Deference and Determination

In his remarkable and underappreciated book on Natural Law in Court, Dick Helmholz observes that when moderns turn to the subject of the relationship between natural law and positive law, they immediately focus on the question whether the former in some sense “trumps” the latter in cases of irreconcilable conflict. To the classical lawyer, however, that question was not central. To be sure, as early as the Institutes of Gaius it was said that “changes to civil law can take away rights belonging to civil law (jus civile), but not rights belonging to natural law (jus naturale)” (I.158). But natural law had many other roles detailed by Helmholz, roles much more central in actual practice, such as supplying interpretive principles and default rules for construing statutes, supplying principles of just procedure, and suggesting remedies.

More broadly, for the classical lawyer, the whole framework within which to discuss the relationship between natural and positive law was different, centering on their complementary roles rather than on potential conflict. The positive law, the ius civile, was understood as a set of rational ordinances promulgated by the public authority for the common good —that is, in order to give more specific content to the general principles of the natural law. In a famous passage, Aquinas distinguished two ways in which positive law might be derived from the natural law:

It must be noted, however, that something may be derived from the natural law in two ways: in one way, as a general conclusion derived from its principles; in another way, as a specific application of that which is expressed in general terms. The first way is similar to that by which, in the sciences, demonstrated conclusions are derived from first principles; while the second way is like that by which, in the arts, general ideas are made particular as to details: for example, the craftsman needs to turn the general idea of a house into the shape of this or that house. Some things are therefore derived from the principles of the natural law as general conclusions: for example, that `one ought not to kill’ may be derived as a conclusion from the principle that `one ought not to harm anyone’; whereas some are derived from it as specific applications: for example, the law of nature has it that he who does evil should be punished; but that he should be punished with this or that penalty is a specific application of the law of nature. Both modes of derivation, then, are found in the human law. Those things which are derived in the first way are not contained in human law simply as belonging to it alone; rather, they have some of their force from the law of nature. But those things which are derived in the second way have their force from human law alone. (Emphases added).

This public “determination” (determinatio) of the natural law, making it more concrete in application, has two important features. First, “those things which are derived in the second way have their force from human law alone.” They only become law because of the command of the legitimate public authority. In this sense, the central role of the positive law is embedded within the determinatio framework as a special case. Classical legal theory is sometimes opposed to positivism, but this is misleading. The latter is merely the myopic view that part of the framework of classical law should be broken out and treated as though it were the whole of law.

Second, the public authority fits the general precepts of the natural law to varying local circumstances; as Aquinas puts it, “[t]he general principles of the natural law cannot be applied to all men in the same way because of the great variety of human circumstances; and hence arises the diversity of positive laws among various people.” This means that determination will inevitably be discretionary within bounds; in the limit, it may even call for a decision of the sort I have elsewhere called “rationally arbitrary.” The basic purpose of making general precepts of the natural law concrete in particular circumstances is like a vector; it sets a broad direction for the public authority, but does not say exactly how far to go, without further specification. Always subject to the nature of lawmaking in the classical conception — reasoned ordination to the common good — there is an irreducible discretion to choose. And that discretion, Aquinas suggests elsewhere, may be exercised through the kind of inarticulate or intuitive judgment that the artifex uses. As John Finnis puts it, referring to the standard image of an architect chosen to build a school or hospital for the well-ordered city, “[t]he kind of rational connection that holds even where the architect has wide freedom to choose amongst indefinitely many alternatives is called by Aquinas a determinatio of principle(s)—a kind of concretization of the general, a particularization yoking the rational necessity of the principle with a freedom (of the law-maker) to choose between alternative concretizations, a freedom which includes even elements of (in a benign sense) arbitrariness” (my emphasis).

I want to suggest that determinatio is an excellent lens for thinking about Chevron deference, Auer deference, and other such structures even within the positive law. There are two distinct points here. One is strictly analogical: determinatio describes not only the relationship between natural law and positive law, but also the relationship between higher and lower levels of positive law, such as statute on the one hand and administrative regulation or adjudication on the other. On this account, agency action often determines — gives concrete form to — general principles laid out in statutes or agency legislative rules (binding unless and until changed through valid rulemaking procedure). This is close, theoretically, to the point that executive action is a “completion power” that carries general legislative commands into execution. Agencies may even make rationally arbitrary choices within the scope of the broad boundaries of the general statutory or regulatory principle.

And this account also has implications for the scope of judicial review. Precisely because agency action often takes the form of determination of general statutory principles, agencies are often in the position of architects carrying out a commission whose broad principles have been set by Congress. Judges should thus afford agencies leeway to carry out the task of the architect. They should allow agencies to specify policy within the bounds and for the purposes of the overall statutory task, even when the agency, choosing between alternative ways of making the statute concrete, cannot offer reasons for preferring one specification to another. Remarkably, the Supreme Court more or less endorsed exactly this picture by holding, in FCC v. Fox Television Stations in 2009, that an agency charged with giving specificity to a general statutory limit on “indecent” broadcasting could change its policy without adducing reasons to show that its new policy was better than the old. Barring issues of reliance interests or a change of specific factual findings, an agency charged with giving form to a broad, general or ambiguous statutory authority is entitled to simply show that its (change of) policy amounts to a valid specification. Deference flows from determination.

The other way of cashing out determinatio is somewhat more ambitious. So far we have been discussing cases in which agency action makes more concrete higher sources of binding positive law, such as a statute or the agency’s own prior binding legislative rule. Perhaps agency rules and adjudications can also, sometimes, be seen as ways in which agencies themselves supply specifying content to the natural law directly, rather than through the medium of some earlier act of positive lawmaking. Agencies, that is, could be taken to have the authority to invoke general principles of the natural law as starting points for their own actions (always assuming they otherwise have legal authority to take action of that kind, such as authority to make rules or to adjudicate).

Two obvious doctrinal contexts come to mind. First, subject to minimum legal requirements in statutes and the Constitution, agencies typically have broad discretion to make procedural rules for their own rulemaking and adjudication (whether at the level of administrative law judges or the agency itself). Agencies may thus draw upon principles of natural law in order to supply general principles for formulating procedures and for interpreting their statutory and constitutional obligations. Even where, for example, background requirements of constitutional due process in adjudication do not require a pre-deprivation hearing, there is nothing that prevents an agency from affording one anyway, as a determination of the natural procedural presumption in favor of hearings. Helmholz explains that such a presumption was deeply embedded in the ius commune: “Justice required that every person affected by litigation be cited and heard by an impartial judge. The ability to speak in defense of one’s own person and property was a right that could not be denied to any person consistently with the law of nature.” The natural-law presumption could be overcome by a sufficient showing that the common good required it, but it would be a perfectly valid starting point for the agency’s reasoning, one to which courts should defer. (I have argued at length elsewhere that a deferential framework should apply to constitutional due process cases).

A second doctrinal context involves agency consideration of the “relevant factors” under arbitrariness review, as laid out in the seminal Overton Park decision. It has always been slightly mysterious where, exactly, the relevant factors come from, putting aside factors whose consideration Congress has either mandated or barred. The basic principle is that discretionary factors must be logically relevant to the agency’s decision, leaving open the question of their source. Under the conception of determination laid out here, general principles of natural law should be straightforwardly eligible as “relevant factors.” An example might be EPA v. EME Homer, a complex air regulation case in which statutes left open what principle EPA should choose to allocate responsibility for interstate air pollution among states. The Court observed that the principle EPA chose was both efficient and “equitable … because, by imposing uniform cost thresholds on regulated States, EPA’s rule subjects to stricter regulation those States that have done less in the past to control their pollution.” EPA here drew upon a natural principle of justice, that responsibility should at least presumptively lie where fault lies, to make concrete a general statutory command.

The connection between deference and determination obviously extends to other contexts within and without administrative law. Thayerian deference in constitutional judicial review, for example, is easily susceptible of a similar analysis. I will not belabor the point more than I already have, but simply observe that dim echoes or professional memories of determinatio in the ius commune may well form part of the deep background matrix of legal principles out of which modern deference doctrines arose.

Adrian Vermeule