Court enforces more rules that seem to turn on functional categories—as when it disallows the legislative veto or line-item veto in favor of “a sort of assembly-line fidelity”196—it might be understood as trying to promote the rule of law as defined by Waldron. But when the Court muddies branch boundaries, as in nondelegation cases or removal cases such as Morrison v. Olson, it undermines the rule of law as Waldron appears to understand it.

Finally, it is worth noting that there is no intrinsic tension between the rule of law and the notion of doctrinal cycling. The rule of law prizes stability and predictability, but it does not require that the law remain static. In Lon Fuller’s canonical formulation, the rule of law is undermined when actors “cannot orient” their action in reliance upon a rule because of “frequent changes” to the substance of the law.197 It is certainly true that some of the cases we have identified mark rather abrupt doctrinal pivots that would have been hard to predict ex ante.198 But we do not think that most of the observed separation-of-powers cycling has been so rapid, or so stochastic, as to undermine officials’ or private actors’ capacity to understand and obey the law. Nor do we believe that most of the abrupt discontinuities across lines of precedent undermine the rule of law. Horizontal coherence across the jurisprudence is not generally considered a prerequisite of the rule of law. As a result, legislatures and judges are free to carve up the regulatory landscape based on qualitative and quantitative distinctions that might otherwise seem arbitrary.199

5. Normative Pluralism and the Risk of Tyranny

Recognizing the inevitable friction between competing values coheres with another familiar touchstone of separation-of-powers thinking: the resistance to “tyranny.” In The Federalist 47, Madison famously glossed tyranny as the “accumulation of all powers, legislative, executive, and judiciary, in the same hands . . . .”200 He explicitly did not reject all intermingling or interaction between the branches. Rather, he identified and rejected the corner solution of an absolutist resolution to the problem of interbranch coordination.201 Against the risk of a singular, even celebrated tyrant, Madison positioned the Congress and the courts.202 And against a populist congressional juggernaut, he situated the President and the courts.203 His larger agenda thus focused on preventing absolute concentrations of State power that would be immune from the back-and-forth of politics via interbranch consideration—a concern ultimately about a static institutional equilibrium no longer responsive to divergent values articulated through the democratic process.204 From Madison’s perspective, at least, therefore, a central theoretical difficulty of the Constitution’s separation of powers was the articulation of a static, textual mechanism that would induce a dynamic, fluid equilibrium in practice.

Consistent with his anti-tyranny orientation, Madison intimated a constitutional theory of capture in the most catholic sense of that term—a concern that any one branch or faction could or would entrench itself in ways that prioritized one normative value to the exclusion of all others.205 The domination by any one particular branch or element—by an interest-group lobby, congressional committee, the civil service, the military, or political parties—equates roughly with the domination by a particular value, whether it be efficiency, democratic accountability, or the rule of law.206 Specifically, domination arises when a value is locked in through an institutional arrangement that denies or silences the articulation of other important values. Therefore, a goal of the separation of powers, on Madison’s view, is to preclude this sort of normative monopolization by promoting the ebb and flow of negotiation and compromise. In this respect, the anti-tyranny value conflicts with any and all of the other values insofar as Madisonian fears of domination lead to a general skepticism of the forceful expression of any and all of those values.

***

In sum, the separation of powers is a design for governance in which a constellation of competing values is promoted and blended through institutional arrangements created by the republican process. These values are likely, if not inescapably, incompatible. As a result, they—like the three branches themselves—stand in a perpetually uneasy relation to one other. This would be of little significance if we could parse out the contested values, assigning specific values to specific doctrinal categories. But value pluralism does not work that way. Rather, a diversity of normative concerns permeates each of the five doctrinal categories mapped in Part I. As a result, the Court cannot respond to normative pluralism by treating each of these five lines of jurisprudence as a normatively distinctive “island,” within which the Justices concentrate on promoting a single normative value. Instead, the Court must grapple with normative pluralism within all of the doctrinal categories. It is this imperative for pluralism within and across doctrinal domains that serves as a motor for cycling in so many of the pockets of separation-of-powers jurisprudence.

Perhaps the coexistence of multiple normative values in each and every doctrinal line is best evidenced in the Court’s treatment of individual challenges to national security policies. As many commentators have observed, discrete challenges implicate liberty, democratic accountability, rule-of-law, and effective administration concerns.207 And courts must endeavor to balance, harmonize, or choose among these oft-competing values. Of course, normative pluralism is not confined by any measure to the foreign affairs and national security contexts. In matters of domestic governance the Court has likewise evinced greater or lesser sensitivity to effective administration in the form of bureaucratic expertise; populist administration, as evidenced by special solicitude for public participation; political accountability, as expressed principally through heightened presidential involvement; and the specialness of public governance when considering delegations to potentially self-dealing private actors.208

Normative conflicts, in short, are endemic within and between doctrinal categories in the separation-of-powers context. Madison’s ambition for the separation of powers was not a synthesis dissolving those inevitable tensions. It was rather a device for their mediation via the constant ebb and flow of politics in which competing normative imperatives meet and blend.209 Many of the jurisprudential threads described in Part I are broadly consistent with this goal. The Court defends against perceived concentrations of tyranny on some occasions and promotes effective administration of the laws at other moments.

B. The Thick Political Surround

Our second foundational predicate of the separation of powers concerns the institutional context in which competing and sometimes conflicting values are reconciled. How, that is, are various normative values advanced and tested against one another? There is, we posit, a complex ecosystem of intrabranch and entirely external actors not traditionally accounted for in the separation-of-powers literature that do a lot of the work pushing and pulling, advancing prized values, and jockeying with one another. Vindication of separation-of-powers values, therefore, can hardly be accounted for by looking exclusively at the constitutional branches qua branches. The battleground is much wider and often subterranean. The combatants are also much more diverse.

This Section catalogs a thick political surround of actors both external and internal to the three branches. We demonstrate that denizens of this ecosystem influence the realization of separation-of-powers values. At times they do so indirectly and often seamlessly as integral subunits of one of the three branches. Alternatively, they exert their influence by more direct and sometimes confrontational means, pressing from the outside on one of the branches or those branches’ subunits.

Accounting for the thick political surround represents a radical departure from standard treatments of structural constitutionalism. Those standard treatments focus, myopically we think, upon the branches as fixed units of analysis to the exclusion of other considerations. And, though most observers are by now quite sensitive to the diversity within Congress, they all too often continue treating the executive as a monolithic whole.210 Such standard treatments produce relatively sharp distinctions between constitutionally specified institutions on the one hand, and partisan dynamics on the other hand, and further validate those insisting on a crisp divide between law and politics.

We question these treatments. In our view, it is often better to decompose political life into more granular institutions (not to mention networks and communities of affinity) acting within, across, and outside the branches. Doing so provides more analytic purchase, while remaining relatively tractable.211 It also recognizes that political and institutional dynamics are crosscutting, rather than acoustically separate. To be sure, this does not mean losing sight of the “branch” entirely. But it does mean—contra the standard operating practice of current constitutional law—that strict adherence to the three-branch paradigm risks obscuring deeper dynamics.

Our recognition of the thick political surround’s relevance to structural constitutionalism redeems important elements of a pluralistic tradition that have long animated American political thought. Separation’s early American theorists recognized the significance of intermediating institutions in the promotion, contestation, and realization of all the normative values at play in our constitutional order. In Federalist No. 44, Madison identified states as intermediating institutions capable of frustrating federal tyranny.212 In doing so, however, he warned that there is “no such intermediate body between the State legislatures and the people interested in watching the conduct of the former,” and thus worried that “violations of the State constitutions are more likely to remain unnoticed and unredressed.”213 In this passage, Madison recognized the important role that civil society could play in realizing the plural values embedded in the separation of powers and seemingly lamented the absence of such a bulwark.

Madison need not have struck such a wistful note. Early chroniclers of the American social order recognized what Madison evidently failed to see, namely that civil society was alive and well, regularly shaping political outcomes. Most famously and perceptively, Alexis de Tocqueville marveled at a fledgling nation of joiners, activists, and social and political gadflies, convening, petitioning, and litigating with a frequency and intensity that jolted the young French aristocrat.214 In our view, the thick strata of private actors that made such a strong impression on de Tocqueville have always shaped constitutional structures and influenced the promotion of constitutional values.215

The thick political surround is, to be sure, big and unwieldy. For ease of presentation, we divide the surround into its internal and external components—and take up each component in turn.

1. The Internal Political Surround

Our account starts with the internal institutional surround—the wide array of individuals, groups, and organizational actors who form part of one of the two political branches. These actors are distinguished by their access to specific channels through which they can advance, elaborate, realize, or obstruct federal law and regulation. While such actors can also operate outside of the branch that formally houses them, they are usefully distinguished from other, truly external actors insofar as their legal standing within a constitutional branch vests them with distinctive, privileged means of influencing branch-level affairs.

a. Intra-Executive Actors

The executive branch contains the largest, most diverse, and perhaps most influential contingent of internal actors. Long treated by legal scholars as unitary,216 the executive branch is in fact highly fragmented.217 Most obviously, it is divided between White House offices and line agencies. Each line agency might have an agenda distinct from the President’s and may well seek to evade presidential control.218 And each agency possesses its own distinct institutional culture and, at times, distinct approach to legal questions. Most basically, agencies can be arrayed along a spectrum from purely independent to purely executive bodies.219 Though all executive agencies must be at least potentially responsive to the White House,220 there is some divergence with respect to each of the agencies’ relationship to various congressional oversight committees, other agencies, and the specific communities it regulates or serves. This diversity is reflected in, among other things, differential treatment of agencies by the courts.221 More germanely, this diversity is also reflected in several lines of cases described in Part I. The current mix of standards and rules in the removal context, for example, enables a heterogeneous array of vertical control arrangements. The complex deference landscape after Chevron and Mead, moreover, explicitly accommodates diverse combinations of institutions and institutional actors to participate in administrative policymaking, while at the same time signaling a preference for those combinations that are most inclusive and procedurally robust.222

Nested within each of these agencies, in addition, are political appointees and career civil servants. The latter can be further grouped into lawyers, economists, engineers, and social workers, all serving specific functions and operating according to distinctive professional norms and commitments.223 The ensuing mélange of intra-executive actors invites the forging of strategic alliances and the sharpening of rivalries both within and across agencies and also with actors in other constitutional branches.

It is beyond the scope of our project here to identify and discuss comprehensively each of these intra-executive actors, affinity groups, and institutions and to explain how their manifold interactions shape branch-level behavior and ultimately the separation of powers. Instead, we provide a quick sketch of some key participants, and do so to illuminate our conception of the thick political surround and to underscore that ecosystem’s pertinence to separation-of-powers jurisprudence.

First, presidentially appointed leaders of agencies, whether officially independent or technically beholden to the White House, with their politically appointed deputies, counsels, and assistants, play a decisive role in federal administration and thus also in constitutional governance. Congress often gives these agency leaders the statutory authority to make final decisions regarding the promulgation of legislative-like rules that carry the force of law. Congress also endows these agency leaders with the discretion to dispose of claims adjudicated within the agencies.224

As a practical matter, agency leaders are well positioned to set agencies’ substantive agendas, to decide how to prioritize competing policies, and to select among targets against which to initiate enforcement proceedings. Though generally relied on to advance the President’s agenda225 and, in so doing, to reinforce political accountability within the administrative arena, these officials’ positions may and do diverge from the White House’s.226 White House officials derisively label such divergence “going native.”227 “Going native” implies that some agency leaders have turned their backs on their appointing President and chosen instead to identify with the career civil servants or other long-term stakeholders, such as congressional patrons or the beneficiaries of agency programs. Yet White House frustration (and presidential accountability) aside, agency leadership autonomy often enhances other normatively desirable values such as effective governance and the rule of law insofar as identification with long-term stakeholders promotes stable and predictable agency policies.228 In this spirit, leading accounts of bureaucratic autonomy underscore the entrepreneurial role played by agency leadership in depoliticizing agencies, thereby enabling dispassionate expert administration to flourish.229

Agency leaders also act outside the branch they inhabit, in effect operating like the external actors that are addressed below. They can do so by allying themselves with members of the media, special interests, and members of Congress.230 Agency leaders have shown great dexterity in the legislative arena, influencing controversial and consequential legislation. For example, the commissioner of the mid-century federal Bureau of Narcotics, Harry Anslinger, played a pivotal role in framing the use of opiates and marijuana as criminal law problems, and then arguing to the public for punitive legislation such as mandatory minimum sentences.231 Anslinger’s example shows how officials can foster a sufficiently broad and engaged political base of their own, allowing them to make direct appeals to the electorate and Congress. They can also collaborate with similarly positioned officials at the state or municipal level to secure policy goals232 or even collaborate with their counterparts in other nations.233 These internal actors’ successes on external stages is aided in no small part by their status as executive officials, which bespeaks authority, competence, and deep (budgetary) pockets.

Second, with or without agency-leadership support, career civil servants play a central role in shaping branch-level action. They have considerable influence over the design, drafting, and administration of agency rules as well as decisions about how to enforce those rules in favor of would-be beneficiaries and against perceived transgressors.234 Civil servants are numerous, heavily relied upon by agency leaders (whose average tenure is approximately two years235), and—most importantly—insulated from politically motivated personnel actions.236 Their independence helps further, enrich, or obstruct the President’s administrative interests.237 Among other things, civil servants can prioritize legalistic values and professional norms—and discount arguments that sound in political expedience and public opinion polls. On this optimistic account, the constructive yet combative influence of civil servants may moderate the partisan political nature of the executive branch, changing the mix of values championed by that branch in the separation-of-powers arena. Bureaucratic insulation creates opportunity for mandarin expertise to infuse American public policy, seemingly in ways that conflict with assurances of democratic accountability, but also in ways that may promote effective administration and the rule of law.238 Indeed, in Harold Bruff’s telling, it is the civil servants who counterbalance the agency leaders and thus function as a “bulwark to the rule of law.”239

This understanding of the potentially rivalrous nature of the civil service-agency leadership relationship seems to be one that the courts appreciate and implicitly endorse. In cases such as Motor Vehicle Manufacturers Ass’n v. State Farm Mutual Automobile Insurance Co.240and Massachusetts v. EPA,241 the Supreme Court has worried about politicized agency decision making and thus elicited supplemental input from career administrative experts. As Jody Freeman and Adrian Vermeule write, the Court in these cases “overr[o]de executive positions that they found untrustworthy, in the sense that executive expertise ha[d] been subordinated to politics.”242 Finally, like agency heads, civil servants also can act externally by forging relationships with members of the media, members of Congress, beneficiary communities, regulated industries, or their counterparts overseas or at the state or local levels.243

The net effect of these intrabranch and cross-branch entanglements can, however, be ambiguous. Sometimes, industrious civil servants save the day; on other occasions they use their legal insulation and mandarin reputations to slack or advance their own parochial agendas. Consider, for example, the pivotal role that elements of the national security bureaucracy have played in impeding releases from the detention facility at Guantánamo Naval Base, effectively derailing President Obama’s plan to close the facility by the end of his tenure in office.244 As one of us has demonstrated empirically, bureaucratic entrepreneurship (via backchanneling, lobbying, and leaking) has fueled congressional opposition to Guantánamo’s closure.245 Although disagreement still abounds about how best to address the Guantánamo facility’s future, cross-branch entrepreneurship by the national security bureaucracy has, at least in this one case, resulted in a poisoned and distorted political atmosphere and perpetuated detentions that civilian and military officials can no longer justify on policy grounds alone.246

Additional actors with more discrete or domain-specific tasks also abound. In addition to military officials,247 there are inspectors general (IGs),248 participants in congressionally commissioned or agency-commissioned advisory groups,249 and a fleet of service contractors and other private actors deputized to advance the State’s aims.250 Each of these actors also influences how the executive branch presents itself to other branches and the public, and thus further complicates how separation-of-powers values are realized. For example, although organizationally subordinate to the civilian leadership in the Pentagon, the military has expertise, its own extensive legal codes, cultural practices, and operational authorities, and no shortage of political soapboxes.251 This de facto independence from presidential control252 betrays any conception of the executive branch as monolithic, or even exclusively under civilian control, and courts extending particular deference on military matters seem to signal a judicial appreciation of a somewhat autonomous (and intrinsically valuable) military infrastructure.253 Inspectors general are principally internal auditors of agencies, identifying and, one hopes, deterring agency wrongdoing particularly with respect to the misuse of funds.254 Both IGs and the military leadership have ready access to Congress and thus have occasion and, no doubt, reason to operate externally, in addition to their intra-executive efforts.255 The accelerating use of (and often overwhelming dependence upon) federal contractors is such that some estimates suggest they are now as numerous as federal civilian workers, with many tasked with highly sensitive, discretionary responsibilities in the formulation and implementation of agency policies.

Notwithstanding their overlapping or substitutable responsibilities, contractors and civil servants are very different beasts. Whereas civil servants are protected by law and custom from adverse employment actions absent cause, contractors generally depend quite literally on agency leaders to renew their contracts, thus ensuring the continuation of their work.256 As a result, they are quite rationally presumed to be much more politically compliant.257 Thus the choice to deploy contractors instead of civil servants is a consequential one, in part because contractors are more likely to advance a political, probably presidentialist, unitary administrative agenda (over one that is more rivalrous and disaggregated, moderated by longer-term, professionalized bureaucratic interests regularly at odds with the interests and commitments of the incumbent administration).258

b. Intra-Congressional Actors

The bicameral Congress also has a roster of internal players subsumed within and across the two chambers. These players warrant consideration in any analysis of the separation of powers because they not only shape important policy outcomes, but also shape those outcomes differently than might be predicted from an analysis of Congress proper. There are, to begin, party-selected House and Senate leaders, who wield considerable power over their institutions and over coalitions that caucus together. Congressional leaders—often serving relatively provincial constituencies, as evidenced by recent Senate Democratic leaders from Nevada and South Dakota—have specific formal powers such as intra-house appointments and access to special briefings and disclosures by the executive branch.259 The congressional leaders also possess informal agenda-setting powers and make committee assignments. They often act as agents of the “party-in-the-government,”260 rather than as agents of the institution they serve. Principally because of their assignment and agenda-setting powers, congressional leaders’ work overlaps with that of congressional committees.261 Ranging from the Senate Foreign Relations Committee to the House Ways and Means Committee, these relatively stable subgroups of senators and House members take the lead in framing and conducting debate—sometimes, as in Chadha,262 in ways the Court finds problematic—directing, monitoring, and funding federal initiatives, investigating wrongdoing,263 and advancing or quashing proposed bills. They also provide influential glosses on enacted legislation via committee reports that some judges use to guide their statutory interpretations.264 Bridging the two chambers, a range of “unorthodox” institutional arrangements such as budget resolutions and the now-infamous reconciliation process have developed in response to the breakdown of the traditional bicameral process.265

Within Congress, a further array of important positions and offices are populated by appointed technocrats rather than the members themselves. The Congressional Budget Office (CBO) and the Government Accountability Office (GAO), for example, are particularly influential congressional entities. Both help define and interpret the annual budget.266 And both sharpen, clarify, and challenge the positions of members of Congress on controversial legislation, such as the Patient Protection and Affordable Care Act.267 Lastly, both directly influence wider political and legal debates. Members of Congress, not to mention executive branch officials (and governors, special-interest groups, and the like), are constrained from making unsubstantiated economic or fiscal claims that the CBO could easily rebut or from taking actions that would invite a stern rebuke from the GAO.268

c. Judicial Actors

Our central focus in this Article is the play of forces within and between the political branches. Yet two of the five lines of cases charted in Part I—namely, those concerning judicial deference to agencies and the constitutionality of non-Article III adjudication—directly involve the federal judiciary. The judicial branch surely experiences internal conflict, due to geographical dispersal, life tenure, and political currents from which judges enjoy no shelter. Nevertheless, the federal judiciary is likely more cloistered from external forces—via rules against ex parte proceedings and professional norms against political engagement. The judiciary is also less subject to internal disruptions, if only because of the clear hierarchy of the federal court system, the general commitment to principles of precedent, and the division between judges and their support staff. (Whereas agency leaders and members of Congress are often outclassed by their seasoned, expert staffs, rarely would we find a similarly inverted dynamic, or even parity, when it comes to federal judges and their still wet-behind-the-ears term clerks.) More generally, the judiciary has a more coherent and stable set of institutional interests that suppress internal, ideological conflict, that prove unwelcoming to the denizens of the political branches, and that thus result in a unified approach more in keeping with the standard treatments of branches as monoliths.269

2. The External Political Surround

The internal ecology of important players within the executive and legislative branches is complemented by a diverse external ecosystem of actors who influence how the separation of powers plays out. These actors are the lineal descendants of Madison’s intermediating institutions (including the states) and de Tocqueville’s civil society.

To begin, we have the public itself acting in its diverse democratic capacities. This democratic public votes, assembles, protests, petitions Congress, speaks, and sues the government. These opportunities for public engagement empower various factions, marginal and median, across a heterogeneous and fractious electorate. The public, or at least those individuals and groupings sophisticated enough to employ the available tools, is also legally empowered to influence the administrative process through requests for agencies to promulgate rules, through its substantive participation in the rulemaking process, through its access to information under the Freedom of Information Act,270 and of course through its ability to bring suits challenging the lawfulness or reasonableness of agency action.271 At times we see the Court seemingly privileging agency actions that have benefitted from public scrutiny and engagement. For example, much to the dismay of unitary executive theorists, Mead establishes a hard-to-rebut presumption that courts grant the less generous Skidmore deference to agency interpretations that have bypassed public notice and comment.272 Mead nudges agency officials to engage in more notice-and-comment rulemaking, since those who do so are far more likely to be rewarded with the more deferential Chevron review.273 Despite this judicial encouragement and solicitude, structural and asymmetric limitations on participation—such as economic or educational barriers to entry—matter for any number of reasons. For our purposes, they matter most insofar as certain voices that we expect to be heard in the separation-of-powers scrum will be unnaturally amplified to the exclusion or muffling of others.

Second, and distinct from the public at large, there are the two main political parties, the influence of which some claim eclipses the formal separation of powers.274 Parties reflect efforts to pool resources and magnify influence among those who hold a common set of views.275 Parties pursue their members’ substantive interests through action in all three branches of government. Partisan influences thus mold the agenda and the output of congressional processes. Partisan incentives shape the jurisdiction and personnel of the federal courts.276 Depending on how representative and inclusive they are, parties can sharpen or obscure the contestation of values and interests in the separation-of-powers arena.

>Third, local and state governments also influence how the separation of powers plays out in practice.277 This is obviously true when federal programs are administered by the states, as is the case with many health, welfare, education, and housing initiatives. In the context of these programs, states often seek to intervene not just at the agency level (where they appeal separately to agency leaders and civil servants), but also with the White House, Congress (where they once again appeal separately to their states’ contingent of elected legislators as well as to the relevant committee chairs), and the courts.278 Indeed, the Court has shown itself to be particularly receptive to states’ challenges to federal executive action (or inaction) involving immigration, the environment, and health care.279 In one environmental case, Massachusetts v. EPA, the Court conferred what it called “special solicitude” on states, finding states to have Article III standing even when similarly situated private parties may not.280 Local influence is further registered when local and state officials, such as big-city mayors, state governors, or attorneys general, ascend the national stage and influence policy by leveraging their standing with the public and their political parties.281

Fourth, recent work by Daniel Abebe and Ashley Deeks, among others, has brought into focus the diverse ways in which a range of friendly and antagonistic foreign actors, including governments and their equivalents, intervene in the separation of powers to influence policy outcomes.282 In the national security context, Deeks has argued that American intelligence agencies choose among different policy tools in part on the basis of their foreign counterparts’ likely willingness to cooperate or protest.283 Foreign intelligence agencies therefore not only create opportunities to circumvent domestic-law constraints—as occurs when allies agree to spy on each other’s domestic populations—but also supply constraints of their own on policies, such as surveillance and detention, with important separation-of-powers ramifications. Abebe, by contrast, points out that high-level American diplomats, when formulating foreign policy, necessarily account for the interests of foreign allies and opponents.284 Diplomatic interactions between sovereigns influence domestic policy agendas, sometimes with constitutional repercussions. Presidential efforts to manage multilateral relationships, for example, can lead to executive orders that courts later interpret to preempt state laws.285 When diplomatic entreaties to the State Department fail, foreign sovereigns can appeal directly to American courts in the form of amicus briefs (which judges seem to cite at a disproportionately high rate)286 or to Congress and, derivatively, the American people.287 As Abebe shows, it is often infeasible to gauge the actual balance of federal, interbranch power without accounting for foreign and trans-national actors.288

C. Normative Pluralism and the Thick Political Surround in Separation-of-Powers Case Law

For the reasons just discussed, normative pluralism and the thick political surround are indeed key background predicates for understanding the separation of powers in practice. But are those predicates relevant to the work of the courts? We conclude this Part by showing how the jurisprudential shifts detailed in Part I suggest a judicial sensitivity not just to normative pluralism (which appears on the surface of opinions and in everyday juridical patter) but also, more subtly, to the thick political surround. This Section thus presents evidence that in each of the five lines of precedent mapped in Part I the Court has grappled—albeit in an inchoate and perhaps unwitting fashion—with the background theoretical predicates we have identified here.

To be very clear at the threshold, our argument in this Section is notthat the Court has conceived the two predicates of the separation of powers in the way we do, or that the Justices, whether as individuals or as a collectivity, have refined some theoretically sophisticated way of accounting for those predicates in the form of workable doctrine. Rather, our more modest claim is that some instances of rules-standards cycling suggest latent awareness of the thick political surround’s potential to generate virtuous or deleterious forms of politics—to the betterment or subversion of the separation of powers. The analytic and normative framework we have so far developed and will further develop in Part III enables us to posit a theory of rules-standards cycling (and a corollary account for assessing where and when such cycling may be most profitably employed).

First, recall the trajectory of the nondelegation doctrine.289 The Court started with an “intelligible principle” norm operating initially as a standard; the doctrine momentarily hardened, pivoting in 1935 in a rule-like direction; then, just as abruptly, a course correction softened the doctrine once again, leaving us with a permissive standard. The new standard was, however, subject to several rule-like carve out exceptions: categorical bans on private delegations, re-delegations to legislators (in the form of legislative vetoes), and assignments of line-item veto powers (over already enacted legislation).

We can understand these oscillations in the following way: The Court initially encouraged widespread institutional experimentation within the burgeoning federal administrative state. The Court hesitated during the “First New Deal” as Congress transferred massive, largely unconstrained power to agencies and private actors alike. Once it became apparent that the administrative domain had grown more orderly, inclusive, and democratically and legally accountable, the Justices felt confident that they could relax the doctrinal strictures, intervening again only surgically to prune away what they took to be the more problematic forms of experimentation.

We therefore think that these doctrinal shifts evince an implicit sensitivity to the thick political surround. When and where the Court balked—again, private delegations, legislative vetoes, and presidential line-item vetoes290—it was with respect to practices that departed in important ways from the ordinary forms of delegations, which typically facilitate exercises of federal power reflecting widespread participation (and broad buy-in) from an inclusive, heterogeneous set of engaged actors. Notwithstanding the Court’s experience with these exceptional and ultimately unacceptable delegations, the Court stayed true to the baseline “intelligible principle” framework, signaling its continued willingness to encourage and endorse exercises of federal power that are the product of a thick and healthily competitive ecosystem of internal and external actors. To be sure, a healthily competitive and inclusive thick political surround does not guarantee that only desirable outcomes emerge from the administrative process. But, in such a context, it is not clear that more rule-like application of the nondelegation doctrine will elicit better results, either on the substantive merits or with respect to the inclusiveness of the administrative process. As a result, the trajectory of the nondelegation doctrine is at the very least congruent with judicial sensitivity to the thick political surround.

Second, in the removal context,291 Humphrey’s Executor can be read as recognizing particular sensitivity to the thick political surround.292 Simply stated and contra Myers, not all top agency officials are in the same position vis-à-vis the President and Congress.293 Certain officials play roles that are functionally and normatively inconsistent with close political supervision. (One such role would be adjudication.) Others occupy positions where such supervision would yield negligible benefits, given the pressures emanating from internal professional and civil-service constituencies, or even be counterproductive in policy terms given the short-term goals of Presidents. (One such realm might be that of monetary policy.) Were it not for judicial flexibility respecting different mechanisms of control, Congress might not have been as willing to experiment with a broad, variegated administrative state tailored to the needs and demands of difficult pressing social and economic dislocations.

The first judicial gesture toward that dynamic in Humphrey’s Executor was rather blunt, but the jurisprudence cycled later toward an even more open-ended standard in Morrison v. Olson.294 Morrison announced a highly textured, fact-dependent standard that permitted institutional differentiation within the executive branch in the form of an independent prosecutor. By validating this element of the internal political surround, the Morrison Court evinced (here quite explicit) sensitivity to the risk of executive misconduct enabled by a concentration of authority within the presidency.295 The Independent Counsel Act, which was challenged in Morrison, had been “a direct byproduct” of the Nixon White House’s thwarting of investigations into the Watergate break-ins.296 An amicus brief filed by the U.S. Senate in the Supreme Court also highlighted the connection between the risk of executive-branch misconduct and the need for an unorthodox structural arrangement.297 Sensitivity to this historical context may thus help explain the softening of the Myers rule into the Morrison standard, which was more capable of accommodating a wider variety of policy concerns.

Morrison seems to remain good law on the scope of the removal power in some contexts, but it has now been circumscribed by the Free Enterprise Fund carve-out for double for-cause removal procedures.298 Can this partial tacking back to rules in Free Enterprise be justified in terms of shifts in the thick political surround? Perhaps the carve-out is justified by the fact that the Court envisions no circumstance under which dual-insulation of agency officials would be warranted. The attenuation of democratic accountability, one of the key normative values, is—at least by the five-Justice majority’s lights—simply too great.299 Perhaps more telling for our purposes is the further carve-out within the carve-out that Chief Justice Roberts recognizes for politically insulated ALJs working in independent agencies. The dual-insulation of ALJs does not seem to raise Article II concerns of the sort that motivated the Free Enterprise Fund Court. This may be because the Court conceived of ALJs as embedded within a healthier political surround (rightly and properly cut off from presidential politics given their judicial responsibilities), or, more cynically, it may be because eliminating ALJs would shift a heap of high-volume, mundane, and even ministerial disputes onto the doorstep of the federal judiciary.300

Third, in the domain of agency interpretative authority,301 the best evidence of judicial sensitivity to normative pluralism in the thick political surround is Mead, which (through a return to the Skidmore standard) acknowledges the diversity of administrative practices and then assigns different normative weights to distinct agency procedures and practices. Mead opened up opportunities for greater federal-court sensitivity to institutional heterogeneity. As a result, it might be praised as allowing Congress a greater menu of design choices from which to choose (responsibly). After all, Congress can be comforted by the fact that courts will more fully supervise the administrative process, encouraging thick, inclusive administrative participation, while discouraging (or at least more aggressively scrutinizing) unilateral decision making that reflects an unwillingness to engage fully with the broad, diverse public and other interested stakeholders.

Indeed, we could drill down further. Just as Mead reflects a greater judicial appreciation of the need to accommodate the increasingly diverse forms of administrative action when it comes to reviewing agencies’ statutory interpretations, cases today are signaling a similar need when it comes to reviewing agencies’ interpretations of their own rules. For decades, courts have reflexively and categorically extended so-called Auer (or Seminole Rock) deference in a rather blunt, one-size-fits-all fashion.302 Calls to abandon Auer deference are today loud and seemingly getting louder.303 Many propose that courts apply something akin to Skidmore deference.304 Motivating these calls is a concern over agencies’ concentration of power—specifically, the power to propose vague rules and the corresponding power to interpret those rules with considerable flexibility and latitude.305 A Skidmore-like approach would give courts greater leave to police those subsequent interpretations (as well as more leverage to encourage agency officials to be more inclusive and solicitous of many opinions when formulating an interpretative decision). Though Auer’s days may well be numbered, at least some commentators recognize that a sea change may not be necessary since “Auer’s ‘domain’ is [already] increasingly limited by a series of important carve-outs—carve-outs that ‘tailor deference to variety,’ just as United States v. Mead limits Chevron.”306 The details of those carve-outs need not detain us here.307 What suffices is the observation that here is yet another context in which the courts seem open to jurisprudential pruning to encourage and promote healthier forms of administrative engagement.

The fourth and fifth examples of cycling evince the least sensitivity to the normative concerns we have identified in this Part.308 In the case of Article III jurisprudence concerning bankruptcy courts, we have shown that the Court has moved from a standard in cases such as Katchen309 to a rule in cases such as Northern Pipeline310 and Stern,311 and then back partway to a standard in Wellness International.312 These fluctuations stand in stark contrast to the relative stability of the Court’s commodious interpretation of Article III in the agency adjudication context.313

We see little explanation of or justification for this cycling in terms of the thick political surround and normative pluralism. The Court’s periodic turn to rules limiting the scope of bankruptcy jurisprudence can be glossed as evidence of its hostility to bankruptcy judges, perhaps due to a perception that those officials cannot be trusted with significant independent authority.314 It is possible that concerns about the capture of the bankruptcy process by parochially minded insiders could justify the Court’s occasional doctrinal shifts toward rules (and indeed, could also reflect a judicial effort to preserve the healthy internal ecosystem of the federal courts). But the Court has never even gestured toward a reason for its suspicion of bankruptcy judges—or why the latter warrant a jaundiced treatment while administrative agencies secure a relatively free pass to adjudicate a host of claims. One possible reason, we surmise, is that the Court thinks very differently about the political surround enveloping the executive branch than it does about the same such surround enveloping the judiciary. It is nevertheless our sense that, if anything, the ubiquity of political actors lurking around administrative adjudicators—and the relative absence of those types of actors in the more austere judicial surround—should lead to the opposite result.

Moreover, bankruptcy law, unlike administrative law, is largely procedural in character. It does not change the metes and bounds of private rights.315 As one of us has argued elsewhere, the scope of bankruptcy courts’ domain can be defined and delimited to the class of cases in which resolution in a federal forum casts no distorting shadow on the private ordering of state-law property and contract rights.316 Lastly, bankruptcy judges comprise a relatively professional and non-ideological caste.317 District court judges have ample tools, if not always sufficient inclination, to oversee their work.318 There is little reason to think that any given aspect of the normative pluralism that the separation of powers aims to promote is imperiled by the institutional heterogeneity of current bankruptcy practice.

This brief analysis suggests that Article III-related separation-of-powers doctrine contains room for improvement, compared against a baseline of preserving normative pluralism and institutional heterogeneity. Given the comparative expertise of the bankruptcy bench, the close supervision available from the federal courts of appeals, and the necessarily sporadic and disjointed attention that the Supreme Court can give to these cases, there is a strong case for treating non-Article III bankruptcy courts the same way we treat non-Article III agency adjudicators: as an occasion for announcing an open-textured standard with ample room to accommodate both new policy considerations and the previously unconsidered policy concerns of different or emerging democratic actors. If anything, the differences between bankruptcy and agency adjudication would suggest that there should be more cycling in the latter context because of the wider diversity of adjudicative venues, substantive rights and interests, and species of legal authority at stake across the panoply of federal agencies.

Finally, we demonstrated in Section I.E that the Court has tacked between rules and standards in its treatment of statutory authorizations for presidential initiatives in the foreign affairs and national security contexts. Whereas the Court has treated statutory text as a narrow authorizing rule in cases like Youngstown,319 Hamdan,320 and Medellín,321 it has read statutes as broad standards in other instances, including cases concerning presidential power to settle claims against foreign nations322 and concerning the executive’s ability to detain citizens in military custody.323

This doctrinal instability in national security and foreign affairs is likewise difficult to explain in terms of normative pluralism and the thick political surround. On the one hand, the effects of the thick political surround in national security and foreign affairs are likely to be especially unstable both temporally and substantively because of the fluid and unpredictable nature of geopolitical conditions.324 Consider, by way of illustration, the shifting pressures on immigration policy created by conflict in Central America, the evolving demands of national security policy given the threat from the Islamic State, and the periodic shocks from global economic changes. Hence, a set of doctrinal responses that are sensitive to changes in the thick political surround and to the corresponding changes in the proper balance of the system’s plural values is likely to evince a considerable amount of variability. On the other hand, there is reason to ask whether judicial pivots in fact correspond in some reasoned fashion to the changing institutional ecosystem in this particular domain. Even if judges have all the necessary information, it still may be difficult for them to determine whether a rule or standard will generate better deliberative processes or substantive outcomes. This question of institutional competence is one to which we return in the Conclusion.

In summary, the Court’s work product in the separation of powers reflects a measure of sensitivity to both normative pluralism and the thick political surround, albeit imperfectly and inconstantly. We doubt that the Court theorizes the separation of powers in the terms we have developed; we doubt too that the Court has analyzed the occasions for rules-standards cycling with any rigor. Indeed, our brief review of the jurisprudence suggests only sporadic, seemingly unwitting, attention to the dynamics our Article highlights. After all, by our lights, some lines of separation-of-powers cycling can certainly be justified as sensible, perhaps intuitive responses to normative pluralism in the context of the thick political surround; other lines, however, seem unjustified, perhaps even a touch lawless.

***

This Part has set forth two basic premises about the separation of powers and traced their palimpsest across the Court’s jurisprudence. By way of conclusion, we reiterate those two foundational principles. First, the separation of powers cannot be reduced to a single normative value. Rather, cross-cutting currents of liberty, administrative efficiency, democratic accountability, and a commitment to the rule of law inform our governmental structure. Each of these normative tugs is evident in several lines of separation-of-powers jurisprudence. Second, threads of public participation, identified by Madison and de Tocqueville, knit together into a thick political surround of interest groups, institutional actors, political factions, and diffuse democratic masses that infuse and surround the branches, and necessarily inform the separation of powers’ practical operation.

Normative pluralism in the thick political surround does not imply that the separation of powers is a free-for-all. The dynamic character of the separation of powers need not mean that any and all institutional arrangements are licit: it is quite clear that some institutional arrangements are out of constitutional bounds. Accordingly, the challenge is to develop a theory that sorts permissible from impermissible institutional settlements, while offering a reasonable explanation of how the Court crafts doctrinal instruments to distinguish these different arrangements in an evolving and complex normative and institutional environment. This is the challenge we now take up in Part III, where we consider justifications for particular species of doctrinal cycling and then suggest possible doctrinal frames for evaluating our institutionally and normatively complex separation of powers.

iii. the logic of rules-standards cycling in separation-of-powers law

This Part demonstrates that, at least as a theoretical matter, rules-standards cycling can emerge from a judicial effort to honor plural separation-of-powers values in the context of a thick (and dynamic) political surround. Like many other regulators, courts may be well advised to employ both rules and standards to elicit a sensible mix of separation-of-powers values within this all-important but ever-changing thick political surround.

We begin by explaining the virtues of standards in separation-of-powers jurisprudence as well as their limitations. We thus identify patterns in which courts start with standards, and then shift to rules (which of course have their own virtues and shortcomings) before cycling back to standards. (The pattern can, to be sure, be reversed, starting with rules, softening into standards, and then hardening once more.) In developing this argument, we draw insights from illuminating discussions of cycling between rules and standards in other legal and social contexts, as well as illustrative examples from the case law discussed in the previous two Parts. Specifically, we identify correspondences and similarities between other observed instances of rules-standards cycling and those we discern in the separation-of-powers context. We conclude by offering some tentative thoughts on how the separation-of-powers doctrine might be organized to manage rule-standards cycling.

Again our claim is not that the mechanisms we identify here explain most instances of jurisprudential cycling. Rather, we seek to demonstrate what a coherent separation-of-powers jurisprudence may or could entail as a matter of first-order normative constitutional theory. If such a jurisprudence is desirable—a point we take up in the Conclusion—we think courts should embrace cycling. In short, our accounts of normative pluralism and the thick political surround combine to provide a single framework for analyzing rules-standards cycling in separation-of-powers jurisprudence. The normative justification for this framework provides a much-needed benchmark against which observed examples of rules-standards cycling can be evaluated and critiqued. It also gives us reason to think more critically about judges’ competence in a separation-of-powers realm complicated and clouded by the hustle and bustle of the thick political surround.

A. The Allure of Standards

Our decision to begin at the standards end of the spectrum is not wholly arbitrary, although, given the circularity of cycling, we could just as easily have played out our argument beginning with rules. The Constitution prescribes some hard-edged rules in the separation-of-powers domain, but it does not fully assign or clearly explain the allocation of many of the political branches’ governing responsibilities and prerogatives.325 Unsurprisingly, litigation tends to arise in the absence of a precise textual settlement.326 Hence, it may be that Justice Kennedy is correct to say in Public Citizen that it is not at all surprising that the Appointments Clause has generated far less reticulated case law than the removal power.327 The resulting tendency of initial litigation, especially at the Supreme Court level, to present difficult cases where there is no precise textual rule provides a compelling reason to begin with standards rather than rules.328

In the separation-of-powers context, difficult cases of first impression—on recess appointments, presidential line-item vetoes, and non-Article III adjudications of pendent state-law counterclaims—are surprisingly common.329 Given the likely posture of most novel constitutional challenges, a judicial inclination to begin by announcing a standard is sensible. The promulgation of rules requires information about what considerations should, and also should not, count in determining whether the law has been violated. But courts, like other decision makers, will often lack this information when a constitutional question first arises.330 Starting with standards not only gives judges an opportunity to find their footing in unfamiliar institutional terrain, but also enables them to invite coordinate branches (and actors inhabiting the thick political surround) to offer responsive clarifications and experimental enhancements before the next separation-of-powers dispute arises. Instances where the Court has imposed a rule upon its first encounter with an institutional practice might well be criticized on this ground alone.331 Hence, to the extent that first-cut judicial interventions limiting, for example, presidential removal powers or presidential line-item vetoes have generated sharp and categorical rules rather than standards, we think that the Court can rightly be criticized for acting without a robust empirical foundation. To be sure, some of those decisions may ultimately be correct, but we think that the Court would have been better served approaching the underlying question more tentatively over time in keeping with a more cautious, common-law-like methodology.

Starting with standards can be justified on grounds beyond this threshold epistemic advantage. To begin with, initially translating the separation of powers into a hard-edged rule rather than a standard might make it difficult to adapt to new and unexpected social or political pressures.332 Excessive reliance on rules rather than standards as an initial matter will likely constrain ongoing democratic governance—prematurely ruling out some potential solutions and perhaps also disabling some elements of the thick political surround, locking in winners and losers, and deterring the losers from regrouping and refining their tactics in ways that could contribute greatly to the separation of powers.

What is more, the American experience with state-building suggests that judicial reliance on rules may also destabilize the constitutional project as a whole.333 Federal responses to the New Deal are a canonical example of virtuous, perhaps even necessary, experimentation in response to dynamic institutional and demographic changes.334 But the New Deal was hardly the first time that such adaptability was called for. In the early nineteenth century, for example, legislators and Presidents struggled to reconcile their constitutional understandings of modest grants of federal power with the practical imperative of creating an effectual network of roads and canals, forming a national bank, and acquiring vast tracts of western lands.335 To be sure, it is always possible to appeal to inchoate notions of constitutional necessity in order to vindicate ultra vires actions.336 But judicial refusal to recognize claims of practical necessity in cases of first impression seems to be a distinct second-best to the ex ante use of capacious and accommodating standards that can more candidly account for the diversity of pressures upon the separation of powers’ plural values and the resulting need for flexibility in governance.337 The case for standards and the experimentation and innovation they enable is particularly strong where we encounter no textual prohibition against the democratic polity’s choice of institutional forms.

Relatedly, standards engage democratic virtues even if no change in circumstances occurs. They allow different political coalitions drawn from the thick political surround, facing diverse social and political dilemmas, to negotiate and jockey among themselves to propose and produce any number of institutional solutions that embody different but quite possibly reasonable permutations of the separation of powers’ normative goals. Standards hence accommodate the normative pluralism represented within democratic contestation better than rigid rules that would lock in certain combatants’ institutional preferences or altogether preempt engagement between and among those would-be combatants in the thick political surround.338

The Court’s jurisprudence applying Article III to agency adjudication339 and its Chevron-Mead jurisprudence,340 in particular, evince such openness to experimentation. In both domains, openness might be analyzed in terms of the evolving thickening political surround. As the internal political surround of intra-agency lawyers, civil servants, and ALJs has become denser, as civil-service protections have taken root, and as democratic forces have demanded an even greater array of interventions from the regulatory state, the case for requiring strict, uniform, and conforming practices within the executive has become weaker. Even if the Court is not consciously responding to those shifts, its relaxation of rules into standards can be justified in those terms.

Finally, standards have a related virtue, which might be termed an anti-Thayerian effect. Famously, James Thayer worried about the emasculating effect of judicial review on legislative and executive incentives to deliberate seriously on the Constitution.341 But standards framed in terms of vague, normatively freighted terminology are not self-applying. They instead require regulated entities to engage in normatively oriented deliberation.342 That is, officials acting in good faith against the backdrop of judge-made standards cannot help but confront the meaning of fraught, contested terms such as democracy, efficiency, and accountability.343 In so doing, they are more likely to articulate publicly how they understand those obligations in a fashion that renders government more transparent and its participants more thoughtful and disciplined.344 This forcing of public deliberation may be especially valuable if courts are reluctant to probe legislative and executive intent directly, as has historically been the case.345 Standards may therefore provide a solution to the arguably pervasive, but difficult to police, problem of “constitutional bad faith.”346

None of this is to say that the Court will inevitably begin with standards rather than rules. As we explore below, cognizance of the thick political surround might also drive the Court to adopt a rule as a threshold matter, as it did in the removal cases beginning with the rule set forth in Myers.347 Such an alternative starting point, however, seems to make sense in only a minority of cases.

B. The Movement from Standards to Rules

Starting with standards does not mean we commit to standards long term. Over time, standards tend to be refined, hardened, or narrowed, and thus begin to transform into rules. We envision at least three dynamic forces potentially at work here, with the caveat that one is more of a theoretical possibility than an observed reality.

First, standards almost invariably harden over time. This is because as both judges and administrators see similar legal challenges recurring with some frequency, there are strong incentives to recall formal and informal precedents,348 develop guidelines, employ “rules of thumb,”349 and rely on “historical gloss,350 if for no other reason than to lower the transaction costs of mundane or repetitive governance.351 When a once-novel problem begins arising regularly, the common-law method of adjudication, with its central reliance on precedent and stare decisis, is likely to nudge in the direction of a rule.352 An example of a standard calcifying into a rule through common-law adjudication can be found in the jurisprudence on judicial deference to agency interpretations. As we explained in Section I.C, the shift from Skidmore to Chevron involved a gradual hardening of an open-textured standard in the course of iterative common-law litigation.353 Even before Chevron, and within the multifactor Skidmore test, the Court had started to delineate a class of cases in which agency interpretations reliably secured a weightier measure of deference.354 As a result, when Chevron was decided it was seen by some knowledgeable commentators as less of a revolution than a clarifying restatement of what everyone had already begun to understand.355 By most accounts Justice Stevens, Chevron’s author, himself never viewed or intended that decision to effect a dramatic shift in the law of agency statutory interpretation.356

Even in the absence of common-law adjudicators, officials tasked with interpreting and applying a standard will likely supplement their initially open-ended directives with “more specific ‘guidelines’ or ‘rules of thumb,’” that reflect their growing body of knowledge and experience.357 These guidelines and “rules of thumb” then filter into judicial opinions in the form of “historical gloss.”358 Hence historical gloss, normally considered a neutral judicial tool for constitutional interpretation, in fact conduces to a drift from separation-of-powers standards to separation-of-powers rules.

>Historical gloss comprises evidence of the “traditional ways of conducting government” used in order to “give meaning” to the Constitution.359 In many cases, these “traditional ways” will be the products of actors in the internal thick political surround. Indeed, it may well be that the prospective reliance on such “traditional ways” in constitutional interpretation creates a marginal incentive for some within the internal thick political surround to develop and publicize practices and adhere to set routines. Courts have long looked to such practice-based evidence in order to give content to vague or ambiguous constitutional norms.360 When courts use such evidence—either to restrict or to allow a challenged governmental course of action—they express the prevailing legal norm a bit more precisely. In short, a norm that was previously open-textured becomes a bit more textured, a bit more nuanced, and a bit more rule-like.

An example of this can be found in the Court’s 2014 decision on recess appointments, NLRB v. Noel Canning, where the majority opinion relied on “historical practice” to permit appointments during intrasession congressional recesses and also to reject a requirement that a vacancy must initially occur during a given recess.361 The Court further invoked historical practice to cast doubt on the President’s power to make recess appointments during recesses of less than ten days’ duration.362 Noel Canning thus simultaneously licensed and restricted presidential discretion. Both elements of the decision, though, added precision to a previously uncertain constitutional text. In this fashion, the Noel Canning decision pressed the law on recess appointments toward precision and away from open-texturedness. Judicial invocation of historical gloss therefore became a mechanism for the movement from standards to rules.

Second, there is a less organic but no less dynamic way in which standards evolve into rules. In some cases, the Court may detect some inhabitants of the thick political surround abusing the Court’s permissive standards in ways that subvert the normative pluralism of the separation of powers. This may be the result of opportunism and disregard for underlying constitutional norms or merely an instance of “self-interest seeking with guile.”363 The Court would therefore step in to limit or proscribe certain exploitative practices or forms of participation by imposing hard-edged rules, tightening a standard, or by creating a rule-like carve out from that still-broad standard’s application. That is, the Court might recognize that its experiment-encouraging standards are subject to abuse in ways that subvert the normative pluralism of the separation of powers.

As intimated above, there are exceptions to the dynamic practice of allowing experimentation and then pruning the most destructive or exploitative forms of such experimentation. We are unlikely to encounter such judicially encouraged experimentation when confronted with a particularly strong first showing of exploitation or abuse (or readily apparent signs of imminent exploitation or abuse). Under such, likely rare, conditions, courts may well begin with a rule, rather than a standard.364 Yet even in these contexts, courts that rush to announce a rule run the risk of preemptively proscribing new and entirely salutary forms of democratic or institutional engagement.

There are several instances in the Court’s separation-of-powers jurisprudence that might be glossed in roughly these standards-to-rules terms, although we caution once more that the Court does not frame its analysis in terms of normative pluralism and the thick political surround. The Court’s resistance to private delegations springs to mind as an obvious instance where concerns about interest-group entrenchment and an absence of healthily competitive institutional dynamics might have motivated the shift from a standard to a rule altogether proscribing that administrative tool.365 The seeming durability of the anti-private-delegation rule, moreover, suggests that the Justices remain categorically skeptical of the potentially corrosive form of politics that private delegations engender.366

Recall too Chadha’s insistence that only the “single, finely wrought and exhaustively considered . . . procedure” of bicameralism and presentment can be used to alter the effect of legal pronouncements.367 Chadha’s expressed logic is in some tension with other, long-standing elements of the Court’s jurisprudence, which render failures of bicameralism immune from judicial scrutiny as a result of the “enrolled bill” doctrine.368 One can gloss the difference between Chadha and the nonjusticiability of bicameralism challenges more generally by postulating that the Court believed that factions in Congress had misused (or stood poised to misuse) the legislative veto. The Court’s decision perhaps thus reflected not a formalist fidelity to Article I, Section 7’s text, but rather a more situated judgment about the operation of Congress’s (dysfunctional) internal political surround,369 as no doubt evidenced by the Court’s clear frustration with the procedural shoddiness and substantive duplicity associated with one-house veto votes.370 Whether the outcome of Chadha should be endorsed or decried, therefore, should turn at least in part on an assessment of the Court’s political judgment.

We stress this less organic path of the hardening of standards, which depends on judicial estimates of contingent institutional dynamics, to underscore that any such abrupt shift to rules comes at the expense of experimentation and innovation by virtuous or even newly arriving members of the thick political surround. This is not to say rule-like interventions are inherently problematic. We readily concede that the thick political surround may need policing. Starting with a standard allows the political dynamics time and space to play out. Disruptions or subversions may never arise. They may be dealt with internally or through the legislative process. But if all else fails, the courts may have to intercede, imposing some rules to discipline a potentially problematic surround. Doing so later in the process allows for the imposition of more surgical rules, prohibiting some targeted activities while leaving collateral forms of political engagement unmolested. We hasten to add that sometimes cycling proves unnecessary. The very threat of judges moving toward a rule may be sufficient to discourage elements of the thick political surround from exploiting a permissive standard that allows considerable pluralistic jockeying. In effect, this supplies a countervailing consideration to the anti-Thayerian argument raised above: not only can rules be justified by their direct consequences, but they are also warranted when deployment of a standard would “fritter[] away resources” in the form of political conflict.371

Third, we briefly note an argument that perhaps ought to have more traction than it does in practice. In theory, standards are preferable to rules on efficiency grounds when the regulated conduct is (and remains) infrequent. There are ex ante savings from using a standard rather than a rule because standards are less costly to formulate as an initial matter: the relevant norm does not need to be distilled into a prospective expression. These ex ante savings will likely outweigh the costs of liquidating the standard seriatim via adjudication, particularly if there are few cases to adjudicate.372 Imagine, for example, if the Constitution insisted only upon “adequate experience” instead of specifying minimum age requirements for those candidates seeking to serve as a member of the House or Senate, or to occupy the office of the President. It makes sense for the Constitution’s drafters to expend considerable effort crafting a specific text to resolve problems that struck them as recurrent governance difficulties. By contrast, the Framers were seemingly less likely to expend effort speculating on hypothetical dilemmas that might arise one far future day. Only once they do arise does it make sense to eschew the original standard in favor of a more predictable rule that is now less costly to apply. Although new constitutional text can be supplied, responsive amendments may be difficult to secure in practice (or imprudent to add).373 Thus, a rule might emerge through judicial innovation.

Very tentatively, we discern some of this kind of adaptive standards-to-rules drift in the removal context, which might be understood as a response to the enlarged administrative state and increased litigiousness over the basic structures and practices of modern administrative governance. The same might be said of the foreign affairs context, where recent case law has clarified the President’s powers to act domestically to further diplomatic ends as America’s hegemonic role on the global stage has developed. Yet, the absence of more examples might also suggest that the Constitution contained an excess of rules, and that institutional and technological change drives us not from standards to rules, but in the other direction—a topic to which we now turn more fulsomely.

C. . . . And Back Again

The various pressures on the way norms are articulated in the separation-of-powers domain do not flow in only one direction. In addition to the dynamics described above, there are also countervailing forces that can catalyze the movement from rules back to standards.374 We highlight three such hydraulic pressures here.

First, a virtue of rules may be that they embody “quite particular compromises,”375 but in the context of normative pluralism, a rule that seemingly locks in one such compromise is not necessarily a positive result. Unlike standards, rules are not well suited to accommodating novel considerations, new developments, or unexpected contingencies. What might today register as an unacceptable intervention by a member of the thick political surround might tomorrow be seen as entirely virtuous due to other sets of changed circumstances.376 As a rule is applied, courts will identify an increasing number of situations in which those applications do not faithfully honor the initial compromise, prompting the courts to soften the hard edges, adding exceptions and permitting new considerations to enter the mix.377 Further, as Carol Rose has observed, “[o]ur law seems to find . . . dramatic losses abhorrent.”378 Such losses need not arise only from exogenous change. In Rose’s account of rules-standards cycling in property law, “ninnies, hard-luck cases, and the occasional scoundrels who take advantage of them” are the culprits who “muck up” rules by presenting cases whose outcomes under those rules the courts find distasteful, or even immoral.379

The thick political surround, we think, is replete with Rose’s characters. Consider, for example, the litigant who brings a claim in a non-Article III forum, and then, upon losing in that forum, invokes a constitutional objection to agency adjudication.380 That litigant’s strategic behavior makes it difficult for the Court to shut the door entirely on non-Article III adjudication of a given strain (even if it were inclined to do so in the first place). But as the memory of the rule-motivating scoundrel fades, the Justices may throttle back the doctrine from a rule to a standard. This responsiveness to “occasional scoundrels” can also be glimpsed in another recent non-Article III adjudication decision, Wellness International.381 Faced with a debtor who repeatedly engaged in “evasive and dilatory tactics,”382 the Supreme Court declined to apply Stern’s “formalistic and unbending rules,” but instead took account of “practical effect[s]” to allow waiver of an Article III objection.383 It is not hard to read the Wellness International Court’s retreat from the Stern rule’s rigor as informed by distaste toward the actions of bad-faith litigants.

Second, the verbal formulations of rules often contain unintentionally ambiguous or vague terms that lend themselves to standard-like treatment. Limited judicial foresight makes a measure of ineffability inevitable whenever a constitutional norm is specified.384 “Even the most specific of rules may be avoided” if the rule-interpreter decides “to engraft an exception to the rule at the moment of its application,”385 especially in a domain characterized by normative pluralism. Even the judges who most embrace the orthodoxies of “formalism” are not immune from this temptation. For example, when the Court announced a firm rule against dual for-cause constraints on the President’s removal authority in Free Enterprise Fund, it took pains to stress that its conclusion did not apply to agency adjudicators serving in independent agencies.386 Although the Free Enterprise Fund Court used a rule to limit Congress’s discretion to some extent, at the same time it fashioned this particular edge of the rule as a standard. Such a partial standard engrafted onto a simultaneously announced rule may imply that the Court is itself hesitant to lock in certain dynamics within the political surround.

Another, slightly different way in which functional instability arises is when statutory interpretation is necessarily embedded within the operation of a separation-of-powers rule, and judges retain large discretion over how to gloss legislative work product. We have suggested that this characterizes both Youngstown and Chevron—two legal regimes that have proved significantly more unstable than their verbal formulation might lead one to expect.387 The result of ambiguous rules coupled with embedded statutory interpretation is predictable enough: their presence makes the respective legal regimes more likely to drift toward standard-like norms.

Third, given the heterogeneity of the thick political surround, it is possible that some of those ostensibly regulated by the rule, or who are handicapped by the rule’s application, will seek to take advantage of the hard-edged character of the rule in order to claim an exemption or otherwise avoid its effects. Efforts to circumvent hard-edged rules will over time provoke what Brannon Denning and Michael Kent call “anti-evasion doctrines,” or “doctrines developed by courts—usually designed as standards, as opposed to rules—that supplement other doctrines (designed as rules) to . . . prevent officials from complying with the form of the previously announced rule, while subverting [its] substance . . . .”388 Although Denning and Kent supply examples from congressional power, federalism, and rights jurisprudences,389 the dynamic they identify can be discerned in the separation-of-powers context too. One famous (albeit unlitigated) example of evasion in the separation-of-powers context is the “Saxbe fix,” which allows a sitting legislator to secure a cabinet position notwithstanding the bright-line rule specified in the text of the Emoluments Clause.390 When regulated entities seek routes around a rule, courts may shift away from hard-edged rules and embrace (or re-embrace) more multifactorial standards. The mere fact of a rule’s exploitation, however, may not be sufficient to justify such a drift. It may be that a hard-edged rule, even when subject to some circumvention, is more manageable and effective in promoting separation-of-powers values than the best available standard. The effect of interest-group circumvention on rules, as a result, depends not only on the extent of such bad-faith behavior, but also on the relative attractiveness of an alternative standard.

This final dynamic, we note in closing, has a more salutary, alternative trajectory. For it is also possible that rules will be relaxed into standards because regulated actors have become sufficiently socialized into a normative disposition that severe and inflexible judicial regulation of the thick political surround is no longer warranted. Whether this more optimistic dynamic is observed, as opposed to its more pessimistic flip side, is—as with all of these dynamics—ultimately a question of empirics.

D. The Motors of Doctrinal Cycling

The preceding Sections have explored how various combinations of the separation of powers’ normative pluralism and its thick political surround can catalyze judicial cycling between rules and standards. By adumbrating these forces, we have strived to illustrate—at least as a theoretical matter—how a rational, good-faith judge could end up moving between rules and standards in ways that seem to transcend ideological or methodological camps. In contrast, that same rational, good-faith judge could not vindicate the same range of normative values under evolving institutional conditions if she merely hewed to a single kind of legal norm, whether rule or standard. Stability in the form of stare decisis would do little to promote the full spectrum of relevant normative values given continuing developments, both good and bad, within the thick political surround. A recurring question surrounding allegations of undue influence by a particular actor or set of actors within the thick political surround is not susceptible to static judicial analysis. Circumstances change. Other actors in the thick political surround become more or less powerful. And other procedures become more or less transparent and inclusive.

Judges looking to what they did in previous cases is the judicial equivalent to what generals are often chided for: that is, fighting the last war. We were, perhaps, all too slow to realize that Vietnam was not World War II. Given growing inequalities in wealth and income, it might behoove judges to treat civil society today very differently from the civil society they encountered at the time the Administrative Procedure Act was enacted in 1946.391 Thus, rather than being an indication of jurisprudential dysfunction, rules-standards cycling in the separation of powers can be a sensible response to a complex judicial task. And, again, we emphasize rules versus standards, and not formalism versus functionalism, to make clear that rational judges can coherently alternate between the two approaches without doing violence to their constitutional theories or normative commitments.

Importantly, the justifications we have developed for cycling in separation-of-powers law do not necessarily or even logically extend to cycling in the statutory interpretation contexts of Chevron392 and Youngstown.393 To the contrary, such oscillations in statutory interpretation contexts impose greater costs, undermining legislatures’ abilities to predict how their interventions will be interpreted and applied.394 Greater stability in the forms and methods of statutory interpretation thus may be independently desirable as a means of lowering the cost of congressional action.395 In the foreign affairs context, moreover, it is our impression that the Court does not stick with a rule or standard for an extended period of time. Rather, it seems to tack rapidly between rules and standards across different issues (e.g., military detention and the death penalty) and even within given issue domains (e.g., immigration).396 Although rapid shifts are observed in other parts of the jurisprudence, the phenomenon attracts less notice because shifts in the modality of statutory interpretation do not have the precedential force, or salience, that changes in first-order constitutional rules possess.397 Cycling in statutory interpretation may therefore result in even greater uncertainty and even higher obstacles to effectual congressional action.

Finally, we stress that our account is not intended to be comprehensive in the sense of exhausting all possible engines of cycling. Indeed, we acknowledge that cycling between rules and standards can emerge for other, less salutary reasons. For example, any decision-making procedure involving more than two participants making choices among more than two options can produce instability, with any choice being vulnerable to defeat by another.398 As one of us has explored elsewhere, the possibility of decisional instability explains many structural constitutional rules.399 Cycling might also occur if the ideological composition of the Court shifted over time in ways that led it to seesaw as a result. Although we have explained why ideological shifts in the Court do not appear to be the root explanation for much of the rules-standards cycling we identify, we concede that it is possible that some instances of instability in the separation-of-powers case law, just like some instances of instability in other constitutional and statutory domains, simply reflect changes in the Court’s personnel.

E. Organizing Frameworks for Judicial Review

We conclude this Part by sketching, albeit at a relatively high level of abstraction, three possible judicial strategies for enforcing the separation of powers in light of the realities of normative pluralism and institutional heterogeneity. The three alternatives developed below are intended as tentative suggestions of how courts might bring order—and make plain that they are bringing order—to the normative and empirical complexity immanent in the separation of powers. Each strategy picks out a different institutional dynamic as a potential focus of judicial attention. Choosing among these alternative strategies demands yet further normative deliberation. We bracket for now the question how to make this election (and indeed, whether to make the election at all), and instead simply illuminate a range of strategic options open to rational, good-faith judges in the separation-of-powers context.

Our analysis suggests that many internal and external actors influence the ebb and flow of interbranch relations, with any number of attendant effects on the separation of powers’ plural values. Again, the separation of powers includes many diverse practices and patterns of institutional behavior. As a normativematter, however, not all these strains of influence merit the same measure of respect or approbation. A normative analysis of the separation of powers—and, more specifically, a judicial treatment of that structure—must recognize that some practices, persons, and outcomes emerging from the thick political surround are more legitimate than others. It must also develop a typology for organizing and analyzing different kinds of inter- and intrabranch dynamics.

Consider first the possibility that such a normative analysis should focus upon which practices inform the separation of powers. That is, evaluating the thick political surround would involve homing in on the form or modality of engagement between one or more of the players we have identified and the branch in question. It also requires asking whether this kind of engagement is legally authorized or otherwise consistent with the norms of democratic intercourse and the plurality of values underlying the separation of powers.

To flesh out what this means in practice, it is useful to consider legality as a criterion of legitimate institutional engagement. On the one hand, acting pursuant to express constitutional or statutory authorization or pursuant to express rights or prerogatives ought readily to satisfy this requirement. International treaty negotiations or commitments, governors’ speeches criticizing federal policy, public comments on pending rules, and agency audits spearheaded by inspectors general would all rank as acceptable practices. So too would those practices conventionally understood as within the legal rights of institutions and individuals, such as foreign governments’ submission of amicus briefs, congressional committees’ guiding of agency spending, and public lobbying of Congress, agencies, or the White House. Conversely, acting in contravention of constitutional or statutory law, or in some instances acting contrary to longstanding practice, would raise red flags about the separation-of-powers merits of that practice. Ignoble practices might include, among other things, lobbying federal judges or agency adjudicators, efforts by mid-level military officers to circumvent the chain of command, and agency actions taken unilaterally by either agency leaders or civil servants, excluding the views of the other (and excluding members of the public who are, of course, legally authorized to file comments, etc.).

A second potential lens for analysis would focus on whether the actor properly belongs in the thick political surround in the first instance. Do foreign lobbyists merit consideration? Should the nationality of the relevant actor matter? Perhaps influence by the Bank of England is appropriate, while entreaties from Iran’s Guardianship Council are not? Likewise, how should government contractors be conceptualized: is their influence tolerable when performing ministerial or clerical duties but not when carrying out sensitive, discretion-laden policymaking responsibilities? We might even be so bold as to quibble with congressional committees. Are those that are truly representative of the plenary houses legitimate, meriting due respect within the thick political surround, whereas those committees that are unrepresentative of the bodies as a whole—such as the agricultural committees stacked with representatives from the Plains States—somehow suspect? These are of course difficult questions—empirically, politically, and often diplomatically—but ones we might need to consider once acknowledging the thick political surround and assessing its effect on a well-functioning separation of powers.

A final approach to assessing the thick political surround would look to outputs, rather than inputs, and would measure these outputs according to some criterion of desirable results. For example, this strategy would require asking whether an extra push by Group of 7 central bankers or North Atlantic Treaty Organization generals properlyemboldens the executive branch as it battles Congress vis-à-vis primacy in matters of American diplomatic, defense, and international economic policy. The approach would also obligate courts to focus less on inquiring whether the composition of congressional committees is sufficiently diverse and more on whether the funneling of legislative work through said committees generates a desirable level of congressional constraint on the executive.400 And finally, it would require asking whether politically compliant contractors, who sideline independent civil servants, should be welcomed as agents of cost-saving, or condemned for consolidating administrative power in ways that make the executive branch problematically forceful (and unnaturally unitary).

Conclusion

The central aim of this Article has been to isolate and analyze an element of separation-of-powers jurisprudence that to date has been ignored or maligned. Across a wide variety of doctrinal contexts, the Court cycles between rules and standards, and back again. This cycling cuts across and blends the categories of formalism and functionalism. As a result, our account suggests that the canonical formalist/functionalist dichotomies generally used to evaluate separation-of-powers jurisprudence have been systematically obscuring a more complicated, more dynamic, and more interesting picture.

To better understand the potential justifications for such doctrinal cycling, we have returned to the first principles of our separation of powers. We have identified two predicate facts about the foundation of our constitutional design—normative pluralism and the thick political surround. Together, these predicates create fruitful conditions for doctrinal cycling between rules and standards. They do so by encouraging robust, inclusive political engagement while disciplining practices and persons deemed exploitative or threatening of a well-functioning separation of powers. We have thus offered here a theoretical framework for understanding and evaluating the normative and institutional pressures that shape the separation of powers. We have also provided links between that framework and the particular doctrinal instances of rules-standards cycling. Our framework not only illuminates the predicate institutional and normative conditions in which the separation of powers unfolds, but also charts the specific mechanisms that might connect normative pluralism and the thick political surround on the one hand to rules-standards cycling on the other. In concluding, we have offered a rough, tentative sketch of the kinds of doctrinal frames that rational, good-faith judges might adopt in a separation-of-powers jurisprudence. This sketch provides a sense of how judicial engagement with normative pluralism in the thick political surround might be conceptualized in a more systematic way.

Choosing among these judicial strategies—and indeed determining whether any one is attractive—requires a prior judgment about the institutional competence of federal courts to make the kind of evaluations that our framework suggests. To make those judgments, we believe courts must be sensitive to the manifold ways in which separation-of-powers goals are implicated in a policy domain. Judges must be cognizant of both the risks and opportunities presented by the thick political surround, ranging from the prospect of institutional capture by interest groups to the possibility that pressure from internal bureaucratic actors can generate salutary democratic accountability. Courts must also remain poised to revise previous judgments, as the elements of the thick political surround respond to prior rules or standards, or as new policy exigencies impinge upon, or even compromise, existing institutional arrangements.

Additional complications arise because federal courts do not stand, at least not fully, outside the thick political surround.401 As Eric Posner and Adrian Vermeule have observed, it is a mistake to diagnose a public-law problem by drawing “upon the political science literature to offer deeply pessimistic accounts of the ambitious, partisan, or self-interested motives of relevant actors,” but then to proceed by assuming that judges can somehow reach “public-spirited solutions” free of the pernicious forces that otherwise shape institutional behavior.402 Staking out a robust role for the judiciary in separation-of-powers debates hence demands an extended defense of the courts’ ability to play the role of a neutral arbiter403 by successfully navigating ideological and institutional pressures of their own. That position also requires a comparative judgment about when judicial supervision will be superior to the arrangements negotiated by the political branches themselves.404

Whether the Court is capable of reaching informed judgments about these institutional dynamics, whether its deficiencies as a group of law office historians, economists, political scientists, and sociologists can be remedied by amicus briefing, and whether ideological preferences will swamp rigorous, principled evaluation are all difficult questions. Even more challenging are questions about the dynamic effect of judicial intervention on the separation of powers: Will such intervention elicit more desirable forms of behavior from the thick political surround, and the branches they act upon, as various actors compete to elicit judicial endorsement by careful, deliberative, and open behavior? Or alternatively, will judicial enforcement of the separation of powers encourage narrowly focused interest groups to invoke that concept in judicial fora for parochial or partisan gain?405 If courts are not able to disentangle sincere from self-interested invocations of the Constitution (or if that distinction is in practice muddled and hard to draw because of pervasive mixed motivations), then judges’ interventions might have undesirable effects.

Reasonable people, we think, can disagree about the answers to these questions. This Article has not aimed to resolve finally those enduring puzzles, but is instead designed to show that previous scholarship has not gone far in identifying the terms on which debate about the judicial role enforcing the separation of powers must proceed. Indeed, having identified the salience of the thick political surround and normative pluralism, we see no turning back to the stale, over-determined formalism/functionalism binary. More importantly, we discern no basis for fixating on the three constitutional branches simply because it is too difficult to navigate the thick political surround.

With those cautionary disclaimers in mind, we hope that this Article broadens our understanding of a much-maligned domain of constitutional jurisprudence. We furthermore hope it has clarified what considerations are needed for the Court to advance successfully the competing and conflicting separation-of-powers values in the fluid, dynamic, and complex context of our federal government’s thick political surround. Having clarified the foundational grounds of analysis, we anticipate that debate about the judicial enforcement of the separation of powers can proceed on less fallible and more clear-eyed terms.