The third bound.

• Author: Vermeule, Adrian
• Position: Symposium: The Bounds of Executive Discretion in the Regulatory State

INTRODUCTION I. EXAMPLES A. Removal B. Directive Power II. CONVENTIONS AS THE THIRD BOUND A. Neither Law Nor Politics B. Intragovernmental Versus Extragovernmental Conventions C. Conventions About Doing Things Versus Conventions About Saying Things CONCLUSION INTRODUCTION

Our topic is the "bounds on executive discretion." A number of the articles assume, implicitly or explicitly, that there are two principal bounds: law and politics. Yet another subset of the articles--a partly overlapping subset--illustrates, implicitly or explicitly, that there is also a distinct third bound on executive discretion: conventions, roughly understood as unwritten but obligatory rules of the political game. (1)

I will argue that conventions pervasively shape and constrain executive discretion and are an indispensable tool for understanding the issues discussed in the articles. Debates among legal academics over executive discretion misfire if and when the role of conventions is overlooked or misunderstood. In particular, legal debates over executive discretion should take account of three distinctions: (l) between contingent politics and conventions; (2) between intragovernmental conventions and extragovernmental conventions; and (3) between conventions against doing things and conventions against saying things. The last distinction, in particular, illuminates the strong resistance, in contexts such as immigration, to executive policy statements that make explicit a pattern of enforcement discretion, one that would otherwise remain implicit. Even holding legal authority constant, making that authority explicit through general policy statements may trigger the normatively inflected political sanctions that are characteristic of conventions.

I do not at all mean to claim that executive-constraining conventions are necessarily desirable from a welfarist standpoint--they may or may not be. Nor do I claim that they cannot be overridden or broken, or anything like that. Conventions have a dimension of weight, often have implicit override conditions, and may be more or less fragile. Indeed, part of the story is that as the political parties have become more polarized over time, a number of intragovernmental conventions resting on reciprocal cooperation have broken down, and the politics of executive action have become both more legalized and more political--in the ordinary convention-independent sense of "political." But I will claim that the role of conventions cannot safely be neglected in any analysis of executive discretion.

1. EXAMPLES

   Let me begin with some examples of conventions that shape and constrain executive discretion. There are also conventions that shape and constrain the behavior of other branches, and indeed the government as a whole, but I will focus strictly on executive-governing conventions in the United States constitutional order. In parliamentary systems lacking a formal and institutional separation of executive power from other powers, of course, the whole question of categorization would have to be approached differently.

   1. Removal

      I will begin with some examples of conventions surrounding the removal of executive officials. In the Free Enterprise Fund litigation, the Supreme Court and the lower courts treated the Commissioners of the Securities and Exchange Commission as though they have for-cause tenure. (2) Legally speaking, however, they simply do not. The relevant statutes say nothing about tenure, and thus fail to override the longstanding background default principle that for-cause tenure must be express. (3) There is a limited exception in which the Court has been willing to imply for-cause tenure for administrative officers who are in essence mini-judges exercising purely adjudicative functions, (4) but the SEC does not fit that description. The independence of the Commissioners, in the sense of for-cause tenure, is not based on a statute but on unwritten rules with normative force. As one lower court put it, it is just "commonly understood" that the Commissioners are independent. (5)

      Somewhat similarly, there is a powerful convention that Presidents cannot fire the Chair of the Federal Reserve except for serious cause. (6) Here, too, there is no legal rule to that effect, but it is a universally agreed-upon rule of the political game--agreed upon in a tacit sense. (7) The most telling evidence is that when Presidents have attempted to maneuver a Fed Chair out of office, they have proceeded covertly and circumspectly, so as not to outrage public opinion. (8)

      In 2007, President George W. Bush failed to renew the terms of seven United States Attorneys, and public opinion rose up in outrage. (9) Bush's action was indisputably within his legal power; the relevant statutes do not grant United States Attorneys for-cause tenure, and indeed the Court ruled early and clearly that no such tenure exists. (10) Yet a network of conventions that limited presidential control over U.S. Attorneys had grown. (11) As described by Mary Jo White, herself a former U.S. Attorney, the President may replace the whole set en masse, at the time of a partisan change of administration, but may not engage in selective replacement during the middle of a presidential term. (12) The rationale for the convention is rather obviously to protect the independence of U.S. Attorneys through a form of herd immunity or safety in numbers, preventing the President from singling out particular prosecutors based on their decisions. (13)

   2. Directive Power

      Now let me turn to conventions governing the so-called "directive authority," raised by Coglianese and Firth in their article. (14) Although the scope and limits of the directive power of the President are among the most contested issues in administrative law, (15) the debates would be greatly improved by the recognition that conventions pervasively shape and constrain that power. The President's legal powers of direction are substantially greater than his actual discretion to direct, and not only because "politics" is an additional constraint--unless we define politics, unhelpfully, to include conventions. I return to the definitional issues shortly, but let me offer some examples.

      There is a clear and powerful convention that restricts presidential direction of agency adjudication, especially in formal proceedings. (16) What makes this convention particularly striking is that it governs even where the department head, herself subject to presidential direction, would otherwise have legal authority to intervene. As Elena Kagan put it, "The only mode of administrative action from which [President] Clinton shrank was adjudication. At no time in his tenure did he attempt publicly to exercise the powers that a department head possesses over an agency's on-the-record determinations." (17) The convention, then, is definitely a constraint on the directive power of the President as President.

      Many executive-branch lawyers have spoken of unwritten conventions protecting the independence of the Office of Legal Counsel (OLC). (18) In the most high-minded version of this account, OLC is supposed to give the President entirely independent legal advice and to act as an impartial legal arbiter in interagency jurisdictional disputes; the President would violate convention by directing OLC to give one opinion or another. (19) There is good reason to think that the high-minded picture is partly aspirational, although there have been important cases in which OLC did actually contradict the President's wishes. (20) It is not that the conventions are unreal, but they have probably weakened over time, and were always subject to bending or breaking when political pressure became sufficiently great. (21)

      In a different example, Michael Livermore and Ricky Revesz discern a tradition of appointing relatively independent voices to the position of Office of Information and Regulatory Affairs (OIRA) Administrator, and opine that the tradition would likely be difficult to break. (22) The cash value of this tradition seems to be that the OIRA Administrator has leeway to make regulatory policy and to apply the cost-benefit framework of relevant executive orders, without fear or favor. (23) Although Livermore and Revesz phrase the "tradition" in terms of appointment, it would seem, a fortiori, that replacing an OIRA Administrator on openly political grounds would probably be even more likely to provoke convention-based political sanctions.

2. CONVENTIONS AS THE THIRD BOUND

   1. Neither Law Nor Politics

      Whatever the details, the relevant point for our purposes is that conventions are a tertium quid, neither law nor politics, at least not necessarily so. They are not "law," or not necessarily "law." The classical British view was that law and conventions are entirely distinct; "law" was by definition enforceable in court, whereas conventions were not. The more nuanced modern view in the Commonwealth is that although conventions may not be enforced in court, they may be "recognized" and used as an aid to interpreting statutes or deciding other legal questions. In our system, the "law" may or may not be enforceable in court, and although it is possible for legal rulemakers to adopt conventions as legal rules, they do not have to be so adopted, so conventions may or may not be enforced in court. (24)

      When will conventions be adopted as law? All else equal, courts are more likely to read conventions into open-ended or ambiguous texts with unclear originalist referents. As Justice Jackson urged, "Usage may sometimes impart changed content to constitutional generalities, such as 'due process of law,' 'equal protection,' or...