An Administrative Jurisprudence: The Rule of Law in the Administrative State

An Administrative Jurisprudence: The Rule of Law in the Administrative State

This Essay offers a specification of the rule of law’s demands of administrative law and government inspired by Professor Peter L. Strauss’s scholarship. It identifies five principles—authorization, notice, justification, coherence, and procedural fairness—which provide a framework for an account of the rule of law’s demands of administrative governance. Together these principles have intriguing results for the eval­uation of administrative law. On the one hand, they reveal rule-of-law foundations for some contested positions, such as a restrictive view of the President’s power to direct subordinate officials and giving weight to an agency’s determination of the scope of its own authority. On the other hand, these rule-of-law principles expose some long-established practices as having troublesome foundations, such as the settled doctrine that agencies need not justify their choice of policymaking form. Consideration of these principles in the context of administrative law and government ultimately shows—like so much of Professor Strauss’s work—the many ways in which government under law ultimately depends on officials taking the rule of law as their highest-order commitment.

INTRODUCTION

  1. THE DIMENSIONS OF THE RULE OF LAW
  2. RULE-OF-LAW PRINCIPLES FOR ADMINISTRATIVE GOVERNANCE
    1. Authorization
      1. Decisional Allocation
      2. Scope of Authority
      3. Conclusion
    2. Notice
      1. The Locus of Notice Demands
      2. Implications for “Choice” Among Policymaking Forms
      3. Obligation to Issue Guidance
      4. Conclusion
    3. Justification
    4. Coherence
      1. Locus of the Demand for Coherence
      2. The Agency’s Duty of Systemic Coherence
    5. Procedural Fairness
    6. Summary

CONCLUSION

Introduction

The rule of law, like democracy, is one of our most basic political commitments. 1 See Jeremy Waldron, The Rule of Law and the Importance of Procedure [hereinafter Waldron, Importance of Procedure], in Getting to the Rule of Law 3, 3 (James E. Fleming ed., 2011) [hereinafter Getting to the Rule of Law] (noting rule of law’s place among “constellation of ideals that dominate our political morality”). While the rule of law is an old ideal, 2 See, e.g., Judith N. Shklar, Political Theory and the Rule of Law 1, 2–4, in The Rule of Law: Ideal or Ideology (Allan C. Hutchinson & Patrick Monahan eds., 1987) (describing Aristotle’s account of rule of law). interest in it has gained renewed energy in recent years in part because it provides a basis for evaluating a wide variety of contemporary institutional arrange­ments. 3 See Richard H. Fallon, Jr., “The Rule of Law” as a Concept in Constitutional Discourse, 97 Colum. L. Rev. 1, 43 (1997) (noting a characteristic purpose is to serve as a basis for evaluating practices—and their distance from ideal). Some scholars have specified the rule of law’s requirements for adjudicative procedure. 4 See, e.g., Waldron, Importance of Procedure, supra note 1, at 18 (providing account of procedure as rule-of-law value). Others have looked to its complexion in constitutional discourse. 5 See, e.g., Fallon, supra note 3, at 24–36 (examining role of invocations of rule of law in constitutional debates). Some have specified the rule-of-law values for new sets of institutions, such as global administrative institutions, 6 See, e.g., Benedict Kingsbury, The Concept of ‘Law’ in Global Administrative Law, 20 Eur. J. Int’l L. 23, 32–33 (2009) (providing account of rule of law for global administrative law). or new models of government action. 7 See, e.g., Leighton McDonald, The Rule of Law in the ‘New Regulatory State’, 33 Common L. World Rev. 197, 215–21 (2004) (evaluating rule of law in light of new governance techniques). Still others have explored its role in fostering legality in conflict-torn societies. 8 See, e.g., Ganesh Sitaraman, The Counterinsurgent’s Constitution: Law in the Age of Small Wars 183–222 (2013) (examining rule of law’s meaning for societies emerging from conflict).

In light of the scope of lawmaking by administrative institutions—our form of government is, importantly, administrative government 9 See, e.g., Congress: Structure and Policy 403 (Matthew D. McCubbins & Terry Sullivan eds., 1987) (“In this century, the nexus of policy making has largely shifted from the constitutionally designated branches of government to the bureaucracy . . . .”); Peter L. Strauss, The Place of Agencies in Government: Separation of Powers and the Fourth Branch, 84 Colum. L. Rev. 573, 575, 580–82 (1984) [hereinafter Strauss, The Place of Agencies] (describing elements of American administrative government). —the rule of law’s demands of administrative government is a critical area of inquiry. When scholars have addressed how the rule of law applies to the admin­istrative state, the conversation has most often taken two forms. Some scholars have proceeded inductively. They have sought to induce a set of legality principles that structure particular administrative practices or institutions. 10 See, e.g., Jerry L. Mashaw, Bureaucratic Justice: Managing Social Security and Disability Claims 213–27 (1983) (providing account of internal legality of Social Security Agency); Nestor M. Davidson & Ethan J. Leib, Regleprudence—at OIRA and Beyond, 103 Geo. L.J. 259, 281–304 (2015) (providing case study of use of precedent and internal law in Office of Information and Regulatory Affairs’s (OIRA) review of agency rulemaking); Trevor W. Morrison, Stare Decisis in the Office of Legal Counsel, 110 Colum. L. Rev. 1448, 1492–511 (2010) (providing account of use of precedent within Office of Legal Counsel (OLC)). Following this course, the account of legality constraints operating within the administrative state is built through case studies that distill the operative constraints for particular administrative actors. At the other extreme, work has fastened on the distance between administrative institutions and some historical ideas of the rule of law. 11 See, e.g., Philip Hamburger, Is Administrative Law Lawful? 5–7, 12–13 (2014) (challenging legality of administrative state based on historical ideals of rule of law); Friedrich Hayek, The Road to Serfdom 72–87 (1994) (arguing government regulation draws society away from rule of law as traditionally understood). These scholars conceive of the rule of law in terms of distinctive virtues of judicial or legislative decision and find that administrative institutions pose a problem. But neither the more inductive studies nor the historically rooted efforts devote much consideration to contemporary administrative law—and to the ways in which it does or could provide a specification of the rule of law’s demands. This leaves some basic questions unanswered: In what ways do current administrative law doctrines provide a specification of the rule of law’s requirements of administrative government? Is there an account of our administrative law that pays particular heed to the values of the rule of law?

To venture answers to these questions, one could not imagine a stronger guide than Professor Peter L. Strauss. His insightful scholarship on administrative government and law is comprehensive in scope and, at every turn, deeply engaged with the values that undergird our commit­ments to law. This Essay develops an account of the rule of law’s demands of administrative government by relying on Professor Strauss’s work as a foundation for understanding how administrative law corresponds—or should correspond—to the rule of law’s most basic principles.

This Essay approaches this task in two steps. The first step, under­taken in Part I, is to identify rule-of-law principles that are most relevant to the administrative state. The focus on administrative institutions allows for some principles, such as those pertaining to criminal prohibitions, to be left aside. Other principles, such as the scope of authorization, procedural fairness, and notice, are recurrent issues of concern for administrative institutions and therefore merit greater emphasis. The result of this analysis is a focus on five dimensions of the rule of law: (1) authorization, (2) notice, (3) justification, (4) coherence, and (5) procedural fairness. The second step, in Part II, develops, with the help of Professor Strauss’s work, an account of how administrative law embodies—or should reflect—these principles. What emerges is the outline of an administrative jurisprudence.

The basic elements of that jurisprudence can be summarized as follows. The first dimension, authorization, is the very basic demand that government action have a valid legal source. A starting premise of this administrative jurisprudence is that authorization is personal to the officeholder, rather than an impersonal vesting of power in the govern­ment as a whole. The idea is that each officer vested with legal authority has responsibility to reach an independent judgment about what the statute requires, a judgment not to be supplanted by that of superiors. This responsibility precludes the specter of a bureaucracy dictated by role-based compliance up the chain of command in which only the highest-level official bears genuine accountability. In other words, when role is defined in terms of independent judgment, role-based compliance privileges an official’s independent evaluation over political loyalty or bureaucratic order. If individual judgment by agency officials is a structural premise for the rule of law within the administrative state, then it makes sense that the other aspects of the rule of law will apply to how agency officials exercise their discretion. That holds with regard to the principles of notice, justification, and coherence.

The second dimension, the cluster of values relating to notice—principles of publicity, clarity, prospectivity, and stability prominent in Professor Lon Fuller’s account of law’s virtues 12 See Lon L. Fuller, The Morality of Law 46–90 (rev. ed. 1969) (arguing these values are fundamental to law). —has been thought to pose particular problems for the regulatory state. Regulatory statutes—the statutes that create and delegate to agencies authority to make law—are famously broad and vague. But because notice values seek to protect law’s capacity to guide action, they should apply to the sources of law that directly bind private parties. In our government, that means a critical battleground for these principles is the law issued by agencies because that regulatory law, rather than the legislation authorizing the agencies to act, bears the weight of imposing obligations on private persons. Once these notice values are seen as applicable to the law agencies issue, there are grounds to ask how well agencies are meeting these demands. In this vein, if rules generally fare better with regard to notice values than adjudication, then there are reasons to require agencies to, at a minimum, justify opting for adjudication instead of rulemaking. This focus also suggests agencies have an obligation, when rulemaking is not practicable, to issue some kind of guidance document reflecting the agency’s best view of the statute’s requirements.

The third and fourth dimensions, justification and coherence, also bear on how agencies exercise their judgment. The coherence of law—law constituting an integral system—is not built into the American style of statute-making. As a result, the rule-of-law burden of creating coherence falls upon agencies and courts; they have a responsibility to implement the statutes they administer in ways that promote the coher­ence of law, including implementing their statutes in a way that is consistent with constitutional considerations and other background values. This suggests a rule-of-law basis for understanding agency statu­tory interpretation as involving what scholars have called administrative constitutionalism. 13 See Gillian E. Metzger, Ordinary Administrative Law as Constitutional Common Law, 110 Colum. L. Rev. 479, 522 (2010) [hereinafter Metzger, Ordinary Administrative Law] (providing account of role of constitutional interpretation in agency statutory implementation). The same holds for the rule-of-law requirement of reasoned justification. If agencies are the primary implementers of statutory law, then law’s demand for justification depends importantly upon their practices. This provides a rule-of-law basis for administrative law’s high demands for reasoned justification.

The final dimension, procedural fairness, makes particular demands of agency adjudicators. While the rule of law does not dictate a particular structure of government—whether parliamentary or separated powers—it does insist on virtues of procedural fairness for adjudicators. The implication is a rule-of-law grounding for insulating adjudicators from political oversight.

Before proceeding, two qualifications are in order. First, a consistent strain of Professor Strauss’s teaching is the importance of contextual understanding of events and legal events in particular. Context comes in, and is deeply informative, at every turn. Accordingly, looking to Professor Strauss’s work with an eye toward specifying a set of rule-of-law principles for administrative governance rests in some tension with his consistent emphasis on contextual understanding, including in case studies. Second, having identified this project, but not one that in so many terms Professor Strauss has directly invited, it should be clear that any of its shortcomings reflect on the present author. Indeed, in places, the Essay highlights the rule-of-law foundations of propositions Professor Strauss defends; at other times it builds upon his ideas to define principles beyond those he has embraced. The hope still is that stepping back in this way will provide a wide lens both for appreciating aspects of Professor Strauss’s contributions, and, at the same time, making some progress toward articulating the broad outlines of the rule-of-law demands of our administrative government.

I. The Dimensions of the Rule of Law

This Part provides a brief account of the ideal of the rule of law and its underlying purposes, and then turns to describe five dimensions of the rule of law particularly salient for assessing administrative governance.

The rule of law retains a place at the center of our political morality; it is an ideal, like democracy, that sits among a small cluster of our most basic commitments. 14 Joseph Raz, The Rule of Law and Its Virtue, in Joseph Raz, The Authority of Law: Essays on Law and Morality 210–11 (2d ed. 2009) (noting rule of law’s status as fundamental commitment); Fallon, supra note 3, at 3 (noting centrality of rule of law to our political traditions); Jeremy Waldron, The Concept and the Rule of Law, 43 Ga. L. Rev. 1, 3 (2008) [hereinafter Waldron, The Concept and the Rule of Law] (linking rule of law and democracy as basic political commitments). But in order to identify the elements of this ideal, one needs to have an account of what purposes the rule of law serves. 15 See Fallon, supra note 3, at 7 (noting efforts to specify meaning of rule of law begin with identifying values it serves). In other words, we need to know what the rule of law is for—what values it protects—before we can identify its core principles. 16 Martin Krygier, Four Puzzles About the Rule of Law, in Getting to the Rule of Law, supra note 1, at 64, 67 (arguing proper place to start is with question of what rule of law is for prior to evaluating its elements).

While the purposes of the ideal are contested, it is possible to identify several underlying values common to most accounts of the rule of law. First, perhaps the most basic, is the idea of constraint, which applies to officials as well as citizens. In this regard, the rule of law is frequently identified with decisionmaking confined by some source other than personal preference, ideology, or a personal sense of justice. 17 Ronald A. Cass, The Rule of Law in America 17 (2001) (noting importance of external constraint); Waldron, The Concept and the Rule of Law, supra note 14, at 6 (noting common feature of rule of law is that government exercises power “within a constraining framework of public norms, rather than on the basis of their own preferences, their own ideology, or their own individual sense of right and wrong”). This con­straint protects against arbitrary decisionmaking, which stands in opposition to law. 18 See, e.g., Fallon, supra note 3, at 8 (identifying protection against official arbitrariness as purpose of rule of law); Krygier, supra note 16, at 76–81 (identifying avoiding arbitrariness as chief value of rule of law). Second, law aims to allow people to plan with some measure of confidence in their capacity to know the legal consequences of their actions. 19 Fallon, supra note 3, at 8 (noting this purpose). Thus, law should give private parties adequate notice and be of a form that they can make sense of so that they can conform their conduct to its requirements. 20 Waldron, Importance of Procedure, supra note 1, at 18. Third, law should provide a mechanism for resolving disputes that is fair. 21 Id. at 6 (specifying procedures necessary for rule of law). Even when law is accessible and clear, and even when decisionmaking is suitably constrained, there is independent value in resolving disputes in a manner that is procedurally fair.

Contemporary accounts of the ideal identify elements or principles that embody or carry forward these rule-of-law purposes. 22 Fallon, supra note 3, at 8 (noting modern theories defend elements of rule of law from account of its purposes). In this regard, rule-of-law theories have a tendency toward lists of elements. Professor Fuller, the fount of much modern thought on the rule of law, identifies eight principles as critical to law: (1) generality, (2) publicity, (3) pro­spectivity, (4) clarity, (5) consistency, (6) stability, (7) capacity to be per­formed, and (8) compliance by officials. 23 See Fuller, supra note 12, at 46–90 (defending these elements of virtues of law). Professor Joseph Raz offers an account with a more institutional complexion. In addition to the virtues of stability, openness, and clarity as virtues of law and lawmaking, which have some cognates within Professor Fuller’s principles, Professor Raz also isolates (1) the independence and impartiality of the judiciary, (2) acces­sibility of courts, and (3) confined discretion of crime-preventing authorities. 24 See Raz, supra note 14, at 214–18 (noting these values). For an account of Albert Venn Dicey’s conception of the rule of law as a formal and procedural view, see Paul P. Craig, Formal and Substantive Conceptions of the Rule of Law: An Analytic Framework, 1997 Pub. L. 467, 470–74 (distinguishing substantive and procedural conceptions). In addition to, or instead of these values, others have empha­sized that law be (1) authorized, 25 See, e.g., Jeffrey Jowell, The Rule of Law and Its Underlying Values, in The Changing Constitution 3, 17–18 (Jeffrey Jowell & Dawn Oliver eds., 7th ed. 2011) (ebook) (noting rule of law requires public officials act within powers conferred upon them). (2) coherent or part of a system, 26 See, e.g., Waldron, The Concept and the Rule of Law, supra note 14, at 32–36 (noting requirement of systematicity or coherence). (3) accompanied by justification, 27 See, e.g., Waldron, Importance of Procedure, supra note 1, at 6 (arguing rule of law requires action “on the basis of evidence” and “right to make legal argument about the bearing of the evidence”). and (4) procedurally fair. 28 See generally, e.g., Waldron, The Concept and the Rule of Law, supra note 14 (“[T]he Rule of Law is violated when due attention is not paid to . . . procedural matters or when the institutions that are supposed to embody these procedures are undermined or interfered with.”).

Focusing on the rule of law’s requirements of administrative insti­tutions, some principles can be left to the side and some have greater importance. 29 See Fallon, supra note 3, at 6 (noting different values are presumptively primary under different conditions). Those elements that pertain to criminal sanctions and processes have less relevance, at least if we focus on those aspects of administrative governance that are not involved in criminal justice. Likewise, other general virtues, such as compliance with the law, do not distinctively apply to administrative institutions. The conditions under which administrative bodies operate make other elements of the rule of law more central. The fact that agencies only have authority that has been delegated to them suggests the critical importance of the principle that official action be authorized. The fact that many statutes that dele­gate regulatory authority grant broad discretion to officials suggests the importance of notice, coherence, and justification. If agency officials are creating law under broad standards, they have obligations to do so in ways that provide adequate notice and justification and also respond to the values of coherence. Finally, to the extent that agencies are engaged in adjudication, the values of procedural fairness apply. This suggests focus on the following five elements or dimensions of the rule of law.

(1) Authorization. Authorization is a bedrock principle of liberalism and the rule of law; it ensures that the state only act or constrict an individual’s liberty when authorized to do so. 30 See Cass, supra note 17, at 12–13 (identifying valid authority as element of rule of law); McDonald, supra note 7, at 204 (noting as fundamental element of rule of law that political power be authorized); see also Franz Neumann, The Democratic and the Authoritarian State: Essays in Political and Legal Theory 163 (1957) (“[T]he state may intervene with the individual’s liberty—but first it must prove that it may do so.”). Authorization demands a positive law source that grants power to the government to act. A system that complies with authorization is one in which official acts are within the scope of powers authorized, or not ultra vires

(2) Notice. Many of the most commonly identified features of the rule of law pertain to a cluster of characteristics that help to ensure that law has the capacity to be practical in the sense of providing guidance to an individual’s actions and allowing individuals to plan with some knowledge of the law. The principles of publicity, clarity, consistency, prospectivity, and stability are among the most important of these values. 31 See Fuller, supra note 12, 46–90 (arguing for place of these values as central to law’s virtue); Raz, supra note 14, at 214–16 (defending requirement that law be pro­spective, open, clear, and general). Some of these principles are nearly categorical. There can be no basis to demand compliance with nonpublic or secret laws. To be action-guiding, laws must be knowable and public. Likewise, a retroactive law cannot purport to guide conduct. Some of these values are a matter of degree; it is more difficult to comply with laws that are unclear, inconsistent, or change so quickly that they cannot (reasonably) claim to be capable of guiding action.

(3) Justification. An important strain of thought about the rule of law focuses on the role of justification and argumentation in law. Justification provides protection against arbitrariness; part of what defines arbitrary action is action that is not justified. As Professor Jeremy Waldron writes,

[L]aw is an argumentative discipline, and no analytic theory of what law is and what distinguishes legal systems from other systems of governance can afford to ignore this aspect of our legal practice and the distinctive role it plays in a legal system’s treating ordinary citizens with respect as active centers of intelligence. 32 Waldron, Importance of Procedure, supra note 1, at 21–22 (citing Neil MacCormick, Rhetoric and the Rule of Law: A Theory of Legal Reasoning 14–15, 26–28 (2005)).

Or, as Professors David Dyzenhaus and Michael Taggart put it, our legal system reflects “the pull of justification, meaning that public power is considered authoritative when and only when it justifies its exercise to those whom it affects.” 33 David Dyzenhaus & Michael Taggart, Reasoned Decisions and Legal Theory, in Common Law Theory 135, 152 (Douglas E. Edlin ed., 2007). This demand for justification, and thus practices of reason-giving and processes for argumentation, are grounded in law’s respect for human agency and dignity, and treatment of law’s subjects as “bearers of reason and intelligence.” 34 Waldron, Importance of Procedure, supra note 1, at 19.

(4) Coherence. Law presents itself as a system in which norms fit together. 35 See Waldron, The Concept and the Rule of Law, supra note 14, at 32–36 (discussing systematicity as legal value). It is from the perspective of the private party—the individual or firm subject to law’s demands—that coherence matters. “For citizens, law is inevitably an integral system, premised in contemporary social expectations and political judgments; a person interested in her legal obligations looks to the whole environment, not a disordered collection of fragmentary, isolated, mutually independent pieces.” 36 Peter L. Strauss, On Resegregating the Worlds of Statute and Common Law, 1994 Sup. Ct. Rev. 429, 442 [hereinafter Strauss, Resegregating]. This does not say how coherence is achieved, but it does emphasize the importance of viewing coherence from the perspective of the private individual subject to law’s obligations.

(5) Procedural Fairness. A central virtue of the rule of law is procedural fairness, that is, the set of institutional arrangements that provide an unbiased determination of one’s rights and duties through transparent procedures with determinations based on evidence. Here the rule of law joins company with the most basic elements of due process, though it can be more demanding. 37 Whereas trial by an elected judge does not violate the Due Process Clause, see Republican Party of Minnesota v. White, 536 U.S. 765, 782–84 (2002) (noting elected judges always face pressure from electorate regarding content of their rulings but rejecting view that Due Process Clause prohibits election of judges), there may be reasons from the perspective of the rule of law to question the impartiality of elected judges.

*     *     *

II. Rule-of-Law Principles for Administrative Governance

The question to ask now—with aid from Professor Strauss—is how we can specify these general rule-of-law principles in the context of administrative governance. In what ways are these principles applied in our administrative law? In what way do they ground arguments for an account of that law? This Part takes on that task organized around the five dimensions of the rule of law just discussed.

A. Authorization

Authorization is a central principle of the rule of law and also a central occupation of administrative law. The rule-of-law principle is that government may act and may constrict an individual’s liberty only when authorized to do so. 38 See supra note 30 (citing sources and describing authorization as element of rule of law). That is, government’s actions must be authorized by some valid source. 39 See Cass, supra note 17, at 12 (noting valid authorization as element of rule of law). Within our system of limited government, with a Constitution creating a government of only enumerated powers, an administrative agency only has those powers Congress confers upon it. 40 See La. Pub. Serv. Comm’n v. FCC, 476 U.S. 355, 374 (1986) (“[A]n agency literally has no power to act . . . unless and until Congress confers power upon it.”). Administrative law thus must provide an account of which officials may exercise delegated statutory power and how the scope of that power is to be judged. Professor Strauss’s answers to these widely contested questions have distinctive grounding in rule-of-law considerations and suggest two rule-of-law principles for administrative government. These principles share a common conception of the value of independent legal judgment for administrative officials and for courts.

1. Decisional Allocation. — One pillar of Professor Strauss’s approach to public law is an insistence that decisional allocation—attention to which official has been vested with power—matters to the chances of government in accordance with law. To be clear, the critical issue is not what occurs when one agency exercises powers delegated to another agency. Nor is the principal issue the commonplace practice of subor­dinate officials acting under the general authority and direction of their superiors; that is a simple fact of organizational life. Rather, in the United States, the significant question of decisional allocation arises with respect to the scope of the President’s powers over officials who are delegated powers and vested with nonministerial duties by statute, whether or not they are protected from removal by good cause pro­tections. What power does the President have to legally bind the discretion of, for example, the Secretary of Transportation, the Administrator of the Environmental Protection Agency, or the members of the Securities Exchange Commission?

Professor Strauss is the leading contemporary defender of the view that when Congress imposes duties and grants discretion to offices or agencies, those duties and that discretion are personal to the officeholder. 41 See Peter L. Strauss, Foreword: Overseer, or “The Decider”? The President in Administrative Law, 75 Geo. Wash. L. Rev. 696, 704–05 (2007) [hereinafter Strauss, The President in Administrative Law] (arguing in ordinary administrative contexts, where Congress delegates to named agency official, President’s role is supervisor, not decider); see also Strauss, The Place of Agencies, supra note 9, at 649 (arguing Congress has power “to place the responsibility for decision in a department rather than the President”); Peter L. Strauss, Was There a Baby in the Bathwater? A Comment on the Supreme Court’s Legislative Veto Decision, 1983 Duke L.J. 789, 808 (“The legal authority to act is then that of the delegate, and even for indisputably executive agencies the President’s power of direction appears limited in ways that make it difficult to characterize him as the delegate.”). This position has a long history. See, e.g., 1 Op. Att’y Gen. 624 (1823), 1823 WL 538, at 625 (“If the laws . . . require a particular officer by name to perform a duty, . . . no other officer can perform it without violation of the law; and were the President to perform it, . . . he would be violating [the law] himself.”); Edward S. Corwin, The President: Office and Powers 1787–1984, at 94–100 (5th rev. ed. 1984) (arguing duties imposed on named offices are not President’s in part to give Congress a choice to delegate to entity other than President). For other explorations, see, e.g., infra notes 43–44 (collecting sources), as well as Harold H. Bruff, Presidential Management of Agency Rulemaking, 57 Geo. Wash. L. Rev. 533, 539 (1989) (arguing President lacks directive authority when Congress delegates to other officials); Richard H. Pildes & Cass R. Sunstein, Reinventing the Regulatory State, 62 U. Chi. L. Rev. 1, 25 n.97 (1995) (arguing against President’s directive authority over agencies). In other words, absent special circumstances, the President does not have legal authority either to supplant that official’s act or to bind the official to a particular action when that official has been granted statutory authority. The power vested by statute is the official’s, not the President’s. As Professor Strauss puts it succinctly, in these cases, the President is the “overseer and not the decider.” 42 Strauss, The President in Administrative Law, supra note 41, at 704–05. This position has both a constitutional and statutory dimension. As to the Constitution, this position rejects the view, commonly associated with a strongly unitary conception of the executive, that the Constitution requires reading any grant of authority to an official as authorizing the President to act in the official’s stead. 43 See, e.g., Steven G. Calabresi & Saikrishna B. Prakash, The President’s Power to Execute the Laws, 104 Yale L.J. 541, 596 & n.210 (1994) (arguing delegations to executive officials should be construed to permit President to exercise officials’ delegated powers directly, for instance, by personally issuing standards of workplace safety in stead of Secretary of Labor). As to statutory construction, this position takes dele­gations to executive branch officials as well as to independent agencies as vesting power in the chosen official, not the President. 44 See Strauss, The President in Administrative Law, supra note 41, 713–17 (arguing delegation to officials does not grant President access to those powers). Others have joined this debate. For arguments that the President has authority to exercise powers delegated to other executive officials, see Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2327–30 (2001) (arguing delegations to executive branch officials authorize President to exercise officials’ powers); Nina A. Mendelson, Another Word on the President’s Statutory Authority over Agency Action, 79 Fordham L. Rev. 2455, 2458–74 (2011) (arguing delegations to executive officials do not imply limit on President’s directive authority). For arguments that the President generally lacks statutory authority to direct the exercise of power granted to other officials, see, e.g., Robert V. Percival, Presidential Management of the Administrative State: The Not-So-Unitary Executive, 51 Duke L.J. 963, 1007–11 (2001) (raising statutory, historical, and policy arguments why President lacks directive authority); Kevin M. Stack, The President’s Statutory Powers to Administer the Laws, 106 Colum. L. Rev. 263, 276–99 (2006) (arguing congressional delegation practices undermine inference of implied delegation to President when Congress names another official). The focus on decisional allocation thus maintains that Congress, in delegating authority, may make a meaningful choice among delegates, including granting power to the President, but also to other officials. 45 See Strauss, The President in Administrative Law, supra note 41, at 713 (discussing view that inability of Congress to delegate authority exclusively to executive officials “render[s] it impossible for Congress . . . to leave anything to the specially trained judgment of a subordinate executive official” without risking politicization of official’s decisions (internal quotation marks omitted) (quoting Edward S. Corwin, The President: Office and Powers 1787–1957, at 80 (4th rev. ed. 1957))).

While this position on decisional allocation has been widely debated in constitutional and statutory terms, it can be seen as grounded in a distinctive set of commitments about the prospects for government under law. “[T]here is a difference between ordinary respect and political defer­ence, on the one hand, and law-compelled obedience, on the other,” Professor Strauss writes; “[t]he subordinate’s understanding which of these is owed, and what is her personal responsibility, has implications for what it means to have a government under laws.” 46 Id. at 704 (emphasis added). The suggestion here is that when an official views her duties under statute as her own, that funda­mentally shapes the “frame of mind” or the “psychology of office” 47 Id. at 712–13. in which the officeholder receives urgings from superiors (and others). At the most basic level, “someone told me to do it” is excluded as a sufficient ground for action by an official vested with delegated power. Such a delegate should generally grant respectful consideration to the views of superiors, but the duty and power of decision are ultimately her own. For a decision to be the official’s own, she must be independently convinced of the action’s legality and appropriateness.

This insistence on decisional allocation thus can be seen as grounded in a pragmatic principle that there is a greater chance of decision in accordance with the law when officials view their duties and powers as personal, requiring their independent judgment, and not subject to supplanting by others. This idea can be put in terms of the definition of role for administrative actors. When the legal role of those delegated statutory power is defined as requiring their independent judgment, the specter of role-based compliance up a chain of command is diminished. Institutionally, this role specification spreads accountability through the bureaucracy. All those with legally delegated authority must exercise their own independent legal judgment; as a result, administrative action will represent the views of many actors, and accountability cannot be confined to the officials at the peak of the institutional hierarchy. These ideas about the foundation for decisional allocation might be formulated in terms of the following rule-of-law principle for administration: Legal authorization (and duty) is relative to officeholders, not an impersonal authorization to government as a whole.

While this principle is articulated at a relatively high level of generality, it takes a stance on the contested question of the President’s powers over law administration. Understanding this legal allocation—that the discretion and duty is personal to the official—clarifies how an official delegated with statutory power is to understand prodding from a President or his immediate advisors. The official is not to take that direction “as a command that she has a legal as well as a political obligation to honor, and for whose justifications she thus has no particular responsibility.” 48 Id. at 712. As opposed to hierarchical military command, the principle of decisional allocation maintains that the dialogue between the President and the agency is necessarily anchored in the requirement of authorization and, consequently, the goals of the under­lying delegating legislation, which are the core positive foundations for statutory law implementation. If the responsibility is the official’s, it is the official who must be convinced and who is ultimately accountable for the decision. The President, then, must persuade the official.

This position does not deny that politically appointed officials are picked and vetted to carry out their duties in accordance with the President’s priorities, 49 Cf. Exec. Order No. 12,866, 3 C.F.R. § 4 (1994), reprinted as amended in 3 U.S.C. § 601 app. at 803 (2012) (providing consistency with President’s priorities is part of regulatory planning process); id. § 6(b) (providing OIRA may review agency action for consistency with President’s priorities). nor that many of them may be fired by the President at will for failing to do so. Nor does it deny that there are relevant differences between the weight of presidential priorities for executive and independent agencies. But it still insists that even for executive officials, as well as those further down the institutional hierar­chy, there is a distinction worth maintaining about whose duty and power is at issue. While this position augments the place of disagreement within the administration, the prospect for disagreement provides an indication and assurance that independent judgment, typically from multiple individuals, has been exercised.

Inquiring into the fundamental rule-of-law demand for author­ization within the administrative context thus reveals the need to make a distinction between authorization as an impersonal grant of powers to government and authorization as delegation to particular officeholders. Viewing obligations as personal to the officeholder opens up a prospect for legal accountability within hierarchical institutions foreclosed by glossing over or denying this distinction.

2. Scope of Authority. — But what is the scope of authority granted? Because agencies only have powers granted to them by statute, the rule of law clearly requires that an agency act within the scope of that delegated power. Indeed, this ultra vires principle—that only authorized action is valid—is, and could be nothing other than, a cornerstone of administrative law.

While all agree that agencies can act only within the scope of their authorization, there is wide disagreement over how that scope is to be determined. This question is at the center of the persisting and generative debate over Chevron and its “first step” inquiry into the statutory permissibility of agency action. Chevron asks the reviewing court to first assess “whether Congress has directly spoken to the precise question at issue.” 50 Chevron U.S.A. Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842 (1984). Many view this first inquiry into the scope of statutory authorization as one that a court could only ask de novo without any form of deference to the agency’s position. 51 See, e.g., Jack M. Beermann, End the Failed Chevron Experiment Now: How Chevron Has Failed and Why It Can and Should Be Overruled, 42 Conn. L. Rev. 779, 782 & n.7 (2010) (arguing Chevron should be overruled as poorly justified and inconsistent with the Administrative Procedure Act (APA)). Others, concerned about judicial overreaching and micromanagement of agencies, suggest that delegation of technical and other matters to the agency qualifies the judicial inquiry. 52 Cf. Lisa Schultz Bressman, Chevron’s Mistake, 58 Duke L.J. 549, 555 (2009) (arguing congressional rationale for interpretive delegation should determine level of scrutiny of statute’s meaning).

On this question, fundamental to the rule of law’s application, Professor Strauss stakes out a middle position that, on the one hand, recognizes the underlying reasons for creation of agencies as part of government and, on the other hand, does not withdraw from the insist­ence on an independent judicial determination of the scope of an agency’s authority. 53 See Peter L. Strauss, “Deference” Is Too Confusing—Let’s Call Them “Chevron Space” and “Skidmore Weight,” 112 Colum. L. Rev. 1143, 1145, 1150 (2012) [hereinafter Strauss, Deference Is Too Confusing] (noting determination of powers vested in agencies is judicial responsibility). Professor Strauss takes as fundamental that the structure of government—a structure created by law—should inform the way in which courts approach the task of determining the legality of the government action they review. His account thus offers a description of the shape of ultra vires review for the administrative context.

Professor Strauss is clear that the question of the scope of an agency’s authority—that is, whether an agency is acting within its “boundaries” 54 Id. at 1150. —is ineluctably and appropriately an issue for inde­pendent judicial evaluation. 55 See, e.g., Kenneth A. Bamberger & Peter L. Strauss, Chevron’s Two Steps, 95 Va. L. Rev. 611, 611 (2009) (noting Chevron’s step-one question of scope of power granted is for independent judicial judgment); Strauss, Deference Is Too Confusing, supra note 53, at 1165 (noting scope of powers vested to agency is matter to be judicially determined); Peter L. Strauss, Overseers or “The Deciders”—The Courts in Administrative Law, 75 U. Chi. L. Rev. 815, 819 (2008) [hereinafter Strauss, Courts in Administrative Law] (“Chevron step one is the terrain of independent (albeit perhaps influenced) judicial judgment . . . .”). But he is equally clear that independent judicial judgment does not exclude—and indeed should include—a court giving “weight” to the agency’s judgment. “The lines defining an agency’s Chevron space must be judicially determined, a determination that is, irreducibly, a statement of what the law is. But that unmistakably judicial determination should be informed by agency judgments in ways that have been conventional at least since 1827.” 56 Strauss, Deference Is Too Confusing, supra note 53, at 1165; Strauss, Courts in Administrative Law, supra note 55, at 818 (explicating precedential grounding for this position); see also Peter L. Strauss, In Search of Skidmore, 83 Fordham L. Rev. 789, 796 (2014) [hereinafter Strauss, In Search of Skidmore] (“One can readily agree . . . that [w]hether Congress has conferred such power is the relevant question[] . . . that must be answered before affording Chevron deference, without . . . having to agree that whether an agency enjoys that authority must be decided by a court, without deference to the agency.” (alterations in original) (internal quotation marks omitted)). In short, courts should review whether the agency has acted within the scope of its authority—the Chevron step-one question—by according Skidmore weight to the agency’s judgments. 57 See Strauss, Deference Is Too Confusing, supra note 53, at 1153–56 (invoking Skidmore weight in reference to formulation of deference in Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944)). Skidmore weight regards “the rulings, interpre­tations and opinions” of the agency as representing “a body of experi­ence and informed judgment to which courts and litigants may properly resort for guidance,” depending on “the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to per­suade, if lacking power to control.” 58 Skidmore, 323 U.S. at 139–40.

Professor Strauss defends this position as the best reading of prece­dent but also the approach that makes most sense of the agency’s position in government. 59 As to precedent, Professor Strauss reminds us that judicial review of agency action did not begin with Chevron or the APA, and that Chevron actually provided an awkward reformulation of principles well established at the time. See Strauss, Deference Is Too Confusing, supra note 53, at 1161–63 (suggesting Chevron fits awkwardly with prior leading decisions). Pre–APA judicial review comprehended that independent judicial review of the agency’s authority, even when understood as an exclusive judicial function, did not prevent giving due consideration to the “contemporaneous construction of a statute by the men [and women] charged with the responsibility of setting its machinery in motion, of making the parts work efficiently and smoothly while they are yet untried and new.” Id. at 1155 (internal quotation marks omitted) (quoting Norwegian Nitrogen Prods. Co. v. United States, 288 U.S. 294, 315 (1933)). Pre–APA review also recognized that when the agency had been allocated authority to establish policy, the court’s role in reviewing agency actions falling within the boundaries of its authority was of oversight and supervision. See id. at 1159–61 (explicating rule of reviewing court under NLRB v. Hearst Publ’ns, Inc., 322 U.S. 111 (1944)); see also Strauss, Courts in Administrative Law, supra note 55, at 818 (same). Professor Strauss argues that lines of current judicial doctrine have unsettled these understandings reflected in the APA by assuming that independent review excludes giving weight to the agency’s view. See Strauss, In Search of Skidmore, supra note 56, at 796 (criticizing City of Arlington on this ground). Congress vests the agency with the statutory duty to definitively implement the statutory scheme and to make sense of it in relation to other laws. 60 See Strauss, Deference Is Too Confusing, supra note 53, at 1146 (observing agencies have authority to act definitively and responsibility to implement statute in coherent way). The justification for granting agencies’ views binding authority within their powers does not vanish when the question is the scope of the agency’s powers. 61 See id. (arguing agency’s responsibilities and authority justify granting Skidmore weight in judicial determination of scope of its powers). In other words, the fact that a statute allocates interpretive authority to the agency to specify the statute’s meaning does not undermine, but rather provides grounds for taking seriously the agency’s views about the scope of that authority. This perspec­tive thus emphasizes the categorization of these issues for the purposes of judicial review—for instance, determining the scope of authority as opposed to review of acts within that scope—should not sweep aside the underlying continuity that it is the same official or agency acting.

This middle position provides a specification of ultra vires review for administrative governance. Acting within the scope of legal authorization is too basic to government under law to evade independent judicial review. But the structure of that review should reflect, not contradict, the underlying place of the agency within government. Accordingly, the justification for a court recognizing that the agency has been vested with power to decide authoritatively within its sphere of powers—a justifi­cation drawing from congressional choice and agency experience—does not vanish when the question is the scope of those powers. In this sense, the court’s commitment to ensuring compliance with the law is not disconnected from an understanding of the legal system as a whole, and the place of the agency within it. This specification could be seen as founded on a more general rule-of-law principle for administrative government: that the shape of ultra vires review should reflect the under­lying legal allocation of authority, such that independent review may still involve respectful consideration of the views of those delegated power in the first instance.

This understanding of the judicial role has deep roots in public law in the United States. In particular, it has strong parallels to Professor James Thayer’s classic position on the narrow role of courts in reviewing the constitutionality of legislation. 62 See generally James B. Thayer, The Origin and Scope of the American Doctrine of Constitutional Law, 7 Harv. L. Rev. 129, 140 (1893) (developing early theory of role of courts in review of legislation for constitutionality). Professor Thayer argued that with regard to the “momentous” power of judicial review of the consti­tutionality of legislation, courts should grant Congress’s views respect because Congress has been expressly entrusted by the Constitution with the exercise of legislative powers, “not merely of enacting laws, but of putting an interpretation on the constitution which shall deeply affect the whole country, enter into, vitally change, even revolutionize the most serious affairs.” 63 Id. at 136. For Professor Thayer, this “respect” was not a matter of mere courtesy, but based on “very solid” grounds of “policy and law.” 64 Id. The Constitution’s entrusting to Congress not merely the power of preliminary or provisional action but presumptively final action, 65 Id. at 135 (“[C]onstitutions not merely intrust to the legislatures a preliminary determination of the question, but [also] contemplate that this determination may be the final one . . . .”). for Professor Thayer, narrows the judicial role. 66 Id. at 135–36 (observing power of “putting an interpretation on the constitution which shall deeply affect the whole country” is given to legislature, and, as a result, legislature’s determinations warrant respect). Much the same logic applies with regard to judicial review of the scope of an agency’s powers. While agencies lack the direct electoral connection of Congress, under many statutory delegations, their actions, too, are presumptively final. Accord­ingly, respect for their judgments is not merely a matter of courtesy, but also grounded in law. Giving agencies’ views “weight” even in the determination of their authority offers fidelity to the law in Professor Thayer’s sense—it recognizes that the allocation of respon­sibility is to the agency to act with presumptive finality. In short, how authority is judged is a function in part of how it is initially allocated.

As Professor Strauss acknowledges, this perspective confronts chal­lenges today. Independent judicial review of an agency’s action is often reflexively understood to exclude giving the agency’s view any weight. 67 Strauss, In Search of Skidmore, supra note 56, at 795–96 (noting City of Arlington Court assumed Chevrons first step could not involve any form of deference to agency’s views). To take one example, consider how the Supreme Court understands the judicial task in Chevron’s first step. As Professor Strauss writes, both the majority and the dissent in City of Arlington v. FCC 68 133 S. Ct. 1863 (2013). pass over the possibility that one of the traditional tools of statutory interpretation applicable at the first step of the Chevron inquiry is according some weight to the agency’s views. 69 See Strauss, In Search of Skidmore, supra note 56, at 795–96 (critiquing City of Arlington on this ground). Instead, both opinions take independent judicial inquiry to exclude weight to the agency’s views. 70 Id. (observing this view among opinions in City of Arlington). As a result, Skidmore’s advice to give due weight to the agency’s views is made relevant only outside of Chevron’s application instead of within it. What is lost is a prospect for greater accommodation of the underlying allocation of legal authority within the framework of judicial review.

3. Conclusion. — Viewing together these two principles of authorization—decisional allocation and deference as to scope—reveals an interesting commonality as to the value of independent legal judg­ment for administrative governance. Independent judgment appears as a virtue that applies to a wider range of legal officials but permits greater consideration of the views of others than many suppose. The principle of decisional allocation is premised in part on the idea that independent judgment is a fundamental value in the executive branch and an unavoidable duty of executive officials, not just a virtue reserved for courts or only particular “independent” agencies. Thus the duty has wide application. At the same time, exercising independent judgment does not preclude giving weight to the views of other actors—regardless of whether that independent judgment is exercised by an executive branch official or a court—and so is less demanding than some assume. This view thus sees an underlying commonality in the legal duties of agency officials and courts; both labor under the burdens of independent judgment, but under the rule of law, such laboring does not require isolation or excluding due consideration of others’ views.

B. Notice

At the center of most accounts of the rule of law is a cluster of formal characteristics that assist law in guiding individuals’ actions. 71 See, e.g., Fuller, supra note 12, at 53 (“Law has to do with the governance of human conduct by rules.”); Fallon, supra note 3, at 8 (“The first element [of the rule of law] is the capacity of legal rules . . . to guide people in the conduct of their affairs.”). Principles of publicity, clarity, consistency, prospectivity, and stability are among the most important. 72 See Fuller, supra note 12, at 46–90 (defending publicity, prospectivity, clarity, consistency, and stability as among requirements for law); Raz, supra note 14, at 214–16 (arguing law must be prospective, open, clear, and general). To the extent the law falls short of these principles, it is difficult to maintain that individuals have reasonable notice.

Administrative government has been thought to pose particular challenges for this cluster of formal values. In particular, scholars argue that delegation of lawmaking authority in extremely broad terms to agencies undermines these notice values. 73 See, e.g., Hayek, supra note 11, at 75–76 (arguing rule of law requires government be “bound by rules fixed and announced beforehand” and is therefore undermined by “discretion left to the executive organs”). As a result, in the admin­istrative context, the first issue is to clarify the type of law to which these principles apply. This section first argues that these notice principles apply to law that binds the public, which in our system is frequently the rules and other law produced by agencies. Based on that premise, this section then discusses two further implications. If rulemaking fares better than adjudication with regard to these notice values, then these principles may impose a prima facie obligation on agencies to engage in rulemaking. In addition, it argues that agencies also have an obligation to issue prospective guidance as a second-best option when rulemaking is not practicable.

1. The Locus of Notice Demands. — If these notice principles of the rule of law—publicity, clarity, consistency, prospectivity, and stability—apply directly to regulatory legislation, such legislation fails to comply in several important respects. A basic feature of our administrative govern­ment is broad legislative delegations to administrative officials and agencies, delegations that are not only broadly worded but also do not impose obligations directly on private parties—characteristics for which Professor Edward Rubin proposes the term “intransitive.” 74 Edward L. Rubin, Law and Legislation in the Administrative State, 89 Colum. L. Rev. 369, 380–85 (1989). Modern legislation “in its essence is an institutional practice by which the legis­lature, as our basic policy-making body, issues directives to the govern­mental mechanisms that implement policy.” 75 Id. at 372; see also Theodore J. Lowi, The End of Liberalism: The Second Republic of the United States 106 (2d ed. 1979) (“Obviously modern law has become a series of instructions to administrators rather than a series of commands to citizens.”). To be sure, Congress does enact some statutes that impose obligations directly on private persons, and some statutes are written with a great deal of specificity. But as administrative lawyers and political scientists have long recognized, the vast weight of modern legislation “regulates the behavior of government agencies, not the conduct of private persons.” 76 Rubin, supra note 74, at 376; see also David Epstein & Sharyn O’Halloran, Delegating Powers: Transaction Cost Politics Approach to Policy Making Under Separate Powers 5 (1999) (noting broad delegation characterizes modern administrative state); McCubbins & Sullivan, supra note 9, at 403 (“[T]he nexus of policy making has largely shifted from the constitutionally designated branches of government to the bureaucracy . . . .”). As opposed to creating primary obligations for private parties, regulatory statutes structure the processes, means, and considerations for agencies. These familiar features of regulatory statutes have important consequences for rule-of-law principles of notice: If the legal system’s compliance with these values depends upon the text of regulatory statutes, we would be forced to conclude either that the system dramatically falls short of these principles or that these principles require revision. 77 Professor Rubin argues in this vein that “[w]hen a transitive statute is enforced by an agency, our normative system simply does not make the demands that Fuller perceives.” Rubin, supra note 74, at 399.

The same result does not follow, however, when we understand these notice principles, as Professor Strauss argues, as “obligation[s] applicable to the system” 78 Peter L. Strauss, Legislative Theory and the Rule of Law: Some Comments on Rubin, 89 Colum. L. Rev. 427, 445 (1989) [hereinafter Strauss, Comments on Rubin]. as opposed to regulatory legislation itself. On this view, “[t]he agency’s development and enunciation of administrative policy” provide the specification of what the law demands of private parties. 79 Id. This position—that we should ask how agency actions imposing obli­gations on private parties comply with these formal rule-of-law values—has strong theoretical foundations. As noted above, a fundamental aspect of law is that it provides guidance as to the behavior of private parties—it aims to impose practical obligations. To be action-guiding, however, law must be accessible, consistent, reasonably clear and stable, and pro­spective. 80 See, e.g., Fuller, supra note 12, at 53 (“To speak of governing or directing conduct today by rules that will be enacted tomorrow is to talk in blank prose.”). Based on the premise that these principles of the rule of law seek to protect law’s action-guiding qualities, it makes sense that these notice demands apply to agency action that binds private parties, and not merely to delegating legislation. Accordingly, principles of notice properly apply to agency action that creates binding obligations for private parties as well as those aspects of statutes that do so. The system as a whole is thus still on the hook for satisfying these formal demands. But these demands apply to legal sources that bind the public, which include, significantly, the rules and decisions issued by administrative agencies.

At a practical level, this insight defuses some lines of challenge to administrative governance that fix upon legislation as the focus of these rule-of-law values. 81 See, e.g., Hayek, supra note 11, at 80–81 (arguing broad delegation threatens rule of law). But more importantly, recognizing that these formal values apply to all sources of law governing private conduct in society frames as a critical inquiry how well agencies comply with these principles in their lawmaking. In other words, a critical element of administrative agencies’ compliance with the rule of law is the ways in which their lawmaking embodies the values of publicity, prospectivity, clarity, and so on—that is, Professor Fuller’s demands of law need to be brought into agency trenches.

Professor Strauss has long insisted on the fundamental rule-of-law requirements of publicity for agency action 82 See, e.g., Peter L. Strauss, Rules, Adjudications, and Other Sources of Law in an Executive Department: Reflections on the Interior Department’s Administration of the Mining Law, 74 Colum. L. Rev. 1231, 1238–40 (1974) [hereinafter Strauss, Rules, Adjudications, and Other Sources of Law] (arguing internal organizational structure and operating procedures should be published and internal operating manuals be publicly available). and has been at the vanguard of a forceful critique of agencies’ reliance on private standards, not practically accessible without a fee, in their regulations. 83 See Peter L. Strauss, Private Standards Organizations and Public Law, 22 Wm. & Mary Bill Rts. J. 497, 560 (2013) (summarizing publicity norms when agencies incorporate private standards); see also Emily S. Bremer, Incorporation by Reference in an Open-Government Age, 36 Harv. J.L. & Pub. Pol’y 131, 153 (2013) (providing assessment of public access problem for standards incorporative by reference in federal rules); Nina A. Mendelson, Private Control over Access to the Law: The Perplexing Federal Regulatory Use of Private Standards, 112 Mich. L. Rev. 737, 800 (2014) (arguing for full digital access without charge to private standards incorporated by reference into agency rules). But questions of the requisite clarity, prospectivity, and stability required of regulations and other agency actions remain areas for further explo­ration as part of the project of specifying the rule of law’s demands of administrative governance.

2. Implications for “Choice” Among Policymaking Forms. — Many agencies may implement their statutes in a variety of policymaking forms. 84 See M. Elizabeth Magill, Agency Choice of Policymaking Form, 71 U. Chi. L. Rev. 1383, 1383, 1386 (2004) [hereinafter Magill, Agency Choice of Policymaking Form] (noting agencies may implement statutes by rule, adjudication, or announcing interpretation). Agencies may promulgate rules, conduct adjudications, issue interpretive statements or guidance documents, compile and publish enforcement manuals, create permitting schemes, conduct auctions, make grants, create pilot projects, engage in research, and so on. Whether a policy is implemented through rulemaking or adjudication often results from organizational and institutional dynamics within the agency more than it follows from a single decisionmaker’s conscious choice. 85 See Strauss, Rules, Adjudications, and Other Sources of Law, supra note 82, at 1258 (exposing how organizational dynamics, not singular agency judgments, led to rulemaking or adjudication within Department of Interior). It is nonetheless still coherent to ask agencies as institutions to develop structures for making informed allocations among policymaking forms.

Different policymaking forms fare better and worse than others with regard to these rule-of-law notice principles. As Professor Strauss notes, case-by-case adjudications, especially when unguided by strong agency internal policy, are not only costly but can threaten “undesirable variation in individual cases.” 86 Peter L. Strauss, The Rulemaking Continuum, 41 Duke L.J. 1463, 1482 (1992) [hereinafter Strauss, Rulemaking Continuum]. More generally, adjudicative decisionmaking processes, like common law processes, strain this cluster of rule-of-law virtues. 87 See Frederick Schauer, Is the Common Law Law?, 77 Cal. L. Rev. 455, 455 (1989) (reviewing Melvin A. Eisenberg, The Nature of the Common Law (1988)) (“[C]entral features of common law method appear inconsistent with some of the primary assumptions of a traditional view of the rule of law.”). In common law adjudication, “rules [are] created in the very process of application” and thus apply “retroactively to facts arising prior to the establishment of the rule.” 88 Id. As to values of prospectivity, clarity, publicity, and generality, common law adjudication often fares worse than prospective legislation. These same deficits would also seem to apply to administrative adjudication in comparison to rulemaking.

Based on the assumption that not all procedural forms are created equal with regard to their compliance with formal rule-of-law princi­ples—and, more specifically, that rulemaking is generally preferable—those principles should supervene on how the agency allocates its activities among procedural forms. Well-established judicial doctrine effectively bars courts from second-guessing the agency’s choice about the best policymaking form to use. 89 See SEC v. Chenery Corp. (Chenery II), 332 U.S. 194, 203 (1947) (“[T]he choice made between proceeding by general rule or by individual, ad hoc litigation is one that lies primarily in the informed discretion of the administrative agency.”). Indeed, as Dean Elizabeth Magill has highlighted, the agency’s discretion to opt for different policymaking forms remains an exception to the general requirement that the agency exercise its discretion in a reasoned way. 90 See Magill, Agency Choice of Policymaking Form, supra note 84, at 1415 (“There is simply no such reason-giving requirement imposed on an agency when it selects its choice of form.”); see also, e.g., Chenery II, 332 U.S. at 203 (“Hence we refuse to say that the Commission, which had not previously been confronted with the problem of management trading during reorganization, was forbidden from utilizing [adjudication] for announcing and applying a new standard of conduct.”). To put the point in the shorthand of students of administrative law, the discretion to select among policymaking forms that the Supreme Court embraced in its Chenery II decision is an exception to the reason-giving requirements of its Chenery I opinion. See SEC v. Chenery Corp. (Chenery I), 318 U.S. 80, 94 (1943) (“[T]he orderly functioning of the process of review requires that the grounds upon which the administrative agency acted be clearly disclosed and adequately sustained.”). Under current law, the agency does not need to offer a justification for its choice among alternative procedural forms.

The diversity of agencies’ organizational structures and the practical necessity of gaining more information about a regulatory environment before developing a rule, among other considerations, 91 See Magill, Agency Choice of Policymaking Form, supra note 84, at 1445–47 (arguing courts can address concerns related to agency’s choice of form through other doctrines); Strauss, Rules, Adjudications, and Other Sources of Law, supra note 82, at 1265–66 (arguing against requirement that agencies formulate policy through rules). counsel against constricting established judicial tolerance for agency choice among policymaking forms with a requirement that agencies utilize rulemaking to the fullest extent possible. But the rule-of-law benefits of rulemaking over adjudication do impose some obligation on the agency. One minimal way that obligation might be specified is an obligation that the agency justify its choice of procedural form. 92 See Lisa Schultz Bressman, Beyond Accountability: Arbitrariness and Legitimacy in the Administrative State, 78 N.Y.U. L. Rev. 461, 544 (2003) (suggesting agencies should be required to justify opting for procedures other than rulemaking). Such a requirement of justification, whether or not judicially enforceable, 93 A useful analogy is the Regulatory Flexibility Act’s requirement that the agency provide a “regulatory flexibility analysis” that includes “a statement of the need for, and objectives of, the rule,” 5 U.S.C. § 604 (2012), published in the Federal Register. 5 U.S.C. § 604(a)(6), (b). These requirements are “[p]urely procedural” and so “require[] nothing more than that the agency file a [final regulatory flexibility analysis] demonstrating a ‘reasonable, good-faith effort to carry out [RFA’s] mandate.’” U.S. Cellular Corp. v. FCC, 254 F.3d 78, 88 (D.C. Cir. 2001) (quoting Alenco Commc’ns, Inc. v. FCC, 201 F.3d 608, 625 (5th Cir. 2000)). would create the occasion for agencies to self-consciously evaluate their chosen policy­making form relative to others available. The agency might justify its choice to proceed through adjudication, for instance, because it does not yet know enough about how the statute impacts the regulated community to make a general rule, or based on considerations of timing. But the requirement to make some comparative assessment to justify its choices bridges internal agency silos and requires the agency make a deliberate decision in light of the full complement of its powers. If not all policymaking forms are created equal, and if agencies can adopt structures that allow for deliberation over policymaking forms, 94 Professor Strauss has cautioned that the search for mandatory controls of the allocation of policymaking between adjudication and rulemaking is illusory in part because many agencies do not have an effective mechanism for choice. See Strauss, Rules, Adjudications, and Other Sources of Law, supra note 82, at 1274–75. As a result, he has disagreed with efforts to mandate this choice, see id. at 1265 (arguing adjudicative function cannot be limited to fact-finding and adjudication and inevitably involves policy choice). Justification of policymaking choice is a more minimal demand, though one that does open the door to judicial second-guessing of agency practices. a requirement to justify the choice of form is a modest means of enforcing these rule-of-law values of notice.

3. Obligation to Issue Guidance. — Attention to these notice values also has implications for guidance documents. Guidance documents include interpretations and policy statements of statutes and regulations that do not have the authority to bind with the force of law, but may instruct agency officials how to set forth the agency’s interpretation of a statute or regulation, or exercise their discretion under a statutory scheme. 95 A commonly used definition of a guidance document is that appearing in President Bush’s (now repealed) executive order on guidance. See Exec. Order No. 13,422, 72 Fed. Reg. 2763(3)(g) (2007) (defining guidance document as “agency statement of general applicability and future effect, other than a regulatory action, that sets forth a policy on a statutory, regulatory, or technical issue or an interpretation of a statutory or regulatory issue”); see also Connor N. Raso, Strategic or Sincere? Analyzing Agency Use of Guidance Documents, 119 Yale L.J. 782, 785 n.1 (2010) (using this definition and noting others’ reliance on same). Agency reliance on guidance documents has prompted considerable criticism and calls for increased scrutiny. 96 E.g., Non-Codified Documents Is the Department of Labor Regulating the Public Through the Backdoor?: Hearing on H.R. 3521 Before the Subcomm. on Nat’l. Econ. Growth, Nat. Res., & Regulatory Affairs of the H. Comm. on Gov’t. Reform, 106th Cong. 1–4 (2000) (examining Labor Department’s guidance practices); Robert A. Anthony, Interpretive Rules, Policy Statements, Guidances, Manuals, and the Like—Should Federal Agencies Use Them to Bind the Public?, 41 Duke L.J. 1311, 1373 (1992) (arguing agencies should proceed through notice-and-comment for any rule of general applicability other than mere interpretations if rule is given effect by agency or establishes mandatory standards).

But guidance documents can have significant rule-of-law benefits. As Professor Strauss observes, “The usual interface between a member of the public and an agency does not involve the agency head, but a relatively low-level member of staff . . . .” 97 Strauss, Rulemaking Continuum, supra note 86, at 1482. That interface is rife with the possibility of inconsistency in application, and thus raises questions about how best to channel the bureaucrat’s discretion. The public and those regulated, Professor Strauss argues, would generally prefer a regime where lower-level bureaucrats are bound to apply instructions issued publicly in the form of guidance documents, even though those instructions do not themselves “bind” the public and were not issued through notice-and-comment. 98 Id. at 1483. In short, published guidance documents that specify how agencies will exercise their discretion have clarity, consistency, and publicity benefits to the public and those regulated—they create a form of internal law that structures the agency’s decision­making. 99 See id. (“[T]hese satisfied consumers of publication rules tend not to appear in court . . . .”). See generally Mashaw, supra note 10, at 213, 223–24 (characterizing internal law within Social Security Administration as providing this form of constraint). By educating the public about how the agency intends to act or its understanding of its powers, such guidance documents also “permit[] important efficiencies to those who must deal with government.” 100 Strauss, Rulemaking Continuum, supra note 86, at 1481. The notice benefits of published guidance suggest that agencies have an affirmative obligation to issue guidance documents that provide the public and the regulated with the agency’s best statement of how the agency plans to apply its statutory and regulatory sources, especially when rulemaking is not practicable. In short, the public and regulated would “prefer having publication rules to not having them,” 101 Id. at 1480. and that preference is importantly grounded in rule-of-law values.

Recognizing a prima facie obligation to issue guidance is particularly timely. For years, under the principles of Alaska Professional Hunters Ass’n v. FAA 102 177 F.3d 1030 (D.C. Cir. 1999). and Paralyzed Veterans of America v. D.C. Arena L.P., 103 117 F.3d 579 (D.C. Cir. 1997). an agency could alter authoritative guidance only through a new notice-and-comment rulemaking. Professor Strauss criticizes this doctrine as a poor reading of the Administrative Procedure Act (APA) and because of the constraint it inserts between the central agency and its field offices; effectively, it inhibits lower-level officials from issuing prospective guidance by requiring the agency act through rulemaking to undo it. 104 See Peter L. Strauss, Publication Rules in the Rulemaking Spectrum: Assuring Proper Respect for an Essential Element, 53 Admin. L. Rev. 803, 846–47 (2001) (noting how doctrine has effect of binding superiors to field office guidance). In March 2015, the Supreme Court’s decision in Perez v. Mortgage Bankers Ass’n overruled Paralyzed Veterans and Alaska Professional Hunters. 105 Perez v. Mortg. Bankers Ass’n, 135 S. Ct. 1199, 1203 (2015) (indicating overruling). As a result, agencies no longer face a disincentive to issue and to reevaluate and update their guidance to ensure that it reflects the agency’s best understanding of the statutory scheme. Perez thus clears the way for agencies to comply with this rule-of-law obligation to provide a prospective statement of the agency’s best understanding of the law in guidance when rulemaking is not practicable.

4. Conclusion. — Consideration of how these rule-of-law principles of notice apply in the administrative context yields one theoretical point and emphasizes the importance of several other projects. The theoretical point is that these rule-of-law demands are appropriately applied to law that binds private parties, and so frequently the law agencies produce, not agencies’ authorizing statutes. This theoretical point saves the admin­istrative state from the kind of condemnation that results from applying these values to regulatory legislation. But this point also brings into focus a sequence of more specific inquiries. First, it suggests the need for scholars as well as policymakers to evaluate agency rules and adjudications with regard to these Fullerian virtues, a project Professor Strauss has initiated. Second, in view of rulemaking’s general superiority with respect to these values, this theoretical point suggests an obligation for agencies to justify their choices when implementing policy outside of rulemaking. Third, it suggests that when rulemaking is impracticable, agencies have an obligation to provide guidance conveying their best understanding to the public of how their statutes and regulations operate.

C. Justification

The demand for justification is a central feature of administrative law and the work of administrative agencies. The difficult question is the extent to which the proceduralization of these requirements ends up undermining the aspiration that the agency’s justification for its actions follows from and responds to public participation.

At a formal level, administrative institutions are the paradigm of reason-giving institutions. Indeed, reason-giving requirements emerged for administrative agencies before courts imposed them, putting pressure on courts to fall in line. 106 See Dyzenhaus & Taggart, supra note 33, at 145 (showing mandated reason-giving arose for agencies before courts). We tend to think of argumentation and reasoned deci­sionmaking as having their historical and conceptual core in common law courts, with administrative actors coming to reason-giving later in time. Professors David Dyzenhaus and the late Michael Taggart argue to the contrary: The requirement for reason-giving was formalized for administrative decisionmakers who in turn “put pressure on judges to bring themselves into line with the trend toward legally enforceable reasoned elaboration.” Id. Administrative law has long taken agencies’ reasoned elaboration of grounds for their action as necessary to the validity of agency action and imposed higher duties of reasoned elabo­ration on agencies than on other government actors, such as lower courts or Congress. 107 See Kevin M. Stack, The Constitutional Foundations of Chenery, 116 Yale L.J. 952, 955 (2007) (observing higher reason-giving demands apply in review of agency action than review of lower court judgments or of legislation). Longstanding principles of administrative law require that agency action be upheld only on the basis of the grounds upon which the agency justified its action, treating reliance on post hoc justifications as exceptional. 108 Id. (providing account of foundation of Chenery I principle that agency action cannot be upheld unless upon grounds upon which agency acted in exercising its power). The APA also imposes procedural requirements that agencies state the reasons for their action, whether they are acting through notice-and-comment rule­making, 109 5 U.S.C. § 553(c) (2012) (requiring agency produce “concise general statement” of basis and purpose of regulations). formal adjudication, 110 Id. § 557 (requiring decider to state reasons for decision). or otherwise. 111 Id. § 555(e) (stating brief statement of grounds of denial is necessary and self-explanatory). Part of the way in which administrative law guards against arbitrary agency action is through these requirements of reason-giving and judicial review of agency action under the “arbitrary and capricious” standard. 112 Id. § 706(2)(A) (requiring reviewing courts to hold unlawful agency action that is “arbitrary, capricious, an abuse of discretion, or not otherwise in accordance with law”). Whatever the rule of law requires by way of reasoned justification appears to be already part of administrative law and built into the way admin­istrative agencies do business.

These principles of reasoned justification have a close connection to the principles of authorization discussed above. The requirement of justification reinforces the principle of decisional allocation by requiring reasons, not just action, from the decider. Requiring reasons makes it more difficult to evade the responsibility for independent judgment; indeed, developing reasons takes the decider a long way toward exercising independent judgment. In addition, when reasoned elabo­ration operates in company with the principle of decisional allocation, it is clear that the duty to give reasons is not a general requirement of reasons to be given by government but a requirement of reasons from the person responsible for the action. When viewed in this light, part of the problem with a President having the power to legally direct an agency’s action is that this direction would sever the connection between the agency’s action and its justification. It would result in an agency action without the agency’s own justification. If valid agency action requires justification, then the President must not simply direct the agency but rather convince the agency official of a particular action in terms of the official’s own duties under the statute. Within those terms, the encounter becomes one of persuasion based on reasons, reasons that the agency official has an independent duty to evaluate.

Reasoned elaboration also underlies the deference—or weight—given to agencies by courts and others. As Professors Dyzenhaus and Taggart write, “[T]o require reasons from such [administrative] officials is to imply that they have an important role in interpreting the law, a role that judges with others should respect as long as the officials do a decent job of justifying their decisions.” 113 Dyzenhaus & Taggart, supra note 33, at 165. Reasoning might be thought of as the coin by which agencies pay for deference. 114 See Kevin M. Stack, Interpreting Regulations, 111 Mich. L. Rev. 355, 399 (2012) (noting role of reason-giving as basis for judicial deference to agencies).

While justification may be well reflected in administrative law’s current demands for reasoned elaboration, difficult questions arise with regard to the connection between justification and genuine opportu­nities for participation. In principle, the rule-of-law value of argumentation is not just a demand for the public official or body to provide their own oracular justification, but also to provide a justification that is the product of a participatory process through which stakeholders have an opportunity to present their positions to the government policymakers. The justification is the culmination of a procedure and participation, not a substitute for it. But, with regard to agency rulemaking, requiring the justification be informed by participation raises familiar, thorny issues when that requirement becomes judicially enforceable. On the one hand, judicial enforcement of an agency’s duty to engage commentators can empower those within the agency that care most about reasoned justification. “[A]gency officials cannot know who their judicial reviewers will be,” as Professor Strauss writes, so they cannot “bend their science to particular supposed judicial tastes.” 115 Strauss, Courts in Administrative Law, supra note 55, at 829. As a result, anticipating that there will be a judicial hard look at their decisions has the effect of endowing “those who care about well-documented and well-reasoned decisionmaking a lever with which to move those who do not.” 116 Id. (internal quotation marks omitted) (quoting William F. Pederson, Jr., Formal Records and Informal Rulemaking, 85 Yale L.J. 38, 60 (1975)). There are thus strong reasons to be reluctant to “give that lever up.” 117 Id. On the other hand, at least in the context of rulemaking, when probing judicial review is combined with the requirement that an agency’s rule not depart significantly from its proposals, the agency will do most substantive vetting of their proposals with select stakeholders prior to public opportunities to comment. 118 Professor Strauss made this observation about the Department of Interior in 1972, see Strauss, Rules, Adjudications, and Other Sources of Law, supra note 82, at 1252–53, and now scholars acknowledge this phenomenon as a weakness of rulemaking practices. See, e.g., E. Donald Elliott, Re-Inventing Rulemaking, 41 Duke L.J. 1490, 1492 (1992) (suggesting no administrator turns to notice-and-comment when he or she is genuinely interested in obtaining input from interested parties). A recent study of ninety hazardous-air-pollutant standards set by the U.S. Environmental Protection Agency (EPA) by Wendy Wagner, Katherine Barnes, and Lisa Peters provides an illustration. Wendy Wagner, Katherine Barnes & Lisa Peters, Rulemaking in the Shade: An Empirical Study of EPA’s Air Toxic Emission Standards, 63 Admin. L. Rev. 99 (2011). The study measured the influence of industry, consumer groups, and the public in the formulation of the proposed rules (during the pre-proposal stage) and the impact of their comments on the final rules. With respect to these rules, EPA had on average 178 contacts with interest groups during the development of the proposal, prior to the publication of the proposed rule—more than double the average number of comments received on these rules. See id. at 124. Industry and industry associations had, on average, 170 times more informal communications in the pre-proposal stage than public interest groups. See id. at 125. Accordingly, a challenge for judicial review of rulemaking is to incentivize well-documented and justified decisions while also making the public process one in which meaningful engagement is possible.

The rule-of-law demand for justification thus has two different faces in the administrative context. On the one hand, if the principle is concerned with detailed justification of the grounds for action, admin­istrative agencies and administrative law are models. On the other hand, a challenge of justification, especially with regard to notice-and-comment rulemaking, is to find a way to combine participation with judicial review that does not end up contradicting the purposes of public participation.

D. Coherence

Law presents itself as a system in which norms fit together. 119 See Waldron, The Concept and the Rule of Law, supra note 14, at 33 (positing coherence as dimension of rule of law). What does law’s claim to coherence demand of administrative governance? In view of the American style of statute-making, this demand for coherence falls on agencies as well as courts, requiring agencies to engage in a synthetic and integrating form of statutory interpretation.

1. Locus of the Demand for Coherence. — In thinking about this demand for systemic coherence in a legal system, it matters what type or source of law predominates. Statutes have long been recognized as a dominant form of law. 120 See, e.g. Strauss, Resegregating, supra note 36, at 442 (noting this fundamental point). And in the United States, statutory law has a distinctive character. It does not take the form of a civil code. A civil code purports to provide an integrated and comprehensive statement of the governing norms. 121 See Peter L. Strauss, The Common Law and Statutes, 70 U. Colo. L. Rev. 225, 235 (1999) [hereinafter Strauss, Common Law and Statutes] (noting civil codes “emerge in a single legislative act, after exquisite intellectual consideration, as an integrated whole” and “are rarely if ever amended; and if amended, only after equivalent study and attention to the integrated effects of change”). As a result, the demand for coherence in a country with such a code falls heavily on the drafters and adopters of the code. In the United States, in contrast, statutes have less comprehensive ambitions; they offer specific directions to specific problems, and, even within that more limited domain, they frequently bear clear marks of political bargaining. 122 See id. at 240 (“[O]ur legislative process is an essentially reactive, pragmatic process, and not a proactive or rational one.”).

This basic contrast between a civil code and the more responsive, ad hoc, situational, and overtly political character of legislation in the United States has clear implications for the legal institutions most responsible for creating law’s coherence. If systemic coherence is not built into the legislative process and drafting, then it falls to the institutions with responsibility for interpreting and implementing statutory law, namely administrative agencies and courts. To the extent that pursuing and realizing this value of systemic coherence invariably involves synthesis and constructive judgment, we can expect coherence to ground a role for statutory implementers that involves bringing a wide range of judgment to bear, in the mode of a common law court, even when dealing with statutory materials.

2. The Agency’s Duty of Systemic Coherence. — Professor Strauss provides a vivid picture of the situation of the agency implementing its statutory mandate, which describes the agency’s basic duty to do so in a way that creates coherence. 123 See Peter L. Strauss, When the Judge Is Not the Primary Official with Responsibility to Read: Agency Interpretation and the Problem of Legislative History, 66 Chi.-Kent L. Rev. 321, 329–30 (1990) [hereinafter Strauss, Agency Interpretations] (characterizing this duty). The agency, as Professor Strauss explains, faces distinctive demands to mediate between past and present commitments. The agency staff frequently plays a role in drafting its own enabling legislation. 124 See Strauss, Deference Is Too Confusing, supra note 53, at 1146 (describing agency involvement in drafting organic statutes). The agency’s task is delimited and anchored by that statutory text as well as guided by the set of understandings, forged in part through its legislative history, which informs “what the statute has ‘always’ been understood to mean.” 125 Strauss, Agency Interpretations, supra note 123, at 330. Yet the agency’s implementation of the statute is by design responsive to contemporary political overseers. At times, the views of an agency’s political overseers will overwhelm the agency’s evolving understanding of the statute and its requirements. 126 See id. at 331 (noting this eventuality). But even when that influence is only supervisory, it is understood to appropriately and legitimately shape the agency’s approach. As Professor Strauss writes, “what distinguishes agencies from courts in the business of statute-reading is that we accept a legitimate role for current politics in the work of agencies.” 127 Id. at 335. As a result, the agency’s job is in part to provide as much coherence as possible between past commitments, 128 Cf. Michael Herz, Purposivism and Institutional Competence in Statutory Interpretation, 2009 Mich. St. L. Rev. 89, 104 (noting agencies are “closer to the legislative process” and “have a keener sense” of the process’s compromises and limits). reflected in the statute and the agency’s past practices, on the one hand, and current policy preferences on the other. Of course, there are sometimes abrupt changes in rules, but even then the agency’s job (or duty) is to expose the coherence of the statutory regime underlying those changes.

This points to a larger respect in which the agency faces a demand for coherence. As many regulatory statutes are intransitive, 129 See supra note 74 and accompanying text (characterizing administrative government as involving broad delegations to administrative officials). agencies have distinctive lawmaking powers. Faced with a broad range of judgment, the agency’s “responsibility is to assist in . . . implementation in a coherent, intelligible way.” 130 Strauss, Deference Is Too Confusing, supra note 53, at 1146. One might view the duty of coherence as a necessary feature of tolerating broad delegation. Just as the agency does not have Congress’s prerogatives of obscurantism, 131 See United States v. Nova Scotia Food Prods. Corp., 568 F.2d 240, 252 (2d Cir. 1977) (“[A]gencies do not have quite the prerogative of obscurantism reserved to legislatures.”). the agency also cannot avoid the duty to implement statutory power in a way that shows how the statute fits together, creating an integrated set of legal require­ments. This coherence is one of the most basic demands judicially enforced through arbitrary-and-capricious review.

Does this duty of coherence apply only to making sense of the agency’s particular statutory powers or does it include a broader obligation to read the statute in light of the legal system as a whole? Professor Strauss’s rendering of the agency’s obligations to achieve coherence within its statutory domain has provided a foundation for other scholars to examine the agency’s broader duties to incorporate constitutional and background legal norms within its reasoning. As part of the inquiry into administrative constitutionalism, Professor Gillian Metzger highlights agencies’ obligation to take constitutional norms seriously when imple­menting statutes, as well as their institutional competence to do so. 132 See Metzger, Ordinary Administrative Law, supra note 13, at 522 (defending constitutional interpretation as part of agencies’ role and competence). Professor Kenneth Bamberger also defends agencies’ capacity to take into account broad background norms, including the constitutional implications of their decisions. 133 See Kenneth A. Bamberger, Normative Canons in the Review of Administrative Policymaking, 118 Yale L.J. 64, 96–97 (2008) (defending agencies’ capacity to engage and incorporate background values). This broader duty fits with the techniques of statutory interpretation developed by Henry Hart and Albert Sacks. 134 See Kevin M. Stack, Purposivism in the Executive Branch: How Agencies Interpret Statutes, 109 Nw. L. Rev 871, 911–13 (2015) (arguing agency statutory interpretation should include Hart and Sacks’s element of integrating statutory obligations with Constitution and background values). Underlying Hart and Sacks’s work is the premise that agencies’ duties as actors within our constitutional scheme require them to read their statutes in light of underlying constitutional commitments and thus to seek systemic coherence within our system of government, rather than mere statutory coherence.

This emphasis on agencies’ duty to take into account the value of systemic coherence does not provide a complete account of what coherence involves and could be subject to different specifications. Some define statutory coherence with more emphasis on its textual features and others with greater emphasis on its policy context. But recognizing this duty clarifies that the demand for coherence should be evaluated from the perspective of the individual or firm that is subject to law’s obligations. That perspective on the private party anchors legal develop­ment within a set of social expectations and political judgments. This perspective also provides a foundation for the agency to incorporate its understanding of the Constitution among the considerations that bear on statutory implementation. In sum, given the limited integral aspira­tions of legislation in the United States, agencies have a particular responsibility, grounded in the rule-of-law value of coherence, to implement their statutory scheme in a way that makes it intelligible in light of their statutory powers, surrounding statutory law, as well as constitutional and background legal values.

E. Procedural Fairness

The rule of law does not require or endorse any particular model of division or balance of governmental powers. It is consistent with parlia­mentary government and presidential systems that divide election of the executive and legislature. But it does make demands on the structure of agency adjudications. At a most basic level, the rule-of-law value of proce­dural fairness requires an impartial decider in adjudications. 135 See Concrete Pipe & Prods. of Cal., Inc. v. Constr. Laborers Pension Tr., 508 U.S. 602, 617 (1993) (plurality opinion) (“[D]ue process requires a ‘neutral and detached judge in the first instance . . . .’” (quoting Ward v. Monroeville, 409 U.S. 57, 61–62 (1972))); see also Hamdi v. Rumsfeld, 542 U.S. 507, 533 (2004) (holding one component of due process is opportunity to rebut factual assertions before neutral decisionmaker). This basic demand has implications for the organization of administrative adjudi­cation. In particular, it suggests separation of personnel; those who investigate and prosecute should not also decide. The principle of separa­tion of persons—and in particular the separation of enforcement staff from those who decide—is enforced by the APA, though not completely. 136 The APA requires separation of the agency’s adversarial enforcement staff from its adjudicative decisionmakers as a matter of personnel, oversight, and communications. See 5 U.S.C. § 554(d) (2012) (providing employees engaged in adversarial investigation or prosecution may not “participate or advise in the decision, recommended decision, or agency review . . . except as witness or counsel in public proceedings”); id. § 554(d)(1) (stating hearing officers “may not consult a person or party on a fact in issue, unless on notice and opportunity for all parties to participate”). Section 554(d) also includes a “command influence rule,” which prohibits agency adjudicators from being “responsible to or subject to the supervision or direction of an employee [who] engage[s] . . . in investigative or prosecuting functions for an agency.” Id. § 554(d)(2). But the APA does not mandate this separation for agency heads, who may hear appeals from initial adjudicators and are generally vested with the powers of initial decisionmakers. See id. § 554(d)(2)(C) (“This subsection does not apply . . . to the agency or a member or members of the body comprising the agency.”).

Impartiality (and its appearance) is also threatened when an adju­dicator faces the prospect of removal based on the merits of his or her decisions. This suggests a rule-of-law foundation for removal protections for adjudicators. This protection is clearly evident in the Supreme Court’s tolerance for—and even implication of—removal protections for those agency officials who adjudicate. As Professor Strauss explains, upholding of removal protections for the Federal Trade Commission (FTC) in Humphrey’s Executor v. United States 137 295 U.S. 602 (1935). is explicable as a grant of removal protections for a body engaged in adjudicative, quasi-judicial tasks. 138 See Strauss, The Place of Agencies, supra note 9, at 613–16 (“The Court [in Humphrey’s Executor] was acutely conscious, however, of the extent to which the Commission acted in circumstances calling for judicial impartiality and the removal from politics that might tend to protect it.”); see also Harold H. Bruff, Balance of Forces: Separation of Powers Law in the Administrative State 425 (2006) (noting FTC’s adjudicative functions provide ground for upholding agency’s removal protections).

The principle that adjudicative decisionmaking powers justify good-cause removal protections finds further support in Wiener v. United States. 139 357 U.S. 349 (1958). Recent scholarship has explored internal separation of powers in agencies. See, e.g., Neal Kumar Katyal, Internal Separation of Powers: Checking Today’s Most Dangerous Branch from Within, 115 Yale L.J. 2314, 2316–17 (2006) (lamenting “paucity of thought” regarding nature of checks on executive branch and identifying bureaucratic agencies as “critical mechanism to promote internal separation of powers [in the executive branch]”); M. Elizabeth Magill, Can Process Cure Substance? A Response to Neal Katyal’s “Internal Separation of Powers,” 116 Yale L.J. Pocket Part 125 (2006), http://yalelawjournal.org/forum/can-process-cure-substance-a-response-to-neal-katyala8217s-a8220internal-separation-of-powersa8221 [http://perma.cc/7BBD-8DDL] (insisting “[w]e already have an internally divided executive . . . characterized by bureaucratic overlap, independent agencies, and perennial complaints by Presidents about their inability to control the bureaucracy”); Gillian E. Metzger, The Interdependent Relationship Between Internal and External Separation of Powers, 59 Emory L.J. 423, 436 (2009) (discussing constitutionality of “vast majority of internal separation of powers mechanisms within the Executive Branch”). Long before the rise of interest in the “internal separation of powers” in agencies, Professor Strauss identified separation of functions as a distinctive strain of separation-of-powers jurisprudence. Whereas traditional separation-of-powers models are concerned with the allocation of government institutions among the branches and the implications that follow from those placement decisions, separation of functions is concerned, for instance, with asking about “what combinations of functions or impacts of external influence will interfere with fair resolution of a particular proceeding.” Strauss, The Place of Agencies, supra note 9, at 622. Even though the statute at issue in Wiener was silent as to removal protections, 140 Wiener, 357 U.S. at 350. the Court held that the President lacked authority to remove a member of the War Claims Commission without cause. The Court emphasized that the Commission’s task was “adjudicat[ion] according to law,” which involved reaching decisions “on the merits of each claim, supported by evidence and governing legal considerations, by a body that was ‘entirely free from . . . control or coercive influence, direct or indirect.’” 141 Id. at 355–56 (quoting Humphrey’s Ex’r, 295 U.S. at 629). Further highlighting the grounds for protecting adjudicators from at-will removal, the Wiener Court emphasized that Humphrey’s Executor had “explicitly ‘disapproved’ the expressions in Myers [v. United States] supporting the President’s inherent constitutional power to remove members of quasi-judicial bodies.” Id. at 352 (citing Humphrey’s Ex’r, 295 U.S. at 626–27). The Court in Myers distinguished quasi-judicial powers but made clear that even when engaging in adjudication, decisions not to the liking of the President would still be grounds for subsequent removal. See Myers v. United States, 272 U.S. 52, 117–18, 134 (1926) (stating power to appoint and remove executive subordinates is “certainly . . . not . . . legislative or judicial” and that “moment [President] loses confidence in the intelligence, ability, judgment or loyalty of any one of them, he must have the power to remove him without delay”). While the President may be restricted from removing an official discharging quasi-judicial functions in the midst of a particular case, the Myers Court wrote, the President “may consider the decision after its rendition as a reason for removing the officer, on the ground that the discretion regularly entrusted to that officer by statute has not been on the whole intelligently or wisely exercised.” Id. at 135.

While it is possible to read the Supreme Court’s most recent removal decision as weakening the principle that adjudicative tasks are a sufficient justification for removal protection, 142 See Kevin M. Stack, Agency Independence After PCAOB, 32 Cardozo L. Rev. 2391, 2409–10 (2011) (noting PCAOB exercised adjudicative task and so adjudication alone was not viewed as sufficient basis for removal protection). the principle is still solidly en­trenched and reflects a core element of the rule of law, namely that the impartiality of adjudication is enhanced when the adjudicator does not act under “the Damocles’ sword of removal by the President” 143 Wiener, 357 U.S. at 356.  based on the content of their decisions. And indeed, today most initial adjudicators within administrative agencies are administrative law judges who enjoy good-cause protection from removal. 144 See 5 U.S.C. § 7521 (2012) (noting actions against administrative law judges may only be taken for good cause). Removal of administration law judges is vested in the Merit Systems Protection Board, whose members are themselves protected from removal from office by a good-cause provision. See id. § 1202(d) (“Any member may be removed by the President only for inefficiency, neglect of duty, or malfeasance in office.”). That structural protection is an element of the demands of the rule of law on internal agency organization—and provides reasons to guard against further weakening of removal protections for adjudicators.

F. Summary

Evaluating administrative law through the lens of these five dimen­sions of the rule of law exposes some long-established practices as having troublesome rule-of-law foundations and reveals that other contested practices are well grounded in rule-of-law values. The closest match between the rule-of-law principles and current doctrine and practice is justification; administrative law and practice represents as well as any domain of law the sense in which law is ultimately argumentative.

The idea that agencies have duties to assist in integrating statutory law into the larger fabric of law, and thus to be partners with courts in implementing the law in a coherent fashion, while not as well-established as the agency’s duties of reasoned elaboration, is steadily gaining recog­nition. This analysis highlights the rule-of-law foundation for that duty.

With respect to notice principles, more groundbreaking work is required. Some of it will take the form of holding agencies to the basic principles of notice, as current scholarship has done with regard to the fundamental value of publicity in agency rules. Further work could also form an executive or judicial requirement for agencies to justify their decision when they opt not to proceed through rulemaking, a departure from long-settled law. It could also usefully involve embracing or even imposing a duty upon agencies to issue, in the form of guidance, their best general statement of the law’s requirements when rulemaking is not practicable.

Perhaps the most controversial analysis pertains to the principle of authorization. Both principles—decisional allocation within the executive branch and courts giving weight to agencies’ views of the scope of their own authority—have waxed and waned in terms of their embrace in the law. Today these positions, at least based on intimations from the Supreme Court, may be on the wane. If so, there is all the more reason to highlight the ways in which officials conceive of their statutory obligations as personal anchors and reinforces the government’s commitment to the law. And once so conceived, the grounds for recognizing that independent review—whether for agency officials or courts—does not require es­chewing respectful consideration of the positions of other government officials become all the stronger.

Conclusion

Law provides benefits to society but also poses risks. Some view those risks as amplified when courts or administrative agencies conceive their roles too broadly. Professor Strauss views those risks as amplified when courts and agencies conceive their roles too narrowly. This assessment is not fundamentally grounded in an expansive view of the size of the state, but rather in the scope of law’s demands on judicial and administrative agents. For Professor Strauss, agencies and courts have arduous duties. They are tasked with making sense of the issue before them while resolving it in a way that integrates it into the larger fabric of law; this frequently requires considering the intelligibility of statutory law, its relationship to other law, and the current context. The exercise of that duty also requires justification and engagement with those affected. Because the duties of government are personal, they create a system of accountability—accountability through the repeated reliance on individual judgment. Recognizing duties of that wide scope may be part of what enables a society to accommodate change without abandoning its best structure. 145 See Strauss, Common Law and Statutes, supra note 121, at 255 (second alteration added) (“[W]hat we mean by law . . . [is] [t]he process by which a society accommodates to change without abandoning its fundamental structure.” (internal quotation marks omitted) (quoting Grant Gilmore, The Ages of American Law 14 (1977))).