There is No Historical Precedent for Prioritization in a Debt Ceiling Crisis

 by Neil H. Buchanan and Michael C. Dorf

Conor Clarke, who will shortly begin his appointment as a faculty member at the Washington University School of Law (and whom we shall accordingly, albeit slightly prematurely, refer to as Professor Clarke) recently posted on SSRN a fascinating draft article titled The Debt Limit. In it, he traces the history of the current debt ceiling statute to origins that long pre-date what is often said to be the first debt ceiling statute in 1917. As he explains, the original purpose of the debt ceiling statute currently codified at 31 U.S.C. § 3101 had nothing to do with fiscal restraint, as Republicans now claim for the current version.

Rather, in prior years Congress had not enacted a general authorization for the Treasury to borrow money, instead passing borrowing authorizations to accompany specific appropriations bills. The debt ceiling statute, Prof Clarke says, was part of a legislative package that obviated the need for new borrowing authority. It was intended to expand, not restrict, the government's ability to borrow. In the current era, he concludes, it serves no useful purpose and therefore should be repealed. Needless to say (because we've said it many times before), we agree with that conclusion.

Despite sharing our policy views about the debt ceiling, Prof Clarke disagrees with our analysis of the situation the President would face (quite possibly within a week) should Congress fail to raise the debt ceiling sufficiently to cover the gap between tax revenues and appropriations. His chief argument is rooted in historical practice.

In The Debt Limit, Prof Clarke recounts a great many interesting details about the pre-1917 experience. He summarizes some of them in a recent essay on The Volokh Conspiracy, where he also links to  popularizations of academic writing and promises additional efforts along those lines for broader public consumption. Both the longer article and the short summary on Volokh endorse what we have called the conventional wisdom: in a debt ceiling crisis, the President should prioritize some appropriations over others, paying bondholders before everyone else and then making discretionary judgments about which bills to pay (and when and to what extent) from the leftover funds on hand. He rejects our view that the exercise of such enormous executive discretion without clear congressional authorization or any statutory guidance would be a much worse usurpation of legislative power than would be simply issuing bonds in excess of the debt ceiling. As Prof Clarke says on Volokh, we "get it wrong on the constitutional 'trilemma.'"

Why? Perhaps Prof Clarke has been holding in reserve a secret killer argument to show how and where we are wrong. However, as we explain below, what he has written so far provides no grounds at all for thinking we're wrong and arguably even reinforces our position.

As we said, the main value added of the Clarke article concerns historical practice. But Prof Clarke also makes an important point regarding accounting with which we want to start. He says that tax revenues will be more than sufficient to cover debt service, even if the government runs out of borrowing authority due to hitting the debt ceiling. Thus, he says, there's no real risk of default--in the sense of the government failing to pay principal and interest owed to lenders on time and in full. We agree with that factual claim about government revenues, as far as it goes, but we don't think it goes very far.

Let's begin with Section 4 of the Fourteenth Amendment. It doesn't use the word "default." It says: "The validity of the public debt of the United States, authorized by law, including debts incurred for payment of pensions and bounties for services in suppressing insurrection or rebellion, shall not be questioned." Putting aside the matter of what it means to "question" the validity of the public debt, Prof Clarke's argument has bite against the Fourteenth Amendment proposition only if one takes a very narrow view of what constitutes "debt." The only case on the subject, Perry v. United States, gestures towards a pretty broad definition, but perhaps there are grounds for thinking "public debt" includes only debt to lenders and not debt to contractors, employees, and others, much less money appropriated for benefits programs. Yet the history of public finance that Prof Clarke recounts sheds little light on the original public (or contemporary) meaning of "public debt" as used in the Fourteenth Amendment.

Nonetheless, and purely for the sake of argument, let's assume that Prof Clarke is right that tax revenues suffice to cover whatever government bills come due that constitute "debt" within the meaning of the Fourteenth Amendment. If so, then he has shown why there is a version of prioritization that would allow the President to pay everyone who is owed money that counts as such debt in full and on time through incoming tax revenues, without the need to borrow in excess of the debt ceiling. Even so, that conclusion does not cast any doubt on the Buchanan/Dorf trilemma argument.

Recall--as we've repeated too many times to mention--that our argument does not depend on the Fourteenth Amendment. Rather, our core point is that if Congress authorizes the collection of taxes totaling X, borrowing totaling Y, and appropriations totaling Z > X + Y, then there's no way for the President to comply with all of the sets of laws. Worse, non-compliance means violating the Constitution. The President has no authority to tax or borrow without Congressional authority. And absent specific statutory authorization and guidance, spending less than Congress has appropriated is as much a usurpation of legislative power as spending more.

As we have previously elaborated, that last point is the central lesson of the impoundment controversy during the Nixon administration. The House considered impeaching President Nixon for impounding (i.e., not spending) appropriated funds, concluding that his doing so was unlawful but uncertain that it was an impeachable offense. After Nixon left office, the Supreme Court, in Train v. City of New York, affirmed what had been the lower court consensus already: that absent specifically delegated discretion, a president must spend all of an appropriation. And the House did impeach a President--Donald Trump in 2019--for "abuse of power" by not releasing "$391 million of United States taxpayer funds that Congress had appropriated on a bipartisan basis for the purpose of providing vital military and security assistance to Ukraine." Citing the Line Item Veto case, a GAO report subsequently confirmed that Trump acted unlawfully, indeed unconstitutionally, because "[t]he Constitution grants the President no unilateral authority to withhold funds" that Congress has appropriated.

And that makes sense as a matter of first principle. The notion that the President--without any specific authorization or guidance from Congress--can pick and choose which of the hundreds of billions of dollars to spend and which to withhold would, as we have argued at length before, be a gross violation of separation of powers, moving the power of the purse from Congress to the President. At the most fundamental level, that is why we conclude that the President's proper course in a debt ceiling crisis is to exercise no discretion by continued borrowing of exactly the amounts needed to cover the gap between revenues and appropriations. Indeed, the current codifications of the statutory borrowing authorities (chiefly here, here, and here) contain exactly that limit: each provision authorizes the Treasury Secretary to borrow "amounts necessary for expenditures authorized by law." That limitation is what renders the separately codified and thus easily severable debt ceiling provision superfluous as a limit on borrowing.

So why does Prof Clarke nonetheless think that prioritization is not merely less unconstitutional than borrowing in excess of the debt ceiling but lawful? He offers two main arguments, neither of which we find remotely persuasive.

(1) Prof Clarke says that the government has engaged in prioritization during government shutdowns and more broadly when Congress hasn't appropriated enough money to accomplish some task. He goes on to acknowledge that prioritization to satisfy the debt ceiling "is different as a matter of degree . . . because it would require prioritizing at the level of government spending as a whole, rather than prioritization within a single spending program" and that this "is not a trivial difference." Nonetheless, he thinks "that unilateral debt issuance is" not "the obvious practical choice."

Note that Prof Clarke doesn't say here that we're wrong to think violating the debt ceiling is the least unconstitutional option. He says it's not obviously the most practical option.

Yet even that timid conclusion is wrong. Of course it's more practical on June 2 (or whatever the X date ends up being) for Treasury to do exactly the same thing it has done on previous days: sell bonds. Meanwhile, for the government to prioritize among billions of dollars of spending across the entire federal budget entails not only the exercise of enormous legislative-style discretion but decisions about numerous practical details about timing, distributive effects, and much more.

And in any event, the core comparison is also wrong. If Congress says to the President spend $X to accomplish some task but it would really take $Y > $X dollars to fully accomplish the task, it's pretty clear that Congress has told the President to accomplish as much of the task as can be accomplished using $X. When the President prioritizes among sub-tasks but still spends no more or less than $X, he has not usurped one iota of Congress's spending authority. Put differently, prioritization to satisfy the debt ceiling is not, as Prof Clarke asserts, different in degree from prioritization during a government shutdown or other appropriations shortfall; it's different in kind.

(2) Prof Clarke also turns to historical practice. As we said above, this is the most interesting and novel aspect of his article. We learned a lot from it and are genuinely grateful for his historical research and lively prose. However, the history does not in any way undermine our position.

The main contribution of Prof Clarke's historical work, as we see it, is the observation that long before 1917, Congress enacted caps on how much money the executive branch could borrow. In some instances, these caps were part of isolated project finance: Congress would authorize a bond issue of some amount, with the funds dedicated to some specific purpose. But in other instances, Congress would grant general borrowing authority capped at some specific dollar amount. We agree with Prof Clarke that the limits in such borrowing authorizations are closely analogous to the current debt ceiling statute.

But, not to put too fine a point on it, so what? Prof Clarke says that "[t]he Executive Branch appears to have respected th[e] limits" on borrowing in the particular authorizations Congress enacted during the pre-1917 period. Well, sure it did, just as every President since 1917 has respected the debt ceiling. Why? Because the borrowing authority was, in each instance, adequate to pay for mandated appropriations. Prof Clarke cites not a single episode in which a President prioritized otherwise mandated spending in deference to the cap on borrowing. Maybe there was such an episode, but we suspect that if there was, Prof Clarke would have found it. And the fact that everybody describes the possibility of hitting the debt ceiling as unprecedented pretty strongly suggests that Prof Clarke's research was exhaustive.

If Prof Clarke or someone else were to find some hitherto unknown episode of a President prioritizing spending when the debt ceiling or one of its predecessors became binding, we would need to consider the weight of that (by definition obscure) episode as measured against the awesome power the President would be usurping were he to follow it as precedent. But as matters stand, because there appears to have been no such episode, Prof Clarke's historical research does not call for any sort of reevaluation.

Prof Clarke does describe what he aptly calls "two particularly delightful vignettes" from the Washington administration, but these vignettes also provide no support for prioritization in a debt ceiling crisis.

In 1789, Treasury Secretary Alexander Hamilton borrowed $191,608 to pay for government expenses, even though Congress had not authorized such borrowing. Prof Clarke says that Hamilton went to Congress "with an air of excuse and apology." We don't detect an apology in Hamilton's report (which mentioned the borrowing as the third item in what was mostly a request for appropriations), but that's a quibble. Even so, we would make two further observations.

First, if the episode proves anything, it's that there is precedent for unilateral borrowing in times of what Hamilton called "necessity." Prof Clarke says that Hamilton "acknowledge[ed] that it wouldn't happen again." That's an overstatement. As the language Prof Clarke quotes indicates, Hamilton said that it would be proper, "in future cases," for Congress to authorize any borrowing by statute. Hamilton did not say that considerations of necessity--to avoid committing an even larger usurpation of legislative authority and mitigate a global economic catastrophe, say--could never create a similar necessity if Congress fails to do what is proper.

Second, even if we were to assume that Hamilton's statement that Congress should authorize borrowing was somehow an acknowledgment that the unilateral borrowing was wrong and a promise binding on all future administrations never to do it again, that would have no bearing on a debt ceiling crisis. Hamilton did not borrow more than a cap Congress had placed on authorized borrowing. He borrowed without any congressional authorization at all. If the vignette proves anything (and we don't think it does), it proves a proposition that is self-evident in Article I, Section 8: that the Constitution assigns the borrowing power to Congress. That's the starting point of the debate, not any sort of basis for an argument within it.

The other Washington administration vignette Prof Clarke cites involves a statute that authorized borrowing at 5 percent. The actual debt was issued at 5 percent, but additional fees totaling 4.5 percent arguably brought it outside the scope of authorization, so Congress chided the administration. However, as Prof Clarke observes, Congress also ratified the loan on grounds of expedience.

What any of that has to do with prioritization or the debt ceiling is mysterious. Prof Clarke does not say or even hint that the Washington administration gave a de facto interest premium beyond what Congress authorized in order to meet spending obligations. And of course, if the administration had done so, that would be further precedent for the proposition that, in extremis, violating the limits on a borrowing authorization is preferable to failing to execute expenditure laws. Meanwhile, the fact that Congress ratified the loan despite the hand-wringing further indicates that necessity can forge authority. If President Biden issues debt in excess of the debt ceiling, this vignette will be precedent supporting Congress in subsequently ratifying it.

* * *

To be clear, our view that borrowing in excess of the debt ceiling is, from a constitutional perspective, much preferable to prioritization, does not in any way depend on historical precedents. We do find it illuminating, however, that the one article that dives deeply into the relevant history not only fails to uncover a historical gloss that contradicts our view. The historical materials, insofar as they are relevant, tend to support it.