ARTICLE CONTENTS INTRODUCTION I. INSIDE THE AGENCY A. Agencies Are a "They," Not an "It" B. Allocating Power Within Agencies: Constitutional Constraints II. HORIZONTAL ALLOCATION A. Judicial Review 1. Chenery 2. Chevron 3. Mead 4. Mead: The Intersection of Chevron and Chenery 5. Cost-Benefit Default Rules 6. "Hard Look" Review 7. The Administrative Law of Emergencies B. Structure and Process 1. Agency Design 2. OIRA Review 3. Separation of Functions 4. Litigating Authority III. VERTICAL ALLOCATION A. Judicial Review 1. Mead Redux 2. The Accardi Principle 3. Massachusetts v. EPA 4. Substantial Evidence: Universal Camera and the Morgan Cases 5. State Secrets Privilege B. Structure and Process 1. Delegations of Authority to Particular Officials 2. Freedom of Information Act, Deliberative Privilege Exception 3. Managing and Controlling Adjudication IV. TRADEOFFS AND IMPLICATIONS A. Horizontal Allocation Decisions B. Vertical Allocation Decisions C. Empowering Courts Empowers Agency Lawyers D. Spillovers E. Agency Choice Between Rulemaking and Adjudication CONCLUSION INTRODUCTION
Perhaps the main topic in administrative law is the allocation of power among legislatures, courts, the President, and various types of agencies. Theorists usually justify their preferred allocations by reference to some conception of expertise, politics, or legalism. Promoters of independent agencies appeal to expertise; promoters of presidential supervision of the bureaucracy appeal to political accountability; promoters of expansive judicial review of agency action appeal to legalism. In all of these standard debates, the main issue is the allocation of power across institutions; agencies are typically treated as unitary entities.
In this Article, we will examine how administrative law allocates power within agencies and how arguments from expertise, legalism, and politics apply inside agencies rather than across institutions. Although commentators have sometimes cracked open the black box of agencies to peer inside, their insights are localized and confined to particular contexts. (1) We will generalize the idea, attempting to show that administrative law allocates power both horizontally and vertically within agencies and offering some hypotheses about the nature of the resulting effects. Horizontally, administrative law directly and indirectly determines the relative influence within agencies of various professionals--lawyers, scientists, civil servants, politicians, and others. Vertically, administrative law directly and indirectly determines the relative influence within agencies of appointed agency heads, lower-level bureaucrats, and line personnel.
This perspective illuminates several of the most puzzling principles and doctrines of administrative law. Among them are the Chenery principle that agency action cannot be upheld in court on the basis of post hoc rationalizations; the Chevron doctrine, which gives deference to reasonable agency interpretations; the Mead doctrine, which amends Chevron by giving agencies more deference if they use more procedural formality; and the Accardi (or Arizona Grocery) principle, which requires agencies to follow their own rules until duly changed. In each of these cases, we will suggest, one of the main effects of the relevant doctrine is to allocate power within agencies-not just among agencies, courts, and other actors. That perspective helps to improve upon, and to critique, existing justifications for the doctrines. Administrative law involves not only doctrines of judicial review but also agency structures and procedures, which are usually established by statute or executive order. Structure and process, we will claim, also have important direct and indirect effects on the allocation of power within agencies.
In offering these claims, our aims are twofold: first, we outline a theoretical framework for understanding how legal rules might affect the allocation of power within agencies; second, we propose a series of hypotheses about the actual allocation effects of administrative law rules. Although we provide anecdotal evidence and insider testimony where it is available, there is very little in the way of systematic empirical work about the questions we discuss. This state of empirical uncertainty cuts neither for nor against our claims. The current regime of administrative law itself rests on unarticulated and unproven suppositions about the internal design of agencies and the effects of law inside agencies. The only implication of this state of uncertainty is that it demands a new research agenda for empirical administrative law--one that should take account of the internal allocation effects of legal rules, among other matters.
Part I identifies the various stakeholders within agencies and the constraints that constitutional law places on the allocation of power among those stakeholders. Part II examines rules, principles, and doctrines of administrative law that affect horizontal allocation within agencies. Part III does the same for vertical allocation. Finally, Part IV generalizes the examples to state some general tradeoffs; the largest tradeoff is between the twin goals of allocating power in desirable ways across institutions and allocating power in desirable ways within agencies. A brief conclusion follows.
INSIDE THE AGENCY
Agencies Are a "They," Not an "It" (2)
Even casual observers of the administrative state recognize that agencies, like nearly all large organizations, are not unitary actors. They are fractured internally. At a minimum, many agencies have the following stakeholders: political appointees, civil servants, front-line decisionmakers, and policy professionals (including attorneys, economists, public policy analysts, or scientists). (3) These are not mutually exclusive categories: one can be a policy professional and a civil servant or a political appointee. One can think of these stakeholders along at least three dimensions: (1) the nature of their selection and tenure (political appointees, civil servants); (2) their professional training and orientation (lawyers, economists, scientists, budget specialists); and (3) their place in the hierarchy of the agency (front-line decisionmakers, top-of-the-heap policymakers).
Several types of stakeholders often will have decisive or crucial authority when the agency acts. "Decisive" and "crucial" are intended to capture a broad range of contributions to agency decisionmaking, including cases where the relevant actor is, formally, the ultimate decisionmaker (such as the top political appointees), cases where the actor has a right to have her views considered or deferred to by others by operation of law, agency rule, or custom, or cases where an actor's input into the decision is inevitably part of the process of agency decisionmaking. There are important differences here, and those differences translate into different levels of influence over given decisions, but only the general (and simple) point is relevant for present purposes. The views and actions of different types of stakeholders shape the agency's performance of its duties.
To illustrate, imagine that the Food and Drug Administration seizes an adulterated drug. Each of the following types of actors is likely to have a significant hand in shaping the overall course of the agency's enforcement action: a front-line enforcement agent who develops the facts and executes the seizure; her supervisors in the regional office and perhaps central agency (who may even have authored a manual instructing enforcement agents on how to conduct seizures); the agency attorney who advises on the enforcement action and defends the agency if the action is challenged in court; and the attorneys, civil servants, and political appointees who help decide whether to seek to appeal if there is an adverse ruling against the government. Enforcement actions in many agencies are routine and initiated at the lower levels of the agency. Even with respect to these types of decisions, a variety of different types of stakeholders will have an influence over the course of the agency's action. The point is even more obvious with less routine decisions. An agency's high-stakes rulemaking or adjudication is likely to involve significant input from civil servants, lawyers, scientists, economists, and political appointees.
These differing stakeholders are likely to disagree, at least sometimes, about the right course for the agency. (4) The conflicts between political appointees and the "bureaucracy"--usually taken to refer to the well-insulated-from-termination members of the professional civil service--are legion. Those at the lower rungs of the agency hierarchy are likely to have different views about proper enforcement strategy than those at the higher rungs, for example. (5) And policy professionals regularly disagree, not only (and predictably) with political appointees, but with other policy professionals as well. Robert Katzmann's well-known study of the Federal Trade Commission describes the varying worldviews of the lawyers and the economists at that agency and the way in which their conflicts worked out. (6)
There is much more to be said about the dynamics of the relationships among these agency stakeholders and (more importantly) about their consequences, but the basic points are simple: agencies contain identifiable constituencies that affect policymaking, and these constituencies can, and do, come into conflict over the proper functioning of the agency.
Allocating Power Within Agencies: Constitutional Constraints
Our aim is to offer some hypotheses about the ways in which factors external to the agency shape the relationships among the agency's internal stakeholders. We will explore three such factors: constitutional law, judicial review of administrative action, and the structure and process of agencies. As the next two Parts detail, these outside forces allocate power...