No Nondelegation at the Founding? Not so fast, by Ilan Wurman
Julian Mortenson and Nicholas Bagley have posted a provocative and thoughtful new paper making the claim that there was no nondelegation doctrine of any kind at the founding. These two authors are careful as usual, and in several places they raise concerns and arguments that may require modifications to existing originalist claims about nondelegation. But their evidence does not quite add up to their ultimate claim: that the founding generation did not adhere to any nondelegation principle at all. In this post, I offer tentative thoughts on a few of the potential flaws in their interpretation of the evidence. I am particularly grateful for the opportunity once again to engage Professor Mortenson in his work on the separation of powers. (For my paper in partial response to his two recent and excellent articles on the meaning of “the executive power,” see here.)
At the outset, it’s important to establish correctly the existing originalist claims. The standard originalist position is that there are certain kinds of things that Congress must do and the executive (or judicial) branch may never do, namely the formulation of rules regulating private conduct, i.e. telling private people (as opposed to government officials) what they can and can’t do or altering their rights or obligations. Yet Mortenson and Bagley often describe the originalist position as being that any “rulemaking” is an exercise of legislative power that cannot be delegated. I know of no originalist who actually holds such a view, and I encourage Mortenson and Bagley to remove the several characterizations to the contrary. (For example, p. 21: “First, the critics have argued that rulemaking is an exercise of legislative power that may not be delegated by the legislature. Second, they insist that rulemaking can’t qualify as an exercise of executive power, which is limited to the particularized application of existing rules.”)
The paper’s central problems, however, relate to its interpretation of the evidence of “what the Founders said” and “what the Founders did.” This post challenges the paper’s interpretation of “what the Founders said” on two counts: their understanding of nonexclusive powers, and the distinction between delegation and alienation. It then argues that the paper’s evidence of “what the Founders did” does not prove what Mortenson and Bagley think it does. Their evidence from the First Congress does, I think, establish that the modern originalist “private conducts/private rights” nondelegation test might have to be modified—and here their paper contributes the most to the scholarship and originalists must take its claims seriously. But the evidence does not prove there was no nondelegation doctrine at all.
Nonexclusive Powers. Starting with the evidence of what the Founders said, Mortenson and Bagley first make much of the fact that exercises of government power are “nonexclusive,” and that the Founders understood that such power could be characterized as either legislative or executive. It is, however, a commonplace that some actions are legislative if done by the legislature and executive if done by the executive. The act involved in the INS v. Chadha case supplies a familiar example. When done by Congress, deciding that certain individuals should or should not be deported was a legislative act, which is why it required bicameral passage and presentment. If done by the INS, it would be mere law execution. It is therefore somewhat surprising that Mortenson and Bagley appear to think the point is novel.
Consider also the following example, which I have written about elsewhere and which Mortenson and Bagley discuss in a different context. One of the very first statutes required the new national government to make the payments to the disabled veterans of the revolutionary war that had been authorized by the Confederation Congress, “under such regulations as the President of the United States may direct.” Act of Sept. 29, 1789, ch. 24, 1 Stat. 95, 95. President Washington and Secretary Knox promulgated a regulation—a “rulemaking,” if you will—stating that the payments shall be made in two equal installments and requiring affidavits as evidence of injury and entitlement to payment. (This early regulation may be viewed here.) Was this something Congress could have done? Of course. This kind of rulemaking is “legislative” in the sense that Congress could have adopted it on its own, or some variant of it. But it is not the kind of regulation that Congress has to be the one to decide. Such a regulation could also be considered executive. Congress had decided all the important questions—that the disabled veterans should be paid pensions, and the amount of those pensions. Washington’s rulemaking was entirely about implementation. This power is partly legislative, yes, but also partly executive—indeed, sufficiently executive that it’s acceptable for the executive to make them. I have previously described the exercise of this rulemaking authority as “nonexclusive legislative power”—Congress may but need not exercise this power itself.
Mortenson and Bagley accurately summarize the Founders’ understanding of “nonexclusive” powers: “Depending on the relationships you focused on, a given act could properly be classified as both legislative and executive at the same time.” (p. 41). “To the question of whether a given act involved an exercise of executive or legislative power, the answer was often both.” (p. 5). These statements are absolutely correct. But note the word: often. The existence of nonexclusive powers does not mean that every exercise of governmental power is nonexclusive, yet that’s what Mortenson and Bagley proceed to assume. They go on to conclude that because all legislative and executive power is “nonexclusive,” there is no problem “delegating” power because as soon as Congress has authorized the executive to do something—anything—any exercise of power pursuant to that authorization is characterized as “executive” power.
In arguing that the Founders understood all legislative and executive power to be “nonexclusive,” they rely on Madison’s statements on the indeterminacy of language and of the line separating legislative and executive power. “The [orginalists’] mistake,” they write, “comes in assuming that executive rulemaking can only be described as an exercise of executive power. To the contrary, sophisticated discussions from the founding recognize that efforts to classify government action in the abstract are irreducibly indeterminate.” (p. 39). For this they principally rely on Madison in Federalist 37 on indeterminacy, but ignore Madison in the nation’s first nondelegation debate, over post roads: “However difficult it may be to determine with precision the exact boundaries of the legislative and executive powers, he [Madison] was of opinion that those arguments were not well founded, for they admit of such a construction as will lead to blending those powers so as to leave no line of separation whatever.” 3 Annals of Cong. 238 (1791). And in the Constitutional Convention, Madison argued that “certain powers were in their nature Executive, and must be given to that departmt.” 1 Farrand 67. Perhaps some in the founding generation thought efforts to classify government action was “irreducibly indeterminate,” but Madison was not among them.
In short, Mortenson and Bagley make a lot of the founding generation’s understanding that power was often nonexclusive. But they haven’t shown that the Founders believed that all power was nonexclusive, nor have they demonstrated that all exercises of government power are in fact nonexclusive. Indeed, Chief Justice Marshall assumed there was a difference between exclusive legislative power—which could not be delegated—and nonexclusive legislative power, which Congress could exercise but could also delegate: “It will not be contended that Congress can delegate to the Courts, or to any other tribunals, powers which are strictly and exclusively legislative,” but “Congress may certainly delegate to others, powers which the legislature may rightfully exercise itself.” Wayman v. Southard, 23 U.S. 1, 42–43 (1825). Mortenson and Bagley point to no evidence that Marshall was wrong.
Delegation versus Alienation. Another error that Mortenson and Bagley appear to make relates to their distinction between delegation and alienation. One of the more interesting parts of Mortenson and Bagley’s paper is how it potentially rediscovers (pp. 29-36) what might have been a founding-era distinction. Delegation is allowing another person or body to exercise authority, but that authority may always be reclaimed by the principal; alienation is transferring and giving up one’s power altogether, after which that power cannot be reclaimed. Summarizing the general rule, “the founding generation saw nothing untoward about provisionally delegating the power to make rules so long as Congress did not permanently alienate its power to make laws.” (p. 3). They argue (pp. 29–30) that Locke’s Second Treatise, often cited by proponents of the nondelegation doctrine, is talking about non-alienation, not non-delegation, when Locke writes that the legislature may not “transfer” the legislative power “to any other hands.”
Yet Mortenson and Bagley’s argument here depends on the equivalence of the word “transfer” with the word “alienate.” In footnote 111 they point to examples where Locke appears to use the words to mean similar things. But Locke also uses the word transfer to mean “delegate.” For example, in section 135, Locke writes,
[The legislature] is not, nor can possibly be absolutely arbitrary over the lives and fortunes of the people: for it being but the joint power of every member of the society given up to that person, or assembly, which is legislator; it can be no more than those persons had in a state of nature before they entered into society, and gave up to the community: for no body can transfer to another more power than he has in himself; and no body has an absolute arbitrary power over himself, or over any other, to destroy his own life, or take away the life or property of another.
Here, Locke talks about the transfer of power from the people to the legislature when they leave the state of nature. Yet Mortenson and Bagley’s entire argument about Locke is that he was arguing that the people delegated their power to the legislature and did not alienate their power, as absolutists like Jean Bodin had argued (pp. 31–32). And now we see that in section 135, when Locke is talking about this original delegation (not alienation), he uses the word “transfer.” In section 135, then, “transfer” means delegation, not alienation.
Even if the distinction were theoretically correct, it does not do much work in practice, for two reasons. First, Madison and John Randolph, anyway, argued that “delegating” legislative power—delegating in the sense in which modern originalists talk about delegation—would in fact be an alienation of legislative power. Put another way, what originalists describe today as “nondelegation” is what at least some in the founding would have referred to as non-alienation. Here again is Madison in the first “nondelegation” debate in Congress, over the post roads, where the question was whether Congress should give the postmaster general the authority to decide where the post roads should be: “there did not appear to be any necessity for alienating the powers of the house—and that if this should take place, it would be a violation of the Constitution.” 3 Annals of Cong. 239 (1791).
Why might Madison have thought that “delegating” the power to decide where the post roads should be would be an “alienation” of legislative power? The simplest answer would of course be that the founding generation did not in fact see a distinction between the two terms. But even if they did, a delegation of legislative power to the executive would still be an “alienation.” Could such a power be reclaimed after it has been given to the executive, if the executive could veto any future attempt to reclaim such a power? It would be possible, yes, but exceedingly difficult. Here is John Randolph in opposition to an 1803 law which gave President Jefferson complete power to make laws and regulations for the Louisiana Territory: “If we give this power out of our hands, it may be irrevocable until Congress shall have made legislative provision; that is, a single branch of the Government, the Executive branch, with a small minority of either House, may prevent its resumption.” 13 Annals of Cong. 498 (1803). Delegating legislative power to the executive might effectively be alienation.
Yet even if not, all Mortenson and Bagley show is that the people could delegate their sovereign powers to agents but could not transfer (alienate) it. What about the agent’s ability to subdelegateor “redelegate”? That is the central question—whether the people’s agents (Congress, the executive, the judiciary) can redelegate the power given to them by the people, and to whom they might be able to redelegate it. Here, however, the paper’s evidence is extraordinarily thin.
Mortenson and Bagley write, rather cursorily, that the executive could clearly redelegate its power by using subordinate magistrates. (p. 26). But that’s an entirely different matter. There is ordinarily no problem when an officer subdelegates power to a subordinate, who is under his supervision and control, because the officer continues to have the real power. Re-delegation is only a problem when power is transferred to someone whom the original delegatee cannot control. When Congress delegates power to the executive, it forfeits control. That may or may not be constitutionally permissible, but it gains no support from analogy to the accepted practice of administrative subdelegation.
Mortenson and Bagley’s other evidence is not much better. As evidence of “redelegations,” they point to Parliament or the Crown delegating legislative authority to the colonies (p. 28), the states delegating legislative authority to the Continental Congress (p. 28), and later on congressional delegations to territories (pp. 68–69). These involve circumstances far removed from the concerns of modern proponents of the nondelegation doctrine. These all involved one government (King-in-Parliament, the states, Congress) delegating legislative authority to a subordinate government (colonial authority, Continental or Confederation Congress, or the territorial governments). Delegating to the executive—where the executive with a small minority in either House may prevent a resumption of any legislative power by Congress, and where the executive is not an agent of or subordinate to Congress—is simply not the same thing as delegating to a subordinate entity.
Evidence from the First Congress. Moving to the evidence of “what the Founders did,” consider three possible conceptions of “nondelegation.” The first conception is the one advocated by many originalists today: any rule governing private conduct or altering private rights is “legislative.” The second conception is the one that Mortenson and Bagley seem to advance: there was no limit on what Congress could delegate. A third conception is Chief Justice Marshall’s in Wayman v. Southard, which is similar to the first conception but more nuanced: there is a category of exclusively legislative power over “important subjects,” but also a nonexclusive power “to fill up the details.” Important subjects on which Congress must legislate might include the formulation of rules of private conduct, and might not include matters involving governmental employees or managing public lands, for instance.
In the last part of Mortenson and Bagley’s paper, where they go through the legislative activity of the First Congress, they demonstrate to my satisfaction that first conception might be too strong of a claim. I’ve tentatively said as much in prior work (see nn. 173–187 and accompanying text here). But they don’t prove the second conception—that there was no limit on congressional delegation at all. Indeed, their evidence is just as consistent with the third conception—with an “important subjects” approach to nondelegation in which private rights and conduct play an important but not necessarily determinative role. Because they have dozens of examples, I cannot explore them all here, but here is a sampling.
Mortenson and Bagley spend a number of pages (pp. 77–79) describing a variety of military pension statutes granting regulatory authority to the President. I mentioned one such statute previously. In that statute, Congress decided all the important subjects—that the disabled veterans shall be paid, and how much. Those were the matters, perhaps not incidentally, that determined the rights of private persons. The President then merely had to decide when the payments should be made (the statute required they be made within one year; President Washington chose two equal payments three months apart) and what proofs would be necessary. To be sure, some statutes appear to have given the President even more discretion, setting for example only upper limits to the pension amounts. But even here most statutes were quite specific, leaving only administrative details to the President—those that didn’t involve the more important questions of whether or how much veterans should be paid.
The paper also describes (pp. 81–82) delegations in the maritime context, including statutes authorizing port-of-entry collectors to let inspectors “examine the cargo or contents” of ships and authorizing them to direct inspectors “to perform such other duties according to law . . . to perform the better securing the collection of the duties.” Another law authorized collectors to conduct searches and seizures when they were “suspicious of fraud” or “suspected a concealment.” In none of these statutes, Mortenson and Bagley argue, “did Congress lay down any meaningful guidance about the circumstances in which ships ought to be searched or the type of evidence that ought to make collectors think that fraud or smuggling was afoot.”
These statutes hardly prove what Mortenson and Bagley seem to think they do. Did any of them delegate to customs officials the power to decide what items shall be subject to a duty? Did any delegate them power to decide tariff rates? Did any delegate to them the power to decide whether fraud was prohibited or not? No—the statutes resolved all these important questions involving the private rights and obligations of private individuals, and even specified the more important of the means that would be used by government officials in executing the law such as the searches and seizures described above. Any subsequent regulation by a collector would merely have been an instruction to other government officers as to how to conduct their law-execution functions. It’s hard to imagine what more Congress could have been expected to do.
On naturalization, Mortenson and Bagley write (p. 83): “Under a 1790 statute, Congress gave to ‘any common law court of record’ the authority to grant U.S. citizenship to any free white persons who had lived in the country for two years after ‘making proof to the satisfaction of such court, that he is a person of good character.’” If this is a delegation of legislative power, then I’m not sure what isn’t. Every statute would be a delegation of legislative power because it authorizes the other two branches to do something. The statute did not authorize a court to decide what types of people should be made citizens, whether there should be naturalization at all, what residency requirements there should be, or anything else of the sort. Congress decided all the important questions, admittedly leaving some discretion in applying the law. But no originalist claims that the existence of discretion in law-application, which is inevitable, suggests an unlawful delegation of legislative power.
Mortenson and Bagley also rely (pp. 75–77) on a statute prohibiting intercourse with Native American tribes without a license from the executive branch, and giving the executive complete discretion to decide whether, to whom, and why to grant such licenses. This was indeed a broad statute that delegated authority to regulate private conduct. But it is also a delegation in the context of the President’s treaty and commander-in-chief power: the idea was to avoid conflict and violence with Native American tribes. As Michael McConnell argues in a forthcoming book, it may have been understood at the time of the founding that Congress had more power to delegate old royal prerogative authorities back to the President than it had power to delegate other legislative powers. And Philip Hamburger has written (Is Administrative Law Unlawful? pp. 104–05) that this licensing scheme may have been justified because it governed behavior outside the domestic territory. I am not sure either view is right, but at a minimum the special context of this delegation militates against drawing any general conclusion from it.
I found the most interesting example to be the first patent statute (p. 73). Congress’s statute authorized the granting of patents that were “useful” and “important.” Clearly this leaves a lot of discretion, and eventually the executive officers charged with enforcing the law came up with rules clarifying these two standards (I’m not sure the rules were ever formally promulgated—Mortenson and Bagley are not clear on this point). Mortenson and Bagley write that Jefferson explained that several rules had been “established by the board,” for example that a patent would not issue for a change in the application of an earlier invention. This is certainly a general rule that alters the rights of private persons as opposed to official conduct. (A number of originalists have proposed to me that perhaps there is a further distinction here between “private rights” and public rights/privileges, like pensions, but I am not sure that should matter for nondelegation purposes.) Such a rule in the patent context might suggest that the executive did make at least some rules specifying the details of more general legislative provision and these rules could and did sometimes affect private rights and conduct. Although Mortenson and Bagley don’t get this far ahead in history, in the 1830s steamboat legislation Congress authorized the making of rules imposing passenger limits on ships and rules for the passing of ships—rules that clearly would have altered private rights and obligations in at least some ways.
Yet, that some originalists might be wrong about their particular test for nondelegation does not prove that there were no limits on delegation at all. The First Congress did not come even close to testing that proposition—its statutes were not nearly as broad as Mortenson and Bagley seem to think they were, it rarely authorized the creation of rules that actually altered private rights and obligations (if it did so at all), and when it might have done so the rules did not alter rights or obligations in any significant way or the rules were in the context of presidential powers.
In short, I lose no sleep over any of the statutes that Mortenson and Bagley unearth. But what of the following hypothetical statutes? “Any common law court shall decide who shall be a citizen, for whatever reason the court sees fit to declare someone a citizen.” Or, “The patent office shall decide whether the United States government should give patents, the term of years, and the grounds on which to grant patents.” Or, “The President may issue regulations carrying into effect any of the powers vested in Congress in Article I, Section 8.” Mortenson and Bagley’s paper does not prove that these kinds of statutes would have been permissible. It does not demonstrate that there was no nondelegation principle at the founding. It proves, at most, that originalists might need to think more about the limitations of their current “private conduct” theory of nondelegation.
Ilan Wurman is a visiting assistant professor at the Sandra Day O’Connor College of Law at Arizona State University, where he teaches constitutional law and administrative law. He is the author of the book A Debt Against the Living: An Introduction to Originalism. You can follow him on Twitter (@ilan_wurman).