Can the Legislative Power be Delegated to the Administrative State?

James R. Stoner, Jr. is Hermann Moyse Professor and Director of the Eric Voegelin Institute for American Renaissance Studies and in the Department of Political Science at Louisiana State University.

In this week’s “Constitutional Conversations” post, James Stoner argues that the legislative power, power granted to the Congress by the people through the Constitution, cannot be delegated to administrative agencies.  Professor Stoner grounds his argument in both the common law and the separation of powers delineated in Article I of the Constitution itself. While Stoner suggests that the administrative state might have usurped the legislative power in some cases, he also suggests that a delegation of legislative power is an abdication of responsibility on the part of the institution entrusted with the responsibility to make laws on behalf of the people and can be recovered if Congress has the will.

            Can the legislative power be delegated?  No, of course not: According to our constitutional principles, the legislative power has been delegated by the American people to Congress, and it is an old maxim of law that “potesta delegata non potest delegare,” that is, a delegated power cannot be delegated—or translated into modern American English, “if you borrow your parents’ car, don’t lend it to a friend.”  Since that maxim was in Latin originally, you can suppose that it can be traced back to Roman law, where it figured in private law, as in the example of the car, and apparently it was also applied in a judicial setting, since a Roman praetor could delegate some of his cases to his clerk, but the clerk could not delegate further to an assistant.

But wait!  What makes the maxim about not delegating the legislative power itself to be law?  That, I think, is where we have to begin if we want to understand why the question of delegating the legislative power to the administrative state has become controversial in modern America and if we want to sort out what ought to be done about it.

            Why, then, are maxims like the one against delegating delegated power and others such as “no one ought to profit from his own wrongs” or “no one ought to be a judge in his own case” themselves law?  They are thought to be basic maxims of justice and thus have long been incorporated into common law, that is, the unwritten, customary law, originally inherited by the American colonists from England, that governs property, exchange, injuries, human relationships, and once even basic criminal matters, as disputes about these things are settled in courts of law, courts that themselves follow due process as defined by common law, including its characteristic mode of trial by jury.  Common law was expounded or developed by following another maxim of justice, that similar cases ought to be similarly decided, and therefore judges at common law follow precedents, inferring a rule from the decision in an earlier case and then applying it in the next one that is similar.  As similar cases are not identical, there is an element of reasoning by analogy as the common law judge moves from case to case; indeed, in hard cases, the issue is often which precedent to follow, that is, which analogy to draw.  Modern cynics say common law is judge-made law, but traditional common-law judges thought they were finding law—the appropriate precedent, the most reasonable analogy—not making law.  Legal historians might trace the original announcement of a rule at common law, but that doesn’t mean they have found its origin; to the common lawyer, the law unfolds according to reason and develops into what one seventeenth-century authority called “the artificial perfection of reason.”

            Now common law can be changed by statute—in fact, it’s a maxim of common law that a statute can override a contrary rule at common law, though the first task of the judge interpreting a statute is to determine whether it declares what was already common law or remedies some mischief that arose in it.  In a common-law environment, then, the legislative power is first and foremost the power to declare or to change the law.  Originally the American colonists inherited from England not only the common law but an unwritten constitution; at the Revolution, they put their new constitutions in writing, but they reaffirmed by declaratory statutes the common law, at least so much of it as was consistent with New World circumstances and unchanged by their statutes—and these statutes often declared in writing some of the basic practices of common law, especially those associated with due process.  In these constitutions, and explicitly in the federal Constitution, the legislative power was said to be delegated by the people to the state assembly or to Congress, as the origin of political power is the consent of the governed—consent being presumed of customary law but inferred through representation in law that is newly made.  The people delegating legislative power and vesting it in representative bodies, that power could not be further delegated without violation of that maxim with which I began.  Indeed, as Columbia law professor Philip Hamburger has argued, the vesting clauses that begin the first three articles of the federal Constitution anchor the separate powers, legislative, executive, and judicial, and I think it is probably correct to say this means that unless specifically provided for in a checking mechanism—for example, the veto, a power legislative in kind but given to the executive in a qualified way, or the impeachment powers given to the House and the Senate, powers executive and judicial in kind but given to Congress for extraordinary use—each branch was intended to exercise only the kind of power given to it.

            But here’s the catch: in the exercise of authority, at least any rational exercise of authority, one cannot help making rules, if only implicitly, since the reason something is done suggests a rule for the future in similar matters.  It might be merely accidental whether one proceeds to formulate a rule, the accident being how frequently the officer encounters the same situation.  In simple times or in rare and extraordinary events such as going to war, the government’s choice is an exercise of prudence, a particular judgment concerning a unique or extraordinary set of circumstances. By contrast, if similar situations are faced again and again, acting justly might seem to require the development of rules.  Take the instance of a tax assessor.  The legislature might pass a tax on real property prorated according to its value, and in a community of a few farms and estates, a judgment could be made about the value of each one, for after all, each would probably sell at its own price.  But if one is considering row houses in a city or single-family homes in a suburban development, it seems likely if not strictly necessary that the assessor would develop rules, factoring in for example the number of rooms, the square footage, the size of the lot, and so forth.  Or take the simple academic example of grading papers: In a seminar, the experienced professor can confidently judge the quality of each, but in larger class one might rely on a grading rubric or assign exercises like multiple-choice tests rather than essays so the grading can proceed by a rule.  In a small setting, one wouldn’t doubt that the assessor or the teacher is a sort of executive official, but insensibly as the stock of property or the size of the school grows, rules emerge.  When along this path of development was the legislative power—originally exercised in passing the tax or establishing a school—delegated?

            Of course, it wasn’t.  In the early cases concerning administrative or non-legislative rule-making, the United States Supreme Court recognized this process and didn’t consider it to be delegation, or unconstitutional delegation, of the legislative power.  In the 1825 case of Wayman v. Southard, Chief Justice John Marshall recognizes that Congress established in the original Judiciary Act a federal rule for judicial procedure: The federal courts in each state were to follow legal practice in that state in 1789.  It was nevertheless understood that judges could adjust practice in their courts as they found necessary, but not that changes by state legislatures be automatically incorporated into federal law as if Congress had delegated its legislative power to state legislatures.  In 1928 in Hampton v. United States, Chief Justice Taft sustained a congressional statute that allowed the president to adjust tariff rates within a set range if he found it necessary to equalize commercial advantage for American firms being undersold by foreign competitors.  The tariff act had established an “intelligible principle” and the executive or administrative rules added only a finding of fact—and the sort of fact that varies too much in its particularities to be covered by a general rule of law.  These decisions seem correct to me, and the rule-making they countenance seems legitimate and non-legislative: on the one hand, based on rules the judiciary would develop about their own courtrooms, on the other, based on changeable facts that an official in the midst of a market could ascertain and adjust easily enough but not necessarily a legislature like Congress.

            But the Court also, I think, was correct in cases like Panama Refining and Schechter Poultry in 1935 to strike down the wholesale grant of power from Congress to other agencies to regulate entire industries without clear principles, defining at least the outer limit of delegation.  The more recent cases seem much less clearly established, as the Court has developed a tangled jurisprudence that generally defers to agency rule-making through various recondite forms of deference—Chevron deference, concerning agencies’ interpretation of their authorizing statutes, or Auer deference, concerning their interpretation of their own regulations—and various forms of scrutiny, as for example the emerging “major questions” doctrine.  Although in principle constitutional rights are protected in court against contrary administrative action, the proliferation of claims of rights seems sometimes to mean that even long-established rights at common law—such as the right to educate one’s children as one thinks best or the right of bodily integrity—get swept aside when faced with the regulations of education and health bureaucracies.  It is easy to understand why in areas of policy as diverse as immigration, environment, and health, people across the political spectrum—or maybe I should say, people within each party when government is in the hands of their opponents—sense that governance is increasingly in the hands of the administrative state rather than subject to laws made by the people’s representatives.  What is to be done?

            First, I think that courts need to be vigilant in the protection of rights (constitutional, common law, or statutory) and not presume that a generally expressed good in a statute authorizing an agency and sketching its purpose (e.g., protecting the environment or insuring workplace safety) overrides and was meant to override claims of rights.

            Second, I think that the proper administrative role involves finding facts, not making policy, and that fact-finding should recover a sense of the distinction, nicely sketched by Hamburger, between settled expertise and emergent science.  Partisanship aside, I would think anyone could acknowledge that unprecedented measures were taken in the past three years in the name of “following the science,” while any genuine scientist will concede the provisional character of most scientific hypotheses, not to mention acknowledge the multiplicity of sciences.  Practical thinking has to weigh scientific claims where evidence is disputable, and the judgment of which experts to consult and when to consult them is profoundly practical or prudential, not easily settled by bureaucratic structures or a decision algorithm or, for that matter, by legislation.

            Third, “arbitrary and capricious” review under the terms of the Administrative Procedure Act seems to me in need of being tightened.  Whatever the merits of the unitary executive—and there are good reasons, originalist and consequentialist, in its favor—the obviously partisan character of what is sometimes touted as expertise ought to be recognizable and suspect.  The stretching of sensible statutes to implement partisan policy objectives seems to me objectionable.  In other words, what is needed is confinement of agencies to their statutory purposes, not permitting regulations wholly ancillary to their main charge; in short, courts need to revive the old doctrine of ultra vires.  Do emergencies allow actions outside the beaten path?  Of course, but routine declarations of emergency are a contradiction in terms and need to be quashed.

            Finally, if the legislative power has indeed been delegated, the principal solution has to be for the legislature itself to recover it.  Actually, that’s almost the nature of the legislative power, for it is, in terms of the history of political theory, the youngest power, courts having existed through all time, the executive since the Renaissance, the legislative only since the English Civil War and the theory of John Locke.  Both in Locke’s account and in that common-law definition I suggested at the outset, legislation involves the correction of abuses or mischiefs that were not adequately addressed under the existing state of affairs.  The administrative state was created to address real social problems; where it has itself created greater problems than it was created to address, Congress has it within its power, if not necessarily its political will or capacity, to alter or abolish its creations and to institute new forms.