Reviewability and the 'law of rules': an essay in honor of Justice Scalia.

• Author: Vermeule, Adrian
• Position: Antonin Scalia

INTRODUCTION

In Washington v. Trump, (1) the Ninth Circuit decision on the Administration's executive order on immigration, (2) a critical moment occurred when the panel had to explain why the order was reviewable at all. Precedents like Kleindienst v. Mandel (3) had said that "when the Executive exercises [immigration authority] on the basis of a facially legitimate and bona fide reason, the courts will [not] look behind the exercise of that discretion." (4) The panel's response was to invoke an important distinction between the reviewability of general rules, on the one hand, and the reviewability of specific applications, on the other:

[T]he Mandel standard applies to lawsuits challenging an executive branch official's decision to issue or deny an individual visa based on the application of a congressionally enumerated standard to the particular facts presented by that visa application. The present case, by contrast, is not about the application of a specifically enumerated congressional policy to the particular facts presented in an individual visa application. Rather, the States are challenging the President's promulgation of sweeping immigration policy. Such exercises of policymaking authority at the highest levels of the political branches are plainly not subject to the Mandel standard. (5) As we will see, this conception of reviewability (6)--keyed to a distinction between general policies or rules and specific applications, with the former subject to review even if the latter would not be--is one that Justice Scalia wrestled with throughout his reviewability jurisprudence. Yet it is more or less directly the opposite of Justice Scalia's own conception. Justice Scalia developed a consistent approach to questions of reviewability: roughly, the idea that "general programs" and "general policies" are to be excluded from judicial review, and even general and legally binding agency rules may or may not be reviewable before enforcement. On this approach, the proper business of courts is to review specific applications of agency rules to particular parties. Notice that there are actually three possible subjects of review in play here: nonbinding policies and rules (such as "interpretive" rules); binding general rules ("legislative rules"); and applications. More on this shortly. For now, the focus is on the distinction between review of general and specific agency action, whether that action is embodied in a legally binding rule or in a nonbinding policy.

The approach featured in Washington v. Trump, by contrast, was propounded in several opinions by Justice John Paul Stevens. On that view, agencies should be more afforded more, not less, discretion to apply policies or rules in particular cases, whereas the proper business of the judiciary is to review the general legality of overall programs, policies, and rules. The Scalia view and the Stevens view obviously differ on the proper role of courts in an overall system of administrative law. For Justice Stevens, the role of courts is to say what the general rules of law are, leaving agencies (reasonable) discretion in application. For Justice Scalia, the role of courts is to decide cases, reviewing the legality of rules only insofar as necessary to that function--as a byproduct of deciding cases.

Furthermore, the two approaches differ as to the relationship between reviewability and the idea--perhaps Justice Scalia's most famous contribution to legal theory--that the rule of law is best understood as "a law of rules." (7) On the Stevens view, reviewability should attempt to ensure that courts review the overall legality of programs and policies to keep the administrative state within the broad bounds of the rule of law. (8) On that view, the "rule of law" is a "law of rules" in the sense that judicial power to say what the law is entails a power to examine the legality of general rules.

On the Scalia view, by contrast, the rule of law is a law of rules in a quite different sense. The judicial role is to review particular "cases and controversies," and the law of rules is a constraint on the sorts of reasons, grounds, or rationales courts may use to decide such cases. The rule of law requires courts to articulate grounds of sufficient generality that their disposition of particular cases is not arbitrary or unreasoned. The grounds of judicial decisionmaking must not display an excessively ad hoc, case-specific character. This conception underpins Justice Scalia's hostility to balancing tests across many areas of law.

Crucially, nothing in this conception is inconsistent with a conception of reviewability that is narrower than the Justice Stevens conception. The Justice Scalia conception admits into court only agency action at the point of specific application, but then demands that judges analyze the validity of that specific application under general rules. For Justice Stevens, the rule of law requires expansive judicial review of rules; for Justice Scalia, the rule of law, qua law of rules, operates primarily as a constraint on the types of rationales courts may give for their decisions of particular cases, once those cases are already in court. Both conceptions are internally consistent, but they are also opposites.

What difference does all this make? To see the main doctrinal difference between these two conceptions of reviewability, we have to clarify a tripartite distinction: between or among (1) review of nonbinding policies and rules, (2) review of legally binding general rules, and (3) review of specific applications. As Washington v. Trump explicitly indicates, and as Justice Stevens indicates as well in several decisions, one conception allows review of nonbinding rules and policies that are otherwise sufficiently final and definite to be reviewable. (9) These nonbinding rules and policies may or may not have been formulated with valid statutory authority, adequate reasoning, and so forth.

On Justice Scalia's conception, by contrast, review of policies is never acceptable; broad policies and programs, as we will see, do not amount to reviewable agency "action." Agency applications of binding rules to specific cases, as in enforcement actions, are the paradigm of reviewable action for Justice Scalia. The hard case, then, involves the intermediate category (2): review of general binding rules, for example notice-and-comment rules, before they are applied in particular cases. In other words, the hard case involves the Abbott Laboratories v. Gardner (10) question: When and to what extent are agency rules reviewable before they are enforced against particular parties?

On the Justice Stevens conception, of course, the answer is straightforward: pre-enforcement review is valid. Justice Scalia, by contrast, was reluctant to countenance pre-enforcement review, which threatens to undo his distinction between general programs or policies and specific applications. Justice Scalia, however, was constrained by the existence of adverse precedent, and generally acceded to pre-enforcement review, insofar as necessary. In that sense, consistent with his overall approach to adjudication and precedent, his treatment of reviewability was "faint-hearted." (11)

1. TWO CONCEPTIONS OF REVIEWABILITY--AND OF THE JUDICIAL ROLE

   1. Review of Rules

      Agencies typically (although not necessarily) formulate policies and programs premised on some conception of their legal authority, the boundaries of their legal discretion; they enact general rules in pursuance of those programs; and they enforce the rules in particular cases. Given some commitment to preserving the rule of law, somehow defined, at which point or points in this sequence should judicial review occur? Here too, I will focus for the moment not on the threefold distinction among policies, rules, and applications, but rather on the distinction between general (policies and rules) and particular (applications). The differences among policies, rules, and applications will be explored later, when we discuss pre-enforcement review.

      In the abstract, and stated at a very general level, two different approaches to reviewability are possible. On one view, the important thing is to ensure that the programs, policies, and rules of law under which the agency is proceeding are valid as formulated on their face. If agencies are formulating policies and rules that are generally invalid, there is a serious problem, one that it is imperative for courts to oversee and correct. If agencies are applying valid policies incorrectly in particular cases, that is unfortunate, but less grave.

      On this approach, the application of policies and rules to particular cases is less important to review, for several reasons. First, the stakes are inherently cabined and confined. The application of policies and rules in particular cases can legally bind (if at all) only in the case at hand, whether or not it has defensible precedential effect in other cases. With respect to rules, the stakes are lower even if the rules are not formulated antecedently in legislative style, through notice-and-comment proceedings, so that the agency qua adjudicator formulates rules as rationales in the process of application. In that situation, rules function as precedents, but are still not so binding as legislative rules; rules as precedents can be overturned in future cases (if reasons can be given for doing so), or can be distinguished away, whereas legislative rules are binding even upon the agency itself, unless and until changed through the same notice-and-comment process that created them.

      Second, application of policies and rules to cases is mediated by agency determination of particular facts ("adjudicative facts"). Judicial review of adjudicative facts has always been deferential, especially in on-the-record proceedings where the "substantial evidence" test applies. (12) The difference, then, between a finding of reviewability and a finding of...