Conor Clarke[1]*
Abstract
Every couple of years, it seems, the debt limit shows up to wreak havoc in American law and public finance. By capping the face value of government securities that can be “outstanding at one time,” the statutory limit regularly threatens the Treasury’s ability to raise the revenue needed to fund required government spending. Brinksmanship over the limit has shut down the government, cost the country billions of dollars, and mired financial markets in uncertainty. And yet, despite its obvious and longstanding importance, the limit remains poorly understood. Commentators attribute its beginnings to 1917 and 1941, before which it is assumed that Congress itself designed each debt instrument and did so only during rare emergencies. For some, the presumed recency of the limit makes it easy to dismiss. It is regarded as a quixotic provision that can and should yield to more significant spending mandates, or a provision that is itself unlawful—at odds with the Fourteenth Amendment’s cryptic command that public debts “shall not be questioned.” For others, the limit is regarded as either a cataclysmic threat to the financial system, or a necessary hindrance to runaway government spending, or both. We are warned, by politicians and commentators alike, that a binding debt limit will lead to default, and the cratering of the American economy—and also that having no limit would be an “irresponsible” concession to our profligate spending instincts.
But each piece of this conventional wisdom is wrong. Debt limits—authorities for the Executive Branch to borrow that come with limits attached—have existed since 1790, and flow naturally from the Constitution’s reservation of the borrowing power to Congress. I provide a corrective account of those early limits and draw on public laws and Treasury borrowing records to provide an overview of the Executive Branch’s borrowing authority between 1790 and 1910. That historical excavation has important doctrinal and policy implications for how we think about public finance today and helps clear the myth and confusion surrounding the debt limit. Under current doctrine, the limit is lawful. It is a form of statutory direction and commitment that was common at the ratification of both the Constitution and the Fourteenth Amendment. The limit is binding. When the limit conflicts with spending provisions, longstanding practice suggests that it is spending—and not the limit—that must yield. And, finally, the implications of the modern limit remain woefully misunderstood. There is no good reason to think that default follows from a binding limit: tax revenue is more than sufficient to cover debt service. But there is excellent reason to think that the modern limit has become so divorced from its original appropriations purpose—which was to make spending cheaper, not harder—that the case for reform is ripe.
On January 13, 2023, Treasury Secretary Janet Yellen wrote to congressional leaders with urgent news: the following week, the Treasury would hit the statutory debt limit of almost $32 trillion.[2] Because federal spending relies heavily on borrowing—we spend more than we tax—hitting the limit threatens to deprive the government of sufficient revenue to cover its costs. Once the Treasury hit the limit, Yellen reported, the Department would take so-called “extraordinary measures,” such as deferring federal retirement investments, to finance government operations as long as possible.[3] But such measures would be only temporary. By June 5, Yellen later reported, the government would run short of money to pay for its obligations.[4]
On June 3, two days before that crisis moment and after months of negotiations, the showdown over the limit was resolved. President Biden signed into law a deal that suspended the debt limit in exchange for spending caps and other legislative concessions.[5] But this was not the first time that the debt limit had provoked an extended showdown between the branches and the Republican and Democratic parties, and it is unlikely to be the last. Since the debt limit was enacted in its roughly contemporary form in 1941, the federal government has skirted close to the debt limit and disrupted the ordinary financial operations of the Treasury dozens of times—starting in 1953 and 1957, and continuing regularly in each decade since.[6] In 2021, two years before the most recent crisis, the federal government again came within days of running short of cash to fund the government.[7] The June 2023 legislation that ended the most recent showdown made no effort to resolve the issue permanently; it suspended the limit only until the beginning of 2025.[8]
Indeed, the debt limit may be the most reliable source of legal and financial crisis that the American government faces. The American Treasury is the biggest borrower in the world, and the debt limit is the primary statutory constraint on that dimension of American public finance.[9] Failed negotiations over the debt limit have been responsible for numerous government shutdowns, including most recently in 2013, and have cost the government tens of billions of dollars.[10] In 2011, the country’s credit rating was downgraded for the first time, with Standard & Poor citing debt-limit brinksmanship for “America’s governance and policymaking becoming less stable, less effective, and less predictable than what we previously believed.”[11] And, each time the Executive Branch approaches the brink, it encounters an unusual legal conundrum: what actually happens—if not what must happen—when a statute seems to prevent the government from paying its own bills? President Biden has asserted recently that the limit would be unconstitutional, citing the obscure Public Debt Clause of the Fourteenth Amendment.[12] The conflict between the spending obligations and the debt limit has also been described as presenting a kind of constitutional “trilemma,” in which the Executive Branch must choose between three unenviable options: triage spending and (so it is claimed) usurp Congress’s spending power, or raise additional funds and thereby violate either the taxing power or borrowing power.[13]
The statutory limit thus raises subtle questions that are central to America’s fiscal constitution.[14] And yet, despite that ongoing importance, the debt limit remains poorly understood—and, in certain key respects, misunderstood. The conventional story of the limit always starts in either 1941—when Congress created the statutory limit on outstanding debt that persists to this day—or in 1917, when, at the height of the First World War, Congress expanded and combined certain borrowing authorities in the Second Liberty Bond Act.[15] Before then, it is typically asserted that Congress itself designed and approved individual issuances of debt to finance specific spending projects—rare authorizations that were supposedly limited to wartime and other emergencies—and that there was nothing comparable to the modern debt limit.[16]
But that conventional understanding is wrong. Before 1917, scores of public laws dealt with debt and debt issuance, and the great majority of those issuance statutes authorized and constrained the Executive Branch in a manner that is not different in kind from the legal structure of modern borrowing. Congress authorized the Executive Branch to borrow within limits—typically giving the Executive Branch discretion over the specific terms of that borrowing, and typically doing so to fund the general expenditures of government. Authorizations of this form go back quite literally to the First Congress. Consider the first statutory debt limit. The law making appropriations for the year 1790 laid out specific appropriations for the year and then concluded with the following broad authority: “That the President of the United States be authorized to empower the Secretary of the Treasury, if he shall deem it necessary, to make such loans as may be requisite to carry into effect the foregoing appropriations . . . .”[17] This broad delegation is strikingly similar to contemporary federal borrowing authorities, which provide that, “[w]ith the approval of the President, the Secretary of the Treasury may borrow on the credit of the United States Government amounts necessary for expenditures authorized by law . . . .”[18] Between 1790 and the present, Congress—unsurprisingly—used a variety of different mechanisms to simultaneously grant and limit the Executive Branch’s authority to issue debt. Piecemeal developments occurred along the way. But statutes simultaneously authorizing and limiting Executive Branch debt issuance have as ancient a pedigree as anything in American statutory law. The story of the debt limit starts in 1790, not 1917.
That story, and the emergent practice between the branches, has relevance for how we conceive of federal public finance today. If the debt limit is but a recent misadventure—or at least a twentieth-century one—it is easy to dismiss, as much existing scholarship does, as “an appendage that was added to the system long after the federal government began operating successfully,”[19] and a “ruinous and arbitrary determinant of government policies.”[20] But 230 years of practice between the branches suggest a richer and more complicated picture. It may be that the debt limit has now become a kind of ruinous appendage, but it is an appendage that has its origins in deep features of our constitutional structure and that has evolved over more than 230 years.
Against this backdrop—a problem that is recurring, important, and understudied—this Article has two goals. The first is simply descriptive and corrective. Part I of this paper resituates borrowing authorities and limits in their proper constitutional context: borrowing is a power reserved to Congress, and limited delegations of that authority serve to make borrowing functional and credible. Starting in 1790, Congress would delegate the borrowing authority needed to cover appropriations—an underappreciated link that has always been present in borrowing authorities—but not so much as to make that borrowing crushing or extortionate. I then begin an excavation of debt authorities and limits before 1917, drawing on the details and data of public laws and borrowing records to provide an account of the nature and magnitude of those authorities over time. This study reveals that, far from being a tool monopolized by Congress and limited to wartime, a limited borrowing power was frequently delegated to the Executive Branch to pay for the general appropriations of government, most especially in the early Republic.
With that corrective story in place, I turn to the debates of the present. Part II offers a number of explanatory, doctrinal, and prescriptive insights. One is to help understand why debt limits—which have existed in some form since the beginning of the Republic—have only in recent decades become such a thorn in the government’s side. The fundamental reason cannot be that debt issuance is only now “limited”; that’s nothing new, and a truly limitless borrowing authority may be difficult to square doctrinally with the fact that the borrowing power resides in Article I. Instead, one unappreciated legal development is the long-brewing separation of the lawmaking for spending and revenue—a procedural hurdle that was not present in early borrowing authorities, and that means Congress now votes on spending twice. Amazingly, that development—now central to the limit’s constraints—did not occur for any high-minded fiscal responsibility reason, but so that the Treasury Department might (at the Treasury’s own request) enjoy greater discretion over how it issues debt.
I then turn to the lawfulness, and bindingness, of the debt limit itself. The President and numerous commentators have suggested recently that the debt limit is unlawful—a violation of the Fourteenth Amendment’s Public Debt Clause, which mandates that the public debts of the United States “shall not be questioned.”[21] Others suggest that, should the limit threaten to prevent the government from meeting its spending obligations, violating the limit would be the “least unconstitutional” option. And we are constantly warned—by Presidents, by commentators, by the media—that a binding debt will lead to a cataclysmic default that imperils the global economy.
But each of these common criticisms misses the mark. Statutory limits on debt issuance were familiar in 1790 and at the time the Fourteenth Amendment was ratified. The ordinary tools of constitutional and statutory interpretation suggest that—when faced with a conflict between the debt limit and spending obligations—it is the debt limit, and not the spending, that will prevail. The correct doctrinal response is to triage spending.
That last public-law claim may at first appear particularly surprising, but it should not be. Unilateral presidential action to issue debt or raise taxes beyond the terms of a statute would be truly unprecedented: a usurpation of Article I powers that no one doubts. But spending prioritization is more common even today—if perhaps not familiar—and anticipated in various contexts by both Congress and the courts. There are certainly situations where spending prioritization violates the Constitution, but a situation in which Congress requires it by statute is not one of them.
But does the debt limit really require it? The conflict between a spending command and a revenue limitation may at first appear unsatisfyingly intractable. Recourse to the text, and to basic canons, doesn’t seem to settle it. But practice can. The Executive Branch knows how to prioritize spending when outside circumstances require it; Congress has acquiesced to the inevitability (and perhaps even the occasional desirability) of such triage. By contrast, there is no known case since 1790 of the Executive Branch issuing debt without or beyond a statute; the Executive Branch has acquiesced to the legitimacy (if not always the desirability) of these limits. This “historical gloss” is one in which the Executive has acceded to Congress’s limits.[22]
And yet the consequences of a binding debt limit are often misunderstood. Because debt service is a small fraction of tax revenue, the federal government can continue to prioritize those payments even if total spending obligations cannot be satisfied. (Indeed, the debt limit may be the rare case in which Presidents have a strategic incentive to overstate the nature of their statutory obligations.) But this is not to suggest that the debt limit is unimportant—far from it. The first-order consequences of spending triage are both uncertain and large. Even if debt service can be paid—and must be under the Fourteenth Amendment—everything from Social Security payments to federal salaries cannot. Moreover, the limit presents the remarkable case of an important statute that has become slowly and unintentionally divorced from its original function. Limited debt authorities were once a simple mechanism to ensure that sufficient revenue was raised to cover spending, but not so much that the government would be unable to pay its debts. The modern debt limit has become precisely the opposite. Debt statutes have become an active impediment to the satisfaction of congressional spending and create doubt about whether longstanding debts can be repaid. The case for reform is ripe.
I. The Structure and Development of the Limit
I begin with a conceptual corrective, then turn to a historical one.
A. How to Conceptualize Debt Limits
The terms “debt limit” or “debt ceiling” may give one the misimpression that a statutory obstacle is being placed in front of an underlying authority.[23] But the Constitution grants Congress, and only Congress, the power “[t]o borrow Money on the credit of the United States.”[24] There is no serious constitutional argument that there is a generally applicable Article II authority to issue debt. The only argument approaching a view to the contrary, which I discuss later, is simply that doing so would be the “least unconstitutional option” under certain circumstances.[25] In light of the textual commitment of the borrowing authority to Congress, one might say that there is a default debt limit in American law: absent authorizing legislation, the Executive Branch can borrow nothing. The Constitution’s default debt limit is zero dollars.
Because the Executive Branch possesses no constitutional authority to issue debt, any statutory framework that governs debt issuance must be both an authorization and a limit. The modern statutory framework fits this mold. It grants the Treasury and the President the authority to borrow, and places various limits on that authority. Since 1982, both the authority to borrow and the limits on that authority are in Chapter 31 of Title 31.[26] Section 3102 and those that follow provide affirmative authority to borrow by selling bonds, Treasury bills, and the like. Section 3101 contains a numerical limit on how many of those securities can be “outstanding at one time”—the provision that is typically referred to as the debt limit.
One could likewise imagine debt-issuance schemes that offered far less discretion to the Executive Branch than our current scheme does. For example, an issuance statute might take the form of a statutory command (“shall issue”) rather than authorizations (“may issue”). Before 1917, a small number of debt-issuance statutes did in fact take this form, a fact that may contribute partially to the common misconception that, prior to 1917, Congress itself designed and approved each issuance of new debt.[27] By the same token, present doctrine makes it difficult to imagine a completely untethered debt issuance authority—that is, one that gave complete discretion over issuance to the Executive Branch. The familiar doctrinal minimum is that Congress must offer some “intelligible principle” by which the Executive could issue debt, and that intelligible principle would be the debt limit.[28] There may also be sound policy reasons for resisting a scheme of untethered Executive debt issuance, and there is no history of such practice in the United States.
The central point discussed above—that the Executive Branch has only as much borrowing authority as Congress has seen fit to give—suggests a corollary. There are many different ways in which debt issuance can be authorized and limited—and, thus, many ways to construct a debt limit. In recent years, the overwhelming focus of policy debates and public attention has been on the aggregate statutory limit on the outstanding face value of United States securities. That limit is important because, at least in recent decades, it is what threatens to bind, and thus what threatens to prevent the government from meeting its spending obligations. But our modern statutory scheme in fact contains a number of different limitations on the issuance of debt—all of which are unappreciated, and some of which are just as binding.
Some present limitations concern the form in which debt is issued. As noted above, the statutory structure of Chapter 31 of Title 31 does not simply give free rein to the Treasury so long as the total face value of those obligations is below a certain amount. Instead, it lists the types of obligations that the Treasury may issue, the general features that those instruments may have, and some procedural restrictions on how they must be sold and redeemed.[29] (E.g., the payment date of certificates of indebtedness and Treasury bills may not be more than one year after the date of issue.)[30] The modern Title 31 issuance authorities do afford the Treasury relatively wide discretion in how it goes about selling securities and raising funds—and, as discussed below, became more relaxed over the course of the twentieth century, in large part at the Treasury’s request. But the statutory structure here is also a far cry from a complete delegation of borrowing authority to the Treasury. Indeed, that structure does not authorize some forms of revenue-raising that have been historically important to the United States. Missing, for example, is any statutory authority for the Secretary of the Treasury to negotiate a one-time loan with France or the Netherlands—sources of borrowing authority that would have struck the founding generation as more important and intuitive than selling T-bills, and that were used into the nineteenth century.[31] Restrictions on the form of borrowing can also bind, and bind in important ways. Until 1971, for example, Congress continued to impose a statutory maximum interest rate on long-term government securities.[32] The Treasury expressly asked for that relaxation of the interest rate limit precisely because it did bind and pushed the Treasury out of the market for long-term financing.[33]
Another class of limitations concerns the purpose for which debt may be issued. In present law, this takes the following form: the Treasury may only borrow “amounts necessary for expenditures authorized by law.”[34] This may at first glance seem like a rather trivial restriction, if not a gratuitous one. But it is not. The Spending Clause and the Appropriations Clause do not independently limit the Executive Branch’s ability to raise funds; that is done only through Article I’s reservation of the power to tax and borrow, and there is no necessary constitutional connection (even if there is a practical one) between the exercise of Article I’s revenue-raising and spending authorities.[35] The government may, for example, run a surplus, and have no statutory authority to dispose of the excess funds. To provide such a connection, borrowing authorities have, over the course of American history, almost universally contained a limitation based on what the Treasury or the President may do with the funds. As discussed later, the character of those restrictions has varied, and the absence of a modern link between annual appropriations and debt authorities—that is, a link between the authority to spend and the authority to raise revenue—may go some length toward explaining why crises over the debt limit are now so common. And, as that connection suggests, limited debt authorizations are about more than simply ensuring that Congress’s borrowing power is defended; they are also about ensuring that Congress’s appropriations interests are successfully advanced.[36]
The present statutory restriction on the purpose for which debt can be issued—which allows borrowing only to the extent necessary for appropriations—also continues to serve at least two practical functions. First, it limits the degree to which the Executive Branch can use the robust market for Treasury securities as a tool of monetary, fiscal, or social policy.[37] Second, the “necessary for expenditures” limitation constrains the degree to which the Executive Branch may use debt issuance to stockpile cash. The amount of cash the Treasury has on hand varies greatly and is a function not just of what Congress has decided to appropriate, but of what the Treasury expects Congress may do.[38] Were it not for this restriction on cash accumulation, it’s possible that the numerical limit on outstanding debt would be a nullity. In the periods in which the debt limit has been suspended, the Executive would have borrowing power limited only by the appetite of the market—and could accumulate a large stockpile of cash to fund any future appropriations.
Other statutory restrictions on the purpose of a borrowing authority are possible too. Before 1917, Congress would sometimes—though not always or often—authorize debt issuance for a specific project, although the degree to which such authorizations are different from borrowing authorizations that support appropriations generally will vary.[39] Commonly cited examples of statutes that fund particular projects are the two public laws, enacted in 1902 and 1909, that authorized bond sales for the construction of the Panama Canal.[40]
The general point here is simply to note the basic structure that all Executive Branch borrowing authorities in our constitutional system have had and must have. First, some authorizing legislation is always required for the Executive Branch to borrow, because Article II provides no independent constitutional authority to do so. Second, any authorizing legislation will be both an authority and a limit; an untethered authority would raise serious constitutional questions and policy concerns. Third, there are many ways to structure a statutory limit on borrowing—the numerical limit on how much debt can be outstanding at a given time is just one. Other limits, such as the statutory requirement that borrowing be “necessary to expenditures,” are important and binding too.
B. Why Constitutionalize and Limit Borrowing?
The basic constitutional and statutory structure described above—under which the Executive Branch has only that borrowing authority which Congress is willing to give, and which Congress can limit in varied and complex ways—raises an obvious question: why have this structure? It is, after all, not the only way of structuring sovereign debt issuance. Virtually every state government, for example, has a strict constitutional or statutory limit on debt issuance, often in the form of a balanced budget requirement.[41] There is also international variation in approaches to limiting sovereign debt, with most countries having nothing that resembles a debt limit, while others have a limit with a different structure, such as a ratio of debt to GDP.[42]
While a comprehensive discussion of the constitutional origins of the borrowing power—and the theory and structure of sovereign borrowing more generally—is beyond the scope of this Article, a few initial observations will inform the discussion of practice since 1790 that follows.
A central challenge for the design of any sovereign borrowing authority is how to make it effective: capable of credibly obtaining funds, and persuading creditors that the borrowing can in fact be repaid.[43] That basic design question tracks the relevant American legal history. Like the power to tax, the borrowing power was understood as necessary to the operation of the nascent constitutional republic. And, like the taxing power, the borrowing authority was understood as the power to destroy.
The first major debt-related issue at the founding, after all, was how to pay for the still-lingering debts of the Revolutionary War. Under the Articles of Confederation, Congress had no taxing power, and instead had to rely on “requisitions” from state governments—demands for payment that it had few means to enforce—to fund the portion of the debt that Congress owed.[44] The Confederation Congress did have the power to borrow, but it required a nine-state supermajority to be exercised.[45] The combination of a large debt and a limited ability to raise funds left the financial state of the confederation a shambles.
Experience under the Articles of Confederation had made clear that borrowing would be difficult when not backed by a credible means of raising current revenue. For this reason, Alexander Hamilton in Federalist 30 noted that “[t]he power of creating new funds upon new objects of taxation, by its own authority, would enable the national government to borrow as far as its necessities might require.”[46] After all, he asked, “who would lend to a government that prefaced its overtures for borrowing by an act which demonstrated that no reliance could be placed on the steadiness of its measures for paying?”[47] Both incurring and repaying a sovereign debt are difficult without a taxing power or some other way to credibly raise revenue. At the time of repayment, obviously, the government needs a way of actually repaying the funds. At the time of borrowing, however, the government also needs a way of credibly signaling that it can and will repay the funds.
Several features of our Constitution speak to the issue of ensuring that sovereign borrowing could be credible and effective. First and most obviously—and in most obvious contrast to the Articles of Confederation—the Constitution gives Congress the power to both borrow and tax. Second, Article VI states a commitment that “[a]ll Debts contracted and Engagements entered into, before the Adoption of this Constitution, shall be as valid against the United States under this Constitution, as under the Confederation.”[48] (The assumption of state debts, meanwhile, would be settled by statute—and dealmaking in the proverbial “room where it happens”—in the compromise of August 1790.)[49] The delegates to the Philadelphia convention had also considered constitutional provisions that would have required the payment of public debts, or forbid appropriations from being diverted once applied to that purpose, but these proposals were rejected as too risky a constraint in times of emergency.[50]
At the same time, our constitutional structure is concerned with much more than repaying preexisting debts; it was anticipated that the borrowing power would be used. One striking feature of founding-era discussions of the borrowing power is how many of them are also discussions of the war power. In Federalist 30, Hamilton observed that “[i]n the modern system of war, [even] nations the most wealthy are obliged to have recourse to large loans,” and noted that “[a] country so little opulent as ours, must feel this necessity in a much stronger degree.”[51] Other early American commentators discussed the necessary relationship between the war powers and the borrowing power in similarly resigned terms.[52]
The fact that so much borrowing would be required for national defense was also what made that power dangerous. It was by means of the borrowing power, the anti-Federalist Brutus warned, that the new country “may create a national debt, so large, as to exceed the ability of the country ever to sink [it].”[53] Because he could “scarcely contemplate a greater calamity that could befal[l] this country, than to be loaded with a debt exceeding [its] ability ever to discharge,” he warned that the government should have restricted the borrowing power so “as to have rendered it very difficult for the government to [practice] it.”[54] In making that connection, the founders also would have been familiar with the British experience, whereby the profligate and forcible borrowing of the Crown led to the so-called “financial revolution,” including parliamentary constraints on borrowing.[55] They would have been familiar, likewise, with defaults in Spain.[56]
Against this backdrop, one can understand limited borrowing authorizations as serving a dual function. On the one hand, borrowing was necessary to fund the work of government, especially for a government so “little opulent” as early America. On the other hand, limits were necessary to ensure that contractual borrowing commitments were credible—that they could, and would, be repaid. For a country with the lingering financial baggage of early America, an unchecked borrowing power would have been a less effective one.[57]
Congress did, in fact, put its refurbished Article I borrowing power to quick use. And it did so not, as most commentators have suggested, through micromanaging a small number of issuances to fund wars and specific projects. It did so through a regular series of flexible delegations to the Executive Branch, most of which were made available simply to fund the general appropriations of government.[58]
Moreover, many of those borrowing authorities and limits were of a character that is not fundamentally different in kind from what we see today. Congress did not typically set interest rates, dictate specific issuance amounts, or set the timing of an auction or subscription. Instead, it provided limits of a character that were familiar through the second half of the twentieth century and that remain familiar today. Congress would typically set a maximum amount that could be issued, state a general purpose (typically the satisfaction of appropriations), and set a maximum coupon rate. On some occasions it would give the Executive Branch even more control over the terms of sovereign borrowing than we see today.
On the other hand, there is no period in which the Executive Branch exercised a borrowing authority without limit, and indeed several decades in which Congress gave the Executive Branch no new borrowing authorities at all, a time in which the Executive Branch did not borrow. The Executive Branch appears to have respected those limits. The history of this borrowing thus has an important dual character. On the one hand, the history of Congress simultaneously authorizing the Executive Branch to borrow goes back much further than previously appreciated. On the other, the story is one of Executive acquiescence—those limits have been respected.
1. The Trends in Early Borrowing Authority
Figures 1 and 2 below display all discretionary Executive Branch borrowing authorities between 1790 and 1910 using a histogram that allocates the nominal amount of each borrowing authority to the year of enactment. I allocate the borrowing authorities of eighty-three separate public laws in this fashion—a set that is hopefully comprehensive.[59] In almost all cases, the dollar amount of the borrowing authority appears on the face of the statute. In cases where it does not, I impute an amount using other data.[60]
The purpose of this exercise is not to provide anything approaching a comprehensive account of early American public finance, but simply to highlight the important and overlooked high-level contrasts in when and how borrowing authorities were implemented. First, debt authorities and limits were hardly unknown in early America, and they were not limited to wartime. In each year between 1790 and 1817, for example, the Executive Branch always had some statutory borrowing authority available to it. On the other hand, there are long periods of time where the Executive Branch had no available borrowing authority, particularly in the decades before and after the Civil War.
Between 1790 and the War of 1812, there were thirty-one separate public laws that contained a discretionary borrowing authority for the Executive Branch.[61] Most of these laws also made appropriations, and simply attached borrowing authority to the listed expenses. In 1793, for example, Congress appropriated no more than $1,589,044 for the expenses of the government, and gave the President the authority to borrow “any sum or sums, not exceeding, in the whole, eight hundred thousand dollars . . . to be applied for the purposes aforesaid.”[62] Other borrowing authorities had a purpose that was even more general; the most commonly recurring language used in early authorizations was simply that the borrowing could be applied “to such public purposes as are or may be authorized by law.”[63] And these authorities, moreover, were put to use—they were not gratuitous provisions left to lie fallow.[64]
The great majority of wartime borrowing laws also made no effort to limit the use of funds to defense expenditures. The borrowing authorities during the War of 1812, for example, were simply authorities to defray any expenses of government authorized by law.[65] The same was true of borrowing authorities during the Mexican-American War.[66] So too the Civil War.[67] The Liberty Bond Acts themselves—the central borrowing authorities in the First World War; conventionally thought of as the progenitors of the modern debt limit—likewise use this language.[68] Only the borrowing authorities for the Spanish-American War limit the amount of debt issuance to war-related spending.[69]
The so-called “project finance” model of borrowing—by which Congress authorized debt issuance for a specific spending project—certainly did exist prior to the twentieth century, though it appears rare in early legislation and uncommon throughout the nineteenth century (a particularly prominent example is the Panama Canal, funded by two issuance authorities that were specifically earmarked for the purpose).[70] The great majority of the dozens of debt authorities in the early republic were for general expenditures of government and, secondarily, to refinance the government’s existing debt.
Finally, it is worth noting that even debt authorizations that purport to limit the use of funds provide a source of general support for the expenditures of government. Because the government’s revenue is fungible, providing a source of funding for one project of government provides funding for others, too, since it allows government to shift resources from one area to another more freely. Assuming that the whole of government is adequately funded, there is no difference between a statute that appropriates $100 million for a canal and allows the government to borrow for that purpose, and a statute that appropriates $100 million for a canal and gives the Executive Branch an additional $100 million in general revenue.[71]
Figure 1: Executive Branch Borrowing Authority by Year of Enactment: 1790–1850
Figure 2: Executive Branch Borrowing Authority by Year of Enactment: 1851–1910
2. The Scope of Executive Branch Discretion
But are the terms of these early borrowing authorities really comparable to the debt limit of today? The details matter. If the exercise is to evaluate the practice of the branches over time, that practice should concern a common object. And, while scholarship on the history of the limit is scant, what exists suggests that Congress made virtually all the key decisions.[72]
But this is not the case. The norm of early borrowing authority was to leave many of the key determinations up to the President and others in the Executive Branch. Occasionally—and especially with early borrowing authorities—the restrictions on Executive Branch borrowing were minimal. A good example of this is the law of May 7, 1800, an act “to enable the President of the United States to borrow money for the public service,” which allowed the President to borrow “a sum not exceeding three millions [sic] five hundred thousand dollars,” to be used “for making up any deficiency in any appropriation heretofore made by law, or to be made during the present session of Congress,” and to be borrowed “upon such terms and conditions, as he shall judge most advantageous for the United States.”[73] The law contained a restriction based on amount, only the most general restriction on purpose, and no restriction on the terms of the debt contract. In certain respects, this banal early authority offers strictly greater discretion to the President than the authorities of today.
Still, perhaps the most intuitive difference between such a limit and the debt limit of today is a simple one. The contemporary debt limit at 31 U.S.C. § 3101(b) is a limit on how many obligations can be “outstanding at one time”: a restriction that allows the government to refinance or roll over debt without exceeding the limit and does not put restraints on when obligations are bought and sold. By contrast, many of the authorities described above and throughout late eighteenth and early nineteenth centuries were authorities to issue a certain value of debt obligations, and often to do so within a specific period of time (typically one year after passage).
But the importance of this contrast should not be overstated, for at least two reasons. First, in the nineteenth century, it was not unknown for Congress to set an aggregate limit on the number of certain classes of obligations that could be outstanding at one time. As early as 1840, for example, Congress set an overall limit on the number of short-term debt obligations (“notes”) that could be outstanding and gave the Executive Branch discretion over the timing of issuance and redemption.[74] Other laws later in the century and prior to the First World War contained similar limits.[75] Second, there is no difference in principle between Congress setting an overall cap on the value of outstanding securities—and then raising that cap by regular legislation—and Congress providing discrete authorities in individual pieces of legislation. The same fiscal goals can be achieved by either means.[76]
The most common conditions in early borrowing statutes were conditions that are still familiar today, or at least were familiar for most of the twentieth century:
Amount. Perhaps the most common constraint in early borrowing authorities is a limitation on the total value of debt that can be created. This is the form of limit most intuitively comparable to the statutory debt limit at 31 U.S.C. § 3101(b), which limits the total face value of obligations that can be outstanding at one time. Almost all borrowing authorities before 1917 included such a limit—and when they didn’t, they included a statement of purpose that would have effectively acted as a limit on the amount. For example, an appropriations law might describe expenditures for a given year and limit the related borrowing authority to “any sum or sums not exceeding in the whole, the sums herein appropriated.”[77]
Coupon Rate. The vast majority of borrowing authorities before 1917 also contained a limitation on the interest rate. Not every act did this. In rare cases, such as the 1800 example cited above, Congress provided no limit on the rate, instead allowing the President to borrow on “such terms and conditions, as he shall judge most advantageous.”[78] In other cases—likewise rare—Congress would identify the specific interest that must be charged.[79] In most cases, however, Congress would simply set the limit on how much interest could be charged—usually five or six percent. This is a contrast with the contemporary debt limit—although, as noted above, a maximum interest rate restriction on federal borrowing was not removed until 1971.[80]
Purpose. Almost every borrowing authority between 1790 and today has contained some limitation on what the borrowing can be used for. The great majority of these limitations have simply been to link borrowing to the general expenses of government; the most commonly recurring limits before 1917 would have linked borrowing to “necessary to meet public expenditures”[81] or “to defray any of the public expenses which are, or may be authorized by law.”[82] This broadly mirrors borrowing authorities today, which are likewise limited to “amounts necessary for expenditures authorized by law.”[83] Occasionally, as noted above, Congress would adopt a “project finance” approach to issuance. Likewise, on rare occasions, Congress would provide an authority to borrow that contained no guidance on the purpose to which the revenue should be applied.[84] But the satisfaction of general appropriations has always been the lodestar of executive borrowing authority.
3. Congressional Control and Executive Acquiescence
Debt limits existed. But was the Executive Branch restrained by them? This Section emphasizes three features of what is known about practice between the branches.
First, there were no periods in which the Executive Branch had an unfettered statutory borrowing authority. Every borrowing authority since 1790 has come with limits attached. If anything, the public laws between 1790 and 1917 exhibit an abiding sense of successive congresses concerned with shaping and defending the contours of the borrowing power. In several laws, for example, Congress specifically noted that it “shall be deemed a good execution of the . . . power to borrow” for the President or the Treasury Secretary to issue “certificates of stock” for the amounts borrowed.[85] Others still required a report to Congress on how a particular Executive Branch borrowing authority had been exercised.[86]
Second, Treasury borrowing records appear to confirm that those limits have been respected. While this is a difficult negative to prove to complete satisfaction—that every loan and issuance was pursuant to statutory authority—the available records suggest that, at least since 1790, it is true. Each loan and issuance in the Treasury’s public records between 1790 and 1917, for example, has a statutory authority attached, and each issuance complies with the formal limits of that authority.[87] Since 1917, the Executive Branch has not exceeded the aggregated limit.[88] It is worth noting that many of the early issuance limits were indeed binding, at least in the sense that the Executive Branch would frequently make issuances with an amount and interest rate precisely equal to the relevant statutory maximum (e.g., a statute would allow up to $3.5 million in borrowing, and the President would borrow exactly that).[89]
Third, and perhaps most intriguingly, the early points of friction between the branches tend to support the conclusion that this is an area in which Congress has successfully asserted its prerogatives and the Executive Branch has complied. There are two particularly delightful vignettes: one in which the Executive Branch itself acknowledges the need for lawful authorization, and one in which Congress resolves a possible effort to skirt a limit’s formal terms.
While it is true that all borrowing records since 1790 provide statutory authority for the relevant loan or issuance, this is not the case for a series of loans obtained in the fall of 1789: funds from New York banks totaling $191,608, the contracts for which were “made upon the authority of the Secretary of the Treasury,” rather than (as all later loans and issuances are) under the authority of an act of Congress.[90] Hamilton had obtained these loans to “meet expenses incurred at the beginning of the present government of the United States” (including salaries for Congress and the President) begun in March of that year.[91] The next year, Hamilton reported to Congress on the state of this borrowing with an air of excuse and apology, noting that the loan was “the result of necessity”[92] and acknowledging that it wouldn’t happen again: “Obvious considerations dictate the propriety, in future cases, of making previous provision by law for such loans as the public exigencies may call for, defining their extent, and giving special authority to make them.”[93]
There appears to be only one other early episode in which, plausibly, it happened again. On August 4, 1790, the federal government assumed the Revolutionary War debts of the states with an Act that also created a new $12 million borrowing authority for the President to pay off the federal government’s foreign debts.[94] The next week, on August 12, a second public law gave the President additional authority to repurchase the foreign debt by borrowing a second sum, “not exceeding in the whole two millions of dollars, at an interest not exceeding five per cent . . . [to] be also applied to the purchase of the said debt of the United States.”[95]
The President did, in fact, exercise these authorities, taking out a number of loans from Dutch and Belgian bankers.[96] In March of 1791, however, controversy emerged over the exercise of the President’s August 12 authority. According to Congress, the loan accumulated interest at a rate of five percent annually, but also contained additional “charges” of four and a half percent.[97] As a result, Congress noted the prefatory statement to a law resolving the issue, “a doubt hath arisen, whether the said loan be within the meaning of the said last mentioned act, which limits the rate of interest to five per centum per annum.”[98] Nevertheless, Congress concluded that it would be “expedient that the said doubt be removed,” and thus made it “enacted and declared by the Senate and House of Representatives of the United States of America in Congress assembled, [t]hat the loan aforesaid shall be deemed and construed to be within the true intent and meaning” of the August 12 authorizing act.[99]
There is perhaps some gentle ambiguity about how to best interpret this episode; Congress ratified the loan, after all. But the whole premise of the congressional statement was that expedient and doubt-free borrowing required statutory authority. Both early conflicts ended with the Executive Branch and Congress acknowledging that principle. And, since then, the Executive seems to have respected it.
D. What Changed Between 1917 and 1941?
A central theme of the previous discussion is that many of the legal features thought to exist only after 1917 in fact existed well before then. Executive Branch borrowing authorities have existed since 1790 and were legion in the first decades of the Republic. Moreover, these authorities always contained limits—limits that still allowed the Executive Branch a degree of authority and flexibility that is not different in kind from what we see today, and limits that the Executive Branch seems to have respected, even under conditions where appropriations were otherwise underfunded.
But the period between 1917 and 1941 did contain some changes of degree, most notably an additional increase in the degree of discretion that Congress gave to the Treasury, largely at the Treasury’s request.[100] Somewhat ironically, however, that increase in discretion led unintentionally and imperceptibly to the separation of spending and borrowing authority that now requires us to vote on spending twice—and that makes debt limit votes an important strategic juncture for extracting legislative concessions.
The Second Liberty Bond Act[101]—the wartime borrowing legislation that is generally considered the origin of the modern debt limit—was an example of some of these things, but its importance as a legal innovation has been overstated. In terms of borrowing, the Act did familiar things. It allowed the Treasury Secretary to issue slightly more than $7.5 billion in bonds (with about $3.5 billion of those bonds “in lieu” of bonds that could have been issued under recent public laws), and also allowed an issuance of $4 billion in certificates of indebtedness and $2 billion in “war-savings certificates.”[102] But these basic mechanisms were nothing new. As discussed above, similar authorities and limitations had been used throughout the nineteenth century. Instead, the Second Liberty Bond Act was important not because it represents a great legal innovation, but for a formal reason: subsequent legislation amended it, and around 1935 those amendments began to be referred to publicly as amendments to “the debt limit.”[103]
In the decades that followed, however, the Treasury began agitating for greater freedom in managing the public debt. In the Treasury’s 1930 annual report on the state of American finances, for example, Treasury Secretary Andrew Mellon wrote with recommendations for amendments to the Second Liberty Bond Act: “[I]t is obvious that the orderly and economical management of the public debt requires that the Treasury Department should have complete freedom in determining the character of securities to be issued and should not be confronted with any arbitrary limitation which was not intended to apply to these circumstances.”[104] “Moreover,” he concluded, “it is highly desirable that the authority be provided well in advance of actual needs.”[105]
Congress did in fact relax many of the constraints about which Secretary Mellon complained. Perhaps the most important step came in 1939, when Congress combined the limits on most pieces of short- and long-term obligation.[106] In 1941, Congress combined them all into a single cap that persists today (the 1941 law also characterized that limitation as “the debt limit,” rather than a limitation specific to a category of obligation or a particular issuance).[107] From the Treasury’s perspective, these innovations were desirable because they gave the Department more control over the timing and terms of borrowing (i.e., a $7 million cap on long-term bonds and a separate $7 million cap on short-term certificates offers strictly less flexibility than a $14 million cap on all instruments).
What is perhaps most remarkable about these legislative developments is that there was no sense in the record that they were intended for any righteous fiscal responsibility purpose. The record suggests the opposite. Congressmen understood that the purpose of a combined limit was to give the Treasury more discretion over which debt instruments it could issue.[108] Representatives of both parties understood and stated explicitly that it was not a check on spending. As one Senator put it in a debate over the 1939 law: “The only time to control a debt is when appropriations are made which contribute to the debt.”[109]
E. Continued Evolution Since 1941
The limit of the 1941 Act is what Congress has amended in subsequent legislation. But the limit has also continued to change and should not be viewed as a static entity. Three types of changes are worth keeping in mind.
First, and as suggested above, the debt limit continues to be an agglomeration of restrictions—not just a restriction on the face value of securities outstanding, but also restrictions on the form and purpose of government borrowing. The precise nature of those restrictions has continued to evolve since 1941. As noted above, for example, federal law continued to restrict the maximum interest rate of long-term securities until 1971. That limit was binding. It pushed the Treasury out of the market for long-term securities.[110]
Second, the overall character of the limit still changes. Between 1954 and 1983, and then again between 1987 and 1993, for example, Congress took a different approach to the debt limit. American statutory law contained both a “permanent” debt limit—a default limit on outstanding debt issuance—and a “temporary” debt limit. That is, Congress would raise the permanent debt limit by an additional increment for a period of time, usually somewhere between a couple of months and a couple of years.[111] And, as noted in the Introduction, between 2013 and 2019, Congress departed from the practice of raising the debt limit by a dollar amount—instead opting to suspend the limit for a period of time.[112] Lest the debt limit become regarded an immovable object, large change appears possible still.
Third, Congress itself (or at least the House of Representatives) has attempted to fashion mechanisms that create a restored link between spending and debt issuance. In 1979—during the era in which Congress regularly enacted temporary debt limits, but was beginning to consider the risks of a divergence between spending and financing authority—the House of Representatives instituted what has become known as the “Gephardt Rule” (named after the longtime Missouri Democrat Dick Gephardt).[113] The rule required that, when Congress adopted a concurrent budget resolution, a concomitant increase in the debt limit would be deemed to have received final passage in the House, and an engrossed copy would be sent to the Senate.[114] More recently, in 2019, the House adopted as Rule XXVIII what has been described as a “Modified Gephardt Rule.”[115] When the House adopts a concurrent resolution on the budget under the Budget Act, the Clerk of the House shall prepare an engrossed joint resolution suspending the debt limit until the end of the fiscal year.[116] This remained a rule of the 117th Congress.[117] Both the Gephardt Rule and its modified successor seem notable for the fact that they failed to settle controversies over the debt limit.[118] But they reflect the sense of many in Congress that procedural changes are desirable to avoid continual showdowns over the limit.
II. The Debate over the Modern Debt Limit
A. Why Are There So Many Debt-Limit Crises?
A theme of the previous Sections is that the modern debt limit is actually a composite of several different features: an authorization to borrow subject to a variety of constraints, including constraints on the form, purpose, and amount. A second theme is that the modern structure of borrowing did not emerge fully formed at a single point in time, with a single intention behind it. Instead, it evolved in a piecemeal fashion over the course of decades, in response to changing circumstances and needs.
And yet the United States had debt limits before 1941—scores of them—without the regular controversies of today. Those limits were consistently imposed by Congress and honored by the Executive Branch. While grand political controversies over the public debt were hardly unknown before the second half of the twentieth century—with perhaps the most prominent occurring after the Revolutionary War and the Civil War—those were discrete episodes caused by national crises, in which the public debt had ballooned to a size many times greater than what it had been. Early debt limits, moreover, were clearly a binding constraint on government finance. But the controversies over the debt limit in recent decades stand out for their consistency, intensity, and seeming intractability. Why?
A full explanatory and predictive model of debt-limit showdowns would verge on a theory of everything: partisan polarization; the growth of the modern welfare state; bargaining between the branches; the economic and political tradeoffs between taxation and debt-finance—the list goes on.[119] It may also be that early America was simply more comfortable with spending deficiencies, and that the consequences of a spending shortfall in a smaller government were less dramatic (there is no 1790 analog to Social Security checks not going out). But, without attempting that theory of everything, here is one account.
The simplest explanation for why we now encounter so much friction over the debt limit—though not an ultimate explanation—is twofold. First, we spend more, and second, the growth of spending has exceeded the growth of tax revenue. The United States is much more reliant on public debt today than in the nineteenth century. Early America relied on borrowing, but the scale and timing of issuance authority was highly varied, as Figures 1 and 2 suggest.[120] The government’s debts would fall between wars, and so would the need for large borrowing authorities.[121] Early American government spending, likewise, did not increase monotonically; it fluctuated around wars and other national crises.[122] For long stretches of time, the country ran surpluses, and successfully eliminated the national debt.[123] Early debt statutes convey an abiding sense that a national debt would not be a permanent feature of American life, but rather a temporary feature that the government should and would eliminate.[124] But, since the late 1920s, this has no longer been the case. The national debt has generally increased and has been used to fund the growth of the modern welfare state.[125] That growth—coupled with the fact that it has exceeded the growth of tax revenue—has created a regularly recurring situation in which the debt limit must be raised to keep pace with lawful and required spending. And because the limit must be raised, it is strategically useful for some members of Congress because it creates a veto point at which additional legislative concessions can be extracted.
But one important legal development has contributed to the rise of the debt limit as a strategic tool: the link between appropriations and borrowing has become severed. The earliest issuance authorities were typically attached to appropriations laws.[126] Other borrowing authorities were attached to specific spending projects.[127] All of these laws, it bears emphasizing, were thus simultaneously a limit and an authority. Congress limited what the Executive Branch could issue to protect its constitutional prerogatives and to ensure that borrowing remained credible. But the more banal and straightforward goal was simply to ensure that the Executive Branch would have enough revenue to cover spending. By contrast, modern debt-issuance statutes have become standalone authorities with little temporal or procedural connection to the expenditures they are supposed to serve.[128] Because Congress now votes separately to approve spending and to finance it, it ends up voting on spending twice—a considerable procedural hurdle that was not present with early spending, and that makes votes over the debt limit so strategically important.
That current state of affairs—under which Congress votes for spending twice—is ironic for several reasons. First, as noted above, the original functional purpose for limited debt authorizations was twofold. Congress would not provide so much authority so as to create doubt about whether debts could be repaid but would also provide sufficient borrowing authority to ensure that appropriations could be satisfied. Yet the modern structure of borrowing authorities has transformed into something close to the opposite. Because there is constant doubt about whether the limit will be increased, it has become a source of skepticism as to whether Congress will make good on its spending commitments.
Second, the Executive Branch shares some responsibility for the predicament. The Executive Branch itself pushed for borrowing authorities to be separated from spending projects—and as a means, no less, of expanding Executive Branch discretion over debt issuance.[129] As noted above, an aggregate cap on all forms of debt instruments did, indeed, increase the Treasury’s discretion over the terms on which it could borrow. It gave the Treasury more control over the choice between short-term and long-term borrowing, and the timing of auctions and issuance.[130] This was what Secretary Mellon had asked for in 1930: “complete freedom in determining the character of securities to be issued.”[131] But it came at a significant price. By separating the sources of funding from the spending they are supposed to serve, that cap produced an additional legislative hurdle that all spending must clear. Mellon did not get the second part of what he requested: an authority “provided well in advance of actual needs.”[132] Instead, Congress often bargains over financing after the spending has been approved.
Finally, all of this happened without any sense that Congress was pursuing an intentional policy designed to limit issuance or restrain spending. On the contrary, Congress seemed to share Secretary Mellon’s view that Treasury should enjoy “complete freedom in determining the character of securities to be issued,”[133] and that single cap on issuance would make borrowing cheaper. Much of the modern public debate over the debt limit assumes that it serves some purpose related to restraining spending—but there is no historical basis for that premise.
B. Is the Debt Limit Lawful and Binding?
The many controversies over the debt limit would be much less important if a common criticism of the debt limit were correct: that the limit is unlawful. While debates over the lawfulness of the limit have never fully ripened—Congress has always voted to increase the limit before Treasury ran short of funds—arguments that the limit is avoidable have reportedly gained some discussion and traction within government. Former President Bill Clinton, for example, has stated that during the 1995 and 1996 standoff over the debt limit, his lawyers researched the constitutional issues and he would have unilaterally raised the debt limit “without hesitation, and force[d] the courts to stop me,” seemingly on the theory that a binding limit violates the Public Debt Clause of the Fourteenth Amendment.[134] Some commentators have likewise argued that “the least unconstitutional option”—when faced with a binding debt limit and spending obligations about to go unfulfilled—would be for the President to ignore the debt limit and unilaterally issue new debt.[135]
But the force of these arguments has been greatly overstated. There are, in fact, strong reasons to think the debt limit is both lawful and binding. The Public Debt Clause of the Fourteenth Amendment likely imposes some obligation on the government, but the argument that the limit itself is unconstitutional is at odds with historical practice. The stronger view is that the Fourteenth Amendment simply creates a constitutional priority for debt service. Such spending prioritization, moreover, contains no constitutional defect. If anything, practice suggests that the debt limit should prevail over conflicting spending mandates. I consider these constitutional and statutory issues in turn.
1. Public Debt and the Public Debt Clause
The Public Debt Clause of the Fourteenth Amendment provides in full: “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.”[136]
Of all the recent controversies over the debt limit, the relationship between the Public Debt Clause and the statutory debt limit may be the one that has come closest to a concrete controversy in court. In 2014, one legal activist filed a lawsuit alleging that the debt limit was “unconstitutional and void” because it “does exactly what the Constitution explicitly prohibits; it questions the ‘validity of the public debt of the United States.’”[137] The district court concluded that the plaintiff lacked standing to bring his claim.[138] But the relationship between the Public Debt Clause and the debt limit remains an ongoing source of scholarly controversy, and recourse to the Fourteenth Amendment remains common in public and scholarly arguments over recent debt-limit crises. Some argue that the debt limit—any debt limit—violates the Clause as a facial matter.[139] Others argue that the Clause has no relevance, or only limited relevance, because it might not apply to the current national debt.[140]
The best reading may be somewhere in between. While the Clause is no paragon of clarity, a few generalizations may be hazarded, suggesting that the Public Debt Clause applies to the contemporary debts of the government, and not just those related to the Civil War. In addition, the contemporaneous practice described above makes it difficult to imagine a reading that limits the Clause’s applicability to a certain limited category of debt instrument.
As a historical matter, the Clause emerged from concerns with the size of the Civil War debt.[141] But the text and history generally support a wider application of the Clause. The Clause covers “the public debt of the United States . . . including debts incurred . . . in suppressing insurrection.”[142] Had the drafters wanted a version that covered only war debts, they could have selected any number of other phrases, including something more
like the second introduced version of the Clause, which covered
only “obligations . . . incurred in suppressing insurrection” and related expenses.[143]
The sole case in which the Supreme Court addressed (albeit in dicta) the scope of the Public Debt Clause, Perry v. United States, also supports the notion that the Clause sweeps beyond wartime debts. There, the Court upheld a plaintiff’s claim that Congress’s abrogation of a gold clause in a Treasury bond—essentially, a requirement that the bond be repaid in gold—violated section 4 of the Fourteenth Amendment.[144] “While this provision was undoubtedly inspired by the desire to put beyond question the obligations of the Government issued during the Civil War,” the Court noted, “its language indicates a broader connotation.”[145] The Court concluded by observing that it regarded the Clause “as confirmatory of a fundamental principle, which applies as well to the government bonds in question, and to others duly authorized by the Congress, as to those issued before the Amendment was adopted.”[146]
The general contours of these arguments are familiar.[147] But the broad applicability of the Public Debt Clause may also be supported by two arguments from practice that have gone previously unappreciated.
The first argument is admittedly somewhat speculative. There is an intriguing contrast in the use of statutory borrowing authorities before and after the Fourteenth Amendment. Prior to the Fourteenth Amendment, Congress would not just limit debt issuance—it would regularly include explicit statutory promises to repay the debt. For example: “[T]he faith of the United States shall be and hereby is pledged, to establish sufficient permanent revenues for making up any deficiency that may hereafter appear in the provisions for paying the said interest and principal sums, or any of them, in manner aforesaid.”[148] The idea of a legal commitment mechanism to signal the reliability of American credit was familiar around the Civil War, since Congress drafted statutes with repayment promises during the War itself.[149] After the ratification of the Fourteenth Amendment, however, the statutory promises stopped. While there is no smoking gun as to why, it seems plausible to think that, with the constitutional commitment of the Fourteenth Amendment in place, the statutory promises were gratuitous.[150]
Second, as a practical matter, it would have been difficult, if not impossible, to distinguish between (as the Public Debt Clause of the Fourteenth Amendment puts it) debts “incurred in suppressing insurrection, or in defense of the Union” and other forms of public debts. Consistent with practice in the decades before the Civil War, the wartime borrowing statutes contain only the broadest limitations on what Treasury could do with the funds. They provided fungible government revenue: the typical requirement during the Civil War was simply that the borrowing be “for the public service.”[151] Nor was there an easy temporal connection between the borrowing and the war. The major episode of borrowing in the years that followed, for example, was used to finance and roll over debt accumulated during the war.[152] Debts “incurred in suppressing insurrection,” in other words, were intermingled with all fungible revenue raised before, during, and after the war.
More puzzling is what to make of the Clause’s odd insistence that the validity of debt “shall not be questioned.” That phrasing is stubbornly elusive. Commentators have long disagreed over its scope, and the outer limits of the Clause present interesting and thorny questions that are not readily answered by reference to the drafting history.[153] The phrase appeared for the first time in the third and final introduced version of the Clause, and I have located no discussion of the wording in the congressional record.
An additional complication is that those who pushed for the Public Debt Clause were certainly worried about forms of debt repudiation that extended beyond the intentional cancelling of a government obligation. The states that repudiated their debts following the panic of 1837, for example, did so using a number of tactics, such as refusing to raise taxes and renegotiating the terms of their debt.[154] In 1868, the Republican Party ran and won on a platform that included several planks that emphasized a broad commitment to repaying the public debt, denouncing “all forms of repudiation as a national crime,” and stating that “national honor requires the payment of the public indebtedness in the utmost good faith to all creditors at home and abroad, not only according to the letter, but the spirit of the laws under which it was contracted.”[155]
Because there is so little contemporary evidence that speaks to the “questioned” question—and because the background political concerns encompassed a broad range of government policies—it can feel difficult to make progress on the issue of what the command of the Public Debt Clause actually means.
Nevertheless, two points seem apparent on either extreme.[156] First, the idea that any debt limit violates the Fourteenth Amendment as a facial matter—that is, the idea that any limitation on debt issuance necessarily constitutes a “questioning” of the Public Debt Clause—is rather difficult to reconcile with history and practice in this area. The idea of a fundamental conflict between the debt limit and the Fourteenth Amendment—and the so-called “nuclear option” of ignoring the debt limit because of its supposed unconstitutionality—can be laid to rest.[157] The Public Debt Clause was ratified against the backdrop of numerous statutory limits on debt issuance, and the debates over drafting and ratification betray no hint that those statutory limits would present a constitutional difficulty once the Fourteenth Amendment was ratified. Indeed, after ratification (in July of 1868) those limits continued. Later that month Congress authorized a limited issuance for the purpose of refinancing the national debt: “an additional amount of temporary loan certificates, not exceeding twenty-five millions of dollars.”[158] Moreover, the argument that any limitation on debt issuance constitutes a “questioning” of the public debt would seem to extend, implausibly, to any statutory limitation on the revenue raising activities of the state (such as taxation). Those limitations, too, might be thought to cast doubt on whether the government can repay its debts. That view cannot be right.
Second, and at the other extreme, it seems plausible to obtain some clarity on what the debt limit does forbid: the government failing to honor the express terms of a debt instrument. This form of repudiation would have fallen squarely within the concerns of the drafters and ratifiers, and provides a reasonable floor for violations of the Clause. Such a conclusion also accords with the Supreme Court’s dicta in Perry, in which the Court stated that it could not “perceive any reason for not considering the expression ‘the validity of the public debt’ as embracing whatever concerns the integrity of the public obligations.”[159] In so concluding, the Court rejected the argument that “[t]he word ‘validity’ refers to the essential existence of the obligation”—that is, the fundamental issue of whether it presented a valid claim against the government.[160] It seems fair to think that, if changing the manner of payment violates the Public Debt Clause, failing to make a scheduled payment would too.[161]
But what follows from that conclusion? The answer, for the reasons described above, cannot be that the debt limit is unconstitutional as a facial matter. It would mean simply that the federal government has a constitutional obligation to make payments on the national debt. In certain applications, the debt limit might be a constraint that makes such payments difficult. But even in such a scenario—one where, say, a service payment on the national debt comes due and the government cannot raise adequate funds to pay for everything—it does not follow that the government may simply disregard the debt limit. A more modest conclusion is simply that debt service payments are entitled to constitutional priority.[162] That is, the Executive Branch has a constitutional obligation, when satisfying the spending instructions that Congress has given it, to put them first.[163] A more dramatic solution—the repudiation of the debt limit statute itself—would enter play only if the government had inadequate resources to cover debt service, an outcome that seems highly unlikely as a practical matter, as explained more fully below.
2. The Constitutionality of Spending Triage
The constitutional prioritization of the national debt would not be a desirable solution—or even a particularly constitutional one—if it led to other constitutional infidelities. This is one of the most common legal objections to the limit: not that the statutory debt limit is facially unconstitutional, but that honoring the debt limit under certain conditions would itself work a constitutional violation.[164] When the debt limit binds, it is supposed, the Executive Branch faces a constitutional trilemma. It must either violate Congress’s authority “[t]o borrow Money on the credit of the United States”; “[t]o lay and collect Taxes”; or to “provide for the common Defence and general Welfare of the United States” through spending.[165] In the face of this trilemma, the putative least-unconstitutional option would be for the President to violate the debt limit and issue new debt.
But does a binding debt limit really present a constitutional trilemma that requires resolution? It seems clear that if the President were to raise taxes without statutory authority or sell Treasury bills in excess of the statutory debt limit, he would be violating the Constitution—and a court would conclude as much. It also seems clear that if the President were to purposely ignore congressional spending dictates and instead spend money according to his own policy preferences, he would likewise be violating the Constitution—and that a court would conclude as much too. In the spending case, the President would also likely be violating statutory law, either in the form of the Impoundment Control Act—which limits the ability of the Executive Branch to defer lawful spending requirements—or the Anti-Deficiency Act, which (in relevant part) prevents the Executive Branch from spending in excess of appropriations.[166] In accordance with these general intuitions, courts have unsurprisingly held that an agency cannot spend funds beyond what Congress has authorized.[167] That much seems clear.
But the conclusion above does not mean that all scenarios in which funds go unspent, contrary to a statutory command, result in a violation of the Spending Clause. Such scenarios are in fact common and are resolved without recourse to the grand language of constitutional violations.
One example is a case in which Congress simply has not appropriated enough money to accomplish the stated statutory spending goal. As the Government Accountability Office (GAO, an agent of Congress) acknowledges: “Sometimes the actual funding Congress appropriates for a program may fall short of original expectations.”[168] So what happens then? “If the agency cannot get additional funding and the program legislation fails to provide guidance,” the GAO advises, “the agency may, within its discretion, establish reasonable classifications, priorities, and/or eligibility requirements, as long as it does so on a rational and consistent basis.”[169] This standard also has a measure of endorsement from the Supreme Court, which has held that when an agency faces a shortfall of the type described, “it does not necessarily follow that . . . [an agency] is without power to create reasonable classifications and eligibility requirements in order to allocate the limited funds available” to the agency for Congress’s purpose.[170]
Structurally, the banal circumstance of a funding shortfall has much in common with the conflict between the debt limit and a statutory command to spend. Congress has legislated in a self-frustrating way—requiring that a certain result be achieved but failing to appropriate sufficient funds to achieve it. Yet this familiar circumstance has been encountered and discussed by all three branches without a hint that it raises constitutional peril. Indeed, at least one GAO analysis suggested (albeit in a somewhat conclusory fashion) that the Executive Branch should enjoy a similar degree of flexibility with prioritizing spending during a debt-limit crisis.[171]
Interbranch practice that has developed under the Impoundment Control Act (ICA)—the statute that prohibits the Executive Branch from deciding to spend less than appropriations command—likewise anticipates that there are lawful circumstances in which mandated spending may not occur. There is a strong argument that the debt limit is one of them. In relevant part, the ICA imposes certain substantive and procedural limits on “deferral[s] of budget authority,” which the Act defines to include any “Executive action or inaction which effectively precludes the obligation or expenditure of budget authority . . . as specifically authorized by law.”[172]
But neither Congress nor the Executive Branch has interpreted this provision to reach all cases in which funds go unspent. The GAO, for example, has long distinguished between a covered “deferral” and what it calls, in somewhat ungainly fashion, a “programmatic delay”: a delay “in which operational factors unavoidably impede the obligation of budget authority, notwithstanding the agency’s reasonable and good faith efforts to implement the program.”[173] For the ICA deferral provisions to apply, in other words, the Executive Branch itself must choose to engage in some active “withholding or delaying,” or there must be some other “Executive action or inaction [that] effectively precludes” the spending in question.[174]
A simple example of an acceptable good-faith “delay” might be if an agency’s good-faith implementation of a spending program is interrupted by a natural disaster. A more complicated example might be if an intervening Supreme Court decision or new statute creates new legal requirements or uncertainties to which the agency’s decisions must adjust. In both cases, budget authority may go unused, contrary to the mandates of a statute. But, under GAO’s interpretation of the ICA, there will have been no “Executive action or inaction [that] effectively precludes” the implementation of a spending law.[175]
In some philosophical sense, of course, there will always be an “action or inaction.” But a sweeping and unbounded test is not how practice has developed under the statute (and would likely not be a good interpretation of the statute).[176] Instead, GAO’s own test looks to the substantive motivations and constraints of the Executive Branch in deciding whether the ICA is implicated, and the Executive has run afoul of it. Crucially, it is “[t]he reason for a delay in obligating, not the delay itself, [that] is the key to determining if the requirements of the Act apply.”[177] GAO thus probes the relevant motivations, as well as legal and practical constraints, in deciding whether the ICA is implicated.[178] If the debt limit binds and prevents the government from meeting spending obligations, the situation is fairly akin to a natural disaster—and it seems reasonable to conclude that a delay in spending on account of the debt limit is not the kind of “action or inaction” to which the ICA is directed.
These conclusions are more than simply wooden standards generated under statutes and appropriations frameworks that may feel obscure.[179] There is good constitutional sense behind them. The Spending Clause is Congress’s power to “provide for the common Defence and general Welfare of the United States.”[180] But when Congress legislates in a self-frustrating way, it has made no obvious determination with respect to the “general welfare.” The Appropriations Clause, meanwhile, commands that “[n]o Money shall be drawn from the Treasury, but in Consequence of Appropriations made by Law.”[181] But good-faith spending prioritization does not run afoul of this either: all of the spending will still have flowed “in Consequence of Appropriations made by Law.”[182] Spending triage, long accepted by the branches, contains no necessary constitutional defect.[183]
On the threshold question of whether a spending lapse creates a constitutional violation, then, there are two important takeaways. First, there is a long history of practice to suggest that not all spending shortfalls create constitutional difficulty—and it is reasonable to think that a shortfall caused by the debt limit is one such situation. Congress itself seems to understand that there will be lawful situations in which—indeed, Congress itself seems to create situations in which—spending commands will go unsatisfied. Under current doctrine, at least, that practice matters. The Supreme Court has emphasized that, in deciding “whether the President is acting alone or at least with the acceptance of Congress,” it looks to “the general tenor” of legislation in the area in question, as well as to whether there is a “history of acquiescence” in a particular Executive Branch practice.[184] The general tenor is for Congress to understand that some spending mandates go unmet—a result codified in the ICA and accepted across decades of practice.
Second, the Executive Branch has historically viewed itself as obligated to construe statutes, where possible, so as to avoid constitutional problems.[185] When faced with the tension between the debt limit and spending statutes, that obligation points in only one direction. If spending statutes were to dominate, the constitutional violation would loom large and clear: unilaterally raise taxes or issue debt—a constitutional violation that no one doubts.[186] On the other hand, if we read the constellation of relevant statutes so that the debt limit controls, we have a plausible savings construction. Spending less than Congress commands, in circumstances where a statute requires that result, creates no constitutional problem.
3. Why Spending Should Yield to the Limit
But is the savings construction above really plausible? Some critics of the debt limit have long suggested that it is not—and that, as a matter of ordinary statutory interpretation, the debt limit must yield to spending requirements.[187] But this is not the case. When there is a conflict between spending and the limit, there are also strong statutory arguments that it is the spending, and not the limit, that should yield.
First, when the debt limit is reached, there really is a looming conflict between spending commands and revenue limits. All relevant statutes appear equally obligatory. Annual appropriations command that funds “shall be available” for the appropriated purposes.[188] And the debt limit commands that “[t]he face amount of obligations” issued by Treasury “may not be more than” the current limit (at the time of this writing, about $31.4 trillion).[189]
Can the ordinary tools of statutory interpretation resolve it? Some scholars have suggested that spending should triumph, on the theory that spending statutes are more specific and later in time.[190] But the story is more complicated. For one, it is not always (or even particularly often) true that spending legislation is later in time than debt-limit legislation. There are almost a hundred laws in which Congress has increased the modern debt limit—most recently in June 2023, with a suspension until January 2025.[191] It is easy to conceive of a debt-limit crisis in which the public law raising the debt limit is more proximate in time than the most recent annual appropriation. This would have happened, for example, if Congress had failed to raise the debt limit in December 2021; the most recent debt-limit law would have been from October 2021, while the most recent annual appropriations law would have been from September.[192] It would be strange to suggest that federal spending hinges on timing variables that are as good as random—and that apparent randomness is good reason to think these canons simply don’t apply.
Indeed, applying a “last in time” canon would be worse than random when applied to a large percentage of federal spending, because the most recent debt-limit law will always be more recent in time than the bulk of government spending that is “mandatory” in nature—i.e., entitlement spending that is insulated from the annual appropriations process, and insulated from that process precisely because it is considered too important to be left to the vagaries of annual politics. Looking to which law is more recent would suggest, paradoxically, that Congress’s past efforts to insulate its most important spending priorities were self-defeating. It would also conflict with the GAO’s own guidance to the Executive Branch on how to prioritize spending in certain cases. According to the GAO, entitlement programs should take precedence over annual spending requirements; and, within annual spending, more prescriptive spending programs (“shall”) take precedence over ones that grant discretion (“is authorized”).[193] Resolving the tension between spending statutes and the debt limit by looking to which laws are later in time would flip established practices concerning spending on their head. Good reason, once again, to think that the canon is of limited use.
Recourse to the familiar canon that more specific provisions should govern over more general ones fares little better.[194] The intuition that a narrow appropriations provision is more “specific” than the debt limit statute only holds if one focuses on the question of the government’s outflows rather than inflows. From the perspective of inflows, the debt limit is the more specific statute; appropriations laws (which do not discriminate as to whether they are satisfied by fungible tax dollars and dollars from debt issuance) simply have nothing to say. Perhaps the fairest description of the problem, therefore, is that spending laws and the debt-limit statute simply address different but related issues—government inflows and outflows—rather than the same issue at different levels of generality.
But if these most basic canons of statutory interpretation cannot settle the matter, perhaps practice can—for it is practice that pushes most forcefully and persuasively against the notion that debt-limit legislation must yield to spending legislation. As recounted in Part I, versions of the debt limit have existed since 1790. They have been enacted and reenacted in hundreds of public laws. And there is no known episode of the Executive Branch flouting those limits, with the lone and borderline episodes of Hamilton’s self-acknowledged rogue 1789 loan and the Holland episode of 1791—episodes in which the Executive Branch acknowledged its overstepping and Congress reasserted its prerogatives.[195]
By contrast, it has been understood for centuries that not all spending legislation would be possible to implement on the timeline suggested by that legislation. Some of the earliest revenue-raising statutes, for instance, direct that funds raised shall be applied to spending deficiencies in previous years.[196] The familiar concepts of a “supplemental” or “deficiency” appropriation persist today.[197] Indeed, all three branches have come to understand that spending legislation may be frustrated by all sorts of factors, from legislative miscalculation (like the spending program to which Congress does not appropriate sufficient funds) to unexpected and blameless delay (like the spending program delayed by hurricane). Practice here, once again, seems clear.[198]
The fact that practice seems to lean so strongly in one direction also pushes against a final argument sometimes raised in favor of unilateral debt issuance: that it is the more practical choice.[199] Because unilateral debt issue in violation of a debt limit is something that has never happened, there is uncertainty about how the market will react to it. Consider a situation in which a President orders the Treasury Secretary to issue debt in violation of the statutory limit, and opposing members of Congress proclaim that the issuance is unlawful and will not be honored in the future. The market for such securities would be in the unenviable position of taking sides in a constitutional crisis. And the securities themselves would have a less plausible claim to the protection of the Fourteenth Amendment, since they would not be “authorized by law,” per the scope of the Public Debt Clause.[200]
On the other side of the ledger, the political branches do in fact have substantial experience dealing with deferrals and delays, across many decades, some history of which is recounted above. The debt limit is different as a matter of degree, to be sure, because it would require prioritizing at the level of government spending as a whole, rather than prioritization within a single spending program. That is not a trivial difference. But it seems unrealistic—given the fact that either scenario presents unprecedented details—to suggest that unilateral debt issuance is the obvious practical choice.
This is not to suggest that the frustration of a spending goal is a desirable occurrence. It isn’t. But it is a familiar occurrence—one that the branches have developed tools and practices to deal with. That long history of practice between the branches is a powerful reason for thinking it is spending, and not the debt limit, that should yield.
C. Why the Debt Limit Doesn’t Lead to Default
The preceding Sections have walked through a series of doctrinal arguments. The first is that certain spending requirements may have constitutional priority, including debt service. The second is that the debt-limit statute itself is lawful and that, when the limit conflicts with statutory spending requirements, established practice between the branches suggests that those spending requirements should yield.
This Section turns from law to policy to add an additional point. Default is not just best seen as unconstitutional—a probable violation of the Fourteenth Amendment that the Executive Branch has a duty to avoid—but unlikely: an improbable consequence of what we know about the state of public finance. With some planning, the Treasury will have enough inflowing tax receipts to service payments on the debt.
To be sure, the public is constantly warned that a binding debt limit will lead to default. During the 2013 debt-limit crisis, for example, President Obama repeatedly warned Republicans in Congress to avoid the reckless and irresponsible outcome of a default.[201] His Treasury Secretary Jack Lew has described the limit as a “nihilistic platform for some in Congress to promote the very real risk of a default to advance narrow partisan agendas.”[202] And mainstream press headlines typically describe negotiations over the debt limit as if default were an imminent prospect.[203]
But how real is that risk? It would surely tempt fate to predict that default can never happen. Arguably, America has come close in the past.[204] Binding constraints on the government’s ability to raise emergency funds may not make default less likely. Nevertheless, there are strong practical reasons to think that default can be averted, at least under present circumstances, even if the debt limit binds and requires spending triage.
Annual net interest payments on the national debt—the standard measure of how much debt service costs each year—are a fraction of annual tax revenue. In 2020, for instance, the government’s net interest outlays were $345 billion, and it received about $3.4 trillion in tax revenue; annual debt service amounted to about 10 percent of annual tax revenue.[205] Between 2010 and 2020, this proportion varied between about 9 and 14.5 percent.[206] While reasonable observers can disagree whether debt service constitutes too large a percentage of tax revenue, available data suggests that tax revenue is more than sufficient to pay for debt service.
The timing of the relevant inflows and outflows is more complicated, but again suggests that debt service can be funded by tax revenue with some advance planning. Both federal outlays and receipts are famously “lumpy,” varying from month to month (the lumpiness of receipts and outlays is indeed one reason why federal debt issuance is so useful; in addition to adding to the absolute amount of federal receipts in a year, it also helps smooth out the gaps between when lumps of tax revenue arrive). In Fiscal Year 2021, for instance, monthly tax receipts varied from a low of about $220 billion to a high of about $464 billion; outlays varied from a low of about $364 billion to a high of about $927 billion.[207] But there is no time-period in recent years (and likely much longer) where net interest payments have exceeded tax receipts in a given month, or even come close. Figure 3 displays Treasury’s monthly receipts and net interest expense over the last five years.[208] For debt service to be unaffordable, the average monthly payments would need to increase by an order of magnitude, or tax receipts would need to collapse.
Figure 3: Monthly Treasury Receipts and Net Interest Expense (Billions of Nominal Dollars)
Two additional points may help drive home the point that planning for the debt limit is more feasible than widely appreciated. First, service payments on the debt are predictable; they depend on the government’s preexisting contractual obligations, rather than uncertain individual behavior that generates tax receipts (and certain forms of spending, like Social Security payments). The Treasury can see them coming and plan for the expenditure with precision. Second, the binding debt limit is itself not something that sneaks up. The Treasury typically prepares for its approach months in advance and has a supply of cash on hand to guard against lapses in its borrowing authority. In September of 2021, for example, Secretary Yellen noted that the government would be unable to pay for all spending on October 18, but also noted that the Treasury had been planning for the exigency since August.[209]
Again, this is not to minimize the costs of spending prioritization. But it is to suggest that the central and most dramatic cost mentioned in debates over the debt limits—the likelihood of default—is exaggerated.[210]
The discussion in the previous Sections is not meant to suggest that the costs of reaching the debt limit and prioritizing spending are negligible. Many of those costs—higher borrowing costs, government shutdowns, uncertainty in the markets—have already been realized repeatedly in past showdowns over the limit. Spending prioritization, while so far avoided, appears to be a live risk—and would be a costly one. Even if the Executive Branch can, as a practical matter, successfully prioritize debt service, that is not necessarily true for all entitlement spending—and a month in which Social Security checks failed to go out would certainly be a big deal. But one upshot of the previous discussion is that there are strong doctrinal reasons to think the problems of the statutory debt limit cannot be cured by unilateral Executive action. It will instead take the work of statutory reform.
Imagining that statutory reform should start with one central point: even though the debt limit is lawful and binding, it has become divorced from the original appropriations values that debt authorities were long supposed to serve. Moreover, there is no particularly good reason to think that a numerical limit is necessary to serve those appropriations values.
As described in more detail above, the original debt-issuance authorities—and the limits attached to them—served a dual function. They were partly to keep the outstanding debts of the country at a manageable level. But they also served a basic appropriations function: to ensure that congressional appropriations had sufficient revenue to be satisfied. This basic appropriations function was reflected in the first debt-issuance authority that Congress enacted: authorizing the President, “if he shall deem it necessary, to make such loans as may be requisite to carry into effect the foregoing appropriations.”[211] That focus on ensuring that appropriations could and would be satisfied was a thread that continued in issuance authorities through the Second Liberty Bond Act—legislation that was not just about ensuring that the value of outstanding bonds remained manageable, but also to ensure that sufficient bonds were sold to “meet expenditures for the national security and defense” and to assist “in the prosecution of the war.”[212]
The modern debate over the debt limit has entirely lost sight of the appropriations function that debt authorities are supposed to serve. We think of the debt limit only as, at best, an additional source of bargaining over government spending.[213] We tend not to think that the purpose of giving the Treasury borrowing authority in the first place is to ensure that it “may borrow on the credit of the United States Government amounts necessary for expenditures authorized by law.”[214] If anything, the debt limit now actively frustrates that appropriations purpose by sowing doubt as to whether “expenditures authorized by law” will be satisfied.
But it is not difficult to imagine a debt limit that is more linked to appropriations. It would allow Treasury to borrow “amounts necessary for expenditures authorized by law”—and no more. A debt limit based on appropriations is indeed the most consistent restraint that American statutory borrowing authorities have contained. Some authorities (like the 1790 statute) contain no numerical limit, and there is variation in Congress’s control over timing and interest rates. But there have never been borrowing authorities that contemplate borrowing in excess of what appropriations require. For most of American history, the most common limit on Treasury borrowing was that such borrowing be necessary for lawful expenditures.
Nor is this just ancient history. Between 2013 and the fall of 2021, as noted above, Congress departed from the approach of raising the debt limit by a specific dollar amount every few months or years. Instead, Congress suspended the debt limit until a particular calendar date months or years in the future. Congress did this a total of seven times; the last suspension period expired in the summer of 2021. During these periods, the central limit on the Treasury’s borrowing authority was that such borrowing be “necessary” to pay for required spending.
The background constraint present in Treasury’s borrowing authority—debt can only be issued for “amounts necessary for expenditures authorized by law”—combined with the more than seven years in which we did not have a fixed numerical debt limit, suggest a simple way in which to address the limit: suspend the numerical limit on an indefinite basis, and retain the restriction—in some sense, the original debt limit—that Treasury may only issue debt that is necessary to fund appropriations.
The statutory fix described above might be particularly unrealistic. In the wake of the 2021 debt-limit crisis, after all, Congress reverted to its pre-2013 practice of raising the debt limit by a fixed dollar amount, rather than suspending the limit for a fixed period of time—a shift that leaves the future politics of the debt limit uncertain.[215] For many in Congress, the debt limit serves a valuable strategic purpose, even if not a financial purpose. But even if the numerical limit on outstanding debt remains good law, we need not resign ourselves to a future of such costly legislative brinkmanship. This Section sketches two kinds of middle-ground mechanisms.
First, one could imagine switching the default of which branch needs to act. Instead of setting a fixed limit after which further congressional action is required—difficult to take for granted in an age of gridlock and partisan polarization—one could imagine setting a limit after which executive notice or certification is required, subject to the possibility of future legislative disapproval.
Variations on this theme have also been tried before—though never on a permanent basis. In 2011, for example, Congress created a statutory scheme by which the President himself could, on a temporary basis, raise the debt limit by making a certification to Congress—subject to future congressional disapproval using a fast-track procedure.[216] President Obama submitted certifications using this mechanism in 2011 and 2012 and received no such disapproval.[217] The historical workability of these mechanisms suggests that Congress has a variety of tools available to it when creating law that governs borrowing and the debt limit with a number of potential defaults and triggers that can be used.
Second, even if we accept both the persistence of a limit and legislative brinkmanship as short-run inevitabilities, those episodes of brinkmanship need not be so destructive. A simple way to make them less so would be to provide what statutory law currently lacks: a more explicit statutory framework for what must happen when the debt limit binds and prevents the government from meeting spending obligations.
In this respect, the case of the debt limit may be profitably contrasted with another recurring public finance emergency that can result from congressional gridlock: a lapse in appropriations. A lapse occurs not because Congress has failed to raise the debt limit but because Congress has failed to pass an expected appropriation—thereby depriving the government of some expected lawful authority to spend, though not to raise funds.[218]
But a number of factors make a lapse less challenging as a practical matter and less fraught as a legal matter.[219] One is simply that, during a lapse, the Executive Branch has some direction on how it must and can proceed, in part because Congress has provided a background set of legal rules that apply. While there are many of these scattered across the code, perhaps the most important is the Anti-Deficiency Act’s general prohibition on the government accepting “voluntary services” and “personal services exceeding [those] authorized by law”—with an exception “for emergencies involving the safety of human life or the protection of property.”[220] When combined with the Government Employee Fair Treatment Act[221]—which guarantees that federal employees will be paid after a lapse ends—the Executive Branch has a general rule of thumb. The government can continue to provide certain functions necessary for “the safety of human life or the protection of property,”[222] and employees can provide such services to the government knowing that they will be paid in the future.
As a result, while lapses disrupt much of the ordinary flow of government—they are in no sense desirable or simple—they also do not present the same scope of uncertainties as the debt limit. Those complexities and uncertainties, moreover, are a big source of debt-limit brinkmanship’s financial costs; uncertainty is typically cited as a primary source of market disruption at such junctures.[223]
This need not be the case, and Congress could also create a standard providing some guidance to the Executive Branch on how spending should be prioritized under general conditions of insufficient funds. Such a standard could take many forms and need not be complex. One option would simply be a legislative rule requiring the prioritization of mandatory (entitlement) spending. As noted above, mandatory spending has already been separated from the annual appropriations process—suggesting what is arguably a special solicitude on the part of Congress, a conclusion that seems reflected in GAO’s materials on both prioritization and more broadly. A second option—not mutually exclusive with the first[224]—would be to codify the standard that GAO already uses in its advice and that I have mentioned above: “If the agency cannot get additional funding and the program legislation fails to provide guidance, the agency may, within its discretion, establish reasonable classifications, priorities, and/or eligibility requirements, as long as it does so on a rational and consistent basis.”[225] As noted above, GAO seems to have gestured toward something like this with the debt limit in 1985.[226]
In 1983, in the throes of another crisis over the debt limit, Attorney General William French Smith wrote to Senator Howard Baker with a grim evaluation of the Executive Branch’s legal position:
[W]e know of no comparable instance in which the President has been required to determine his responsibilities under the laws and Constitution of the United States in the event the United States were to run short of cash to respond to the imperatives necessary to continued functioning of the national government.[227]
The argument of this paper is that Smith was both right and wrong. The consequences of a binding debt limit and insufficient revenue may be unprecedented. But they can be informed by 230 years of lawmaking and practice—a history that extends back much farther than previously appreciated and has implications for how we view the limit today.
Those implications have an important dual character. On the one hand, the substance of legal limits on debt issuance has not changed dramatically over the course of American history. That long history, and the (largely) unblemished record of Executive acquiescence, are strong reasons to think the limit binds, even at the expense of competing spending. On the other hand, the connection between borrowing authority and appropriations has become completely attenuated—not because of any intentional high-minded design choice to more scrupulously “limit” debt, but out of a hapless and ill-fated effort to give the Executive Branch more discretion over how and when to borrow. That odd history, and the absence of any good reason to think the numerical limit serves an independent function, are strong reasons to favor change.
-
* Associate Professor of Law, Washington University in St. Louis School of Law. For helpful feedback, I thank Neil Buchanan, Michael Dorf, Daniel Epps, Bill Eskridge, Edward Fox, Anna Gelpern, Henry Hansmann, Daniel Hemel, Dan Koffsky, Zach Liscow, Gillian Metzger, Jon Michaels, Adam Rosenzweig, Kristin Shapiro, and Ryan Watzel, as well as feedback from faculty presentations at Cornell, Emory, Illinois, Michigan, and Washington University in St. Louis. I am also grateful to George Hall for sharing data behind his two papers with Thomas Sargent. Errors are my own. ↑
-
. See Letter from Janet L. Yellen, Sec’y, U.S. Dep’t of the Treasury, to the Hon. Kevin McCarthy, Speaker, U.S. House of Reps. (Jan. 13, 2023), https://home.treasury.gov/system/files/136
/Debt-Limit-Letter-to-Congress-McCarthy-20230113.pdf [https://perma.cc/UBH9-36YD]. ↑ -
. For a summary of those extraordinary measures, see U.S. Dep’t of the Treasury, Description of the Extraordinary Measures (2023), https://home.treasury.gov/system/files/136
/Description_Extraordinary_Measures-2023_01_19.pdf [https://perma.cc/KDG4-VH87]. Secretary Yellen since confirmed that those extraordinary measures had begun. See Letter from Janet L. Yellen, Sec’y, U.S. Dep’t of the Treasury, to the Hon. Kevin McCarthy, Speaker, U.S. House of Reps. (Jan. 19, 2023), https://home.treasury.gov/system/files/136/Debt-Limit-Letter-to-Congress-20230119-McCarthy
.pdf [https://perma.cc/R42E-6X3M]. ↑ -
. See Alan Rappeport, Treasury Expects to Run Out of Cash by June 5, N.Y. Times (May 26, 2023), https://www.nytimes.com/2023/05/26/business/yellen-treasury-x-date-us-debt.html [https://
perma.cc/AVH2-M4PF]. ↑ -
. Fiscal Responsibility Act of 2023, Pub. L. No. 118-5, 137 Stat. 10. ↑
-
. For a description of the first modern debt-limit crisis (in 1953), see Kenneth D. Garbade, Fed. Res. Bank of N.Y., Staff Rep. No. 783, The First Debt Ceiling Crisis (2016), https://www.newyorkfed.org/medialibrary/media/research/staff_reports/sr783.pdf?la=en [https://perma
.cc/LW96-TFPD]. An effort to summarize many of the disruptions since then appears in a Letter from Gregory M. Holloway, Dir. Governmentwide Audits, Gen. Acct. Off., to the Hon. Daniel Patrick Moynihan, Sen., U.S. Senate (Feb. 23, 1996), https://www.gao.gov/assets/aimd-96-49r.pdf [https://perma.cc/GEY6-DN6L]. An account of controversies since 2002 appears in D. Andrew Austin, Cong. Rsch. Serv., RL31967, The Debt Limit: History and Recent Increases (2015). ↑ -
. See Letter from Janet L. Yellen, Sec’y, U.S. Dep’t of the Treasury, to the Hon. Nancy Pelosi, Speaker, U.S. House of Reps. (Sept. 28, 2021), https://home.treasury.gov/system/files/136/Debt-Limit
-Letter-to-Congress-20210928.pdf [https://perma.cc/M5SW-A48B] (reporting that Treasury would exhaust extraordinary measures by October 18, 2021). The first of two public laws raising the limit was signed into law on October 14. See Act of Oct. 14, 2021, Pub. L. No. 117-50, 135 Stat. 407. ↑ -
. Fiscal Responsibility Act of 2023, Pub. L. No. 118-5, 137 Stat. 10. ↑
-
. See Robin Greenwood, Samuel G. Hanson & David Wessel, Preface to The $13 Trillion Question: How America Manages Its Debt, at ix–xi (David Wessel ed., 2016). ↑
-
. See Jacob J. Lew, Managing Our National Debt Responsibly: A Better Way Forward, 54 Harv. J. on Legis. 1, 8–9 (2017); see also U.S. Gov’t Accountability Off., GAO-12-701, Debt Limit: Analysis of 2011–2012 Acts Taken and Effect of Delayed Increase on Borrowing Costs 33 n.9 (2012) (estimating that Treasury’s borrowing costs increased by between $1 billion and $1.7 billion in fiscal year 2011 because of the debt-limit crisis). ↑
-
. Charles Riley, S&P Downgrades U.S. Credit Rating, CNN Money (Aug. 6, 2011, 8:13 PM), http://edition.cnn.com/2011/BUSINESS/08/05/global.economy/index.html [https://perma.cc/L6MV
-U9T2]; see also U.S. Dep’t of the Treasury, The Potential Macroeconomic Effect of Debt Ceiling Brinksmanship 1 (2013), https://home.treasury.gov/system/files/276/POTENTIAL
-MACROECONOMIC-IMPACT-OF-DEBT-CEILING-BRINKMANSHIP.pdf [https://perma.cc
/WED2-X5GS]. ↑ -
. Brett Samuels, Biden Says He Thinks He Has Authority to Use 14th Amendment on Debt Ceiling, Hill (May 21, 2023, 6:54 AM), https://thehill.com/homenews/administration/4014068-biden
-says-he-thinks-he-has-authority-to-use-14th-amendment-on-debt-ceiling/ [https://perma.cc/5JSW
-JDXS]. ↑ -
. Neil H. Buchanan & Michael C. Dorf, How to Choose the Least Unconstitutional Option: Lessons for the President (and Others) from the Debt Ceiling Standoff, 112 Colum. L. Rev. 1175, 1196–97 (2012) [hereinafter The Least Unconstitutional Option]. ↑
-
. See Bill White, America’s Fiscal Constitution: Its Triumph and Collapse ix (2014) (describing certain “well-defined principles” concerning debt management as part of an “unwritten fiscal constitution”); see also William A. Niskanen, The Case for a New Fiscal Constitution, 6 J. Econ. Persps. 13, 13–14 (1992). ↑
-
. See Public Debt Act of 1941, ch. 7, 55 Stat. 7; Second Liberty Bond Act, ch. 56, 40 Stat. 288 (1917). ↑
-
. See, e.g., D. Andrew Austin, Cong. Rsch. Serv., R43389, The Debt Limit Since 2011, at 1 (2021) (“Before 1917 Congress typically controlled individual issues of debt.”); Robert T. Patterson, Federal Debt-Management Policies, 1865–1879, at 84 (1954) (“[I]t was considered the duty of Congress to determine the interest rates and maturities of the government’s obligations. It had always been the custom for the legislative body to decide these terms, and not to leave them to the discretion of the Secretary of the Treasury.”); Jacob Holt, The Power to Borrow Money, in The Powers of the U.S. Congress: Where Constitutional Authority Begins and Ends 23 (Brien Hallett ed., 2016) (“Until 1917 Congress had no statutory limit on federal debt.”); White, supra note 13, at 12 (“[F]or 180 years the federal government never set out to borrow to support routine spending, though at times it did borrow quite a bit for . . . extraordinary purposes . . . .”); George J. Hall & Thomas J. Sargent, Brief History of US Debt Limits Before 1939, 115 Proc. Nat’l Acad. Sciences. 2942, 2942 (2018) (“Before 1917, Congress designed all federal securities.”); Anita S. Krishnakumar, In Defense of the Debt Limit Statute, 42 Harv. J. on Legis. 135, 135–36, 143 (2005) (explaining that “[p]rior to the [1917] statute’s passage, Congress . . . approved each individual issuance of debt made on the nation’s behalf” and “borrowed only when faced with the exigent circumstances of war or economic recession”); Linda K. Kowalcky & Lance T. LeLoup, Congress and the Politics of Statutory Debt Limitation, 53 Pub. Admin. Rev. 14, 15 (1993) (“Congress first enacted legislation to statutorily limit the borrowing authority of the U.S. Treasury as part of the Second Liberty Loan Act in 1917 . . . .”). ↑
-
. Act of Mar. 26, 1790, ch. 4, § 7, 1 Stat. 104, 105–06. ↑
-
. See 31 U.S.C. § 3102(a) (bonds). Identical or similar language is repeated in the borrowing provisions that authorize the issuance of Treasury notes and certificates of indebtedness. See 31 U.S.C. §§ 3103(a), 3104(a). ↑
-
. The Least Unconstitutional Option, supra note 12, at 1195. ↑
-
. Walter W. Heller, Why A Federal Debt Limit?, 51 Nat’l Tax Ass’n 246, 252 (1958) (internal quotations omitted). ↑
-
. U.S. Const. amend. XIV, § 4. ↑
-
. See generally Curtis A. Bradley & Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411 (2012) (describing and evaluating “historical gloss” as a tool of constitutional interpretation that has special prevalence in the context of the separation of powers). ↑
-
. In common parlance, terms “debt limit” and “debt ceiling” are also used interchangeably. In this Article, I typically refer to the “debt limit” because that usage appears slightly more common (based on Google) and because that is what Congress itself has used in the primary operative provision. See 31 U.S.C. § 3101 (“Public debt limit”). But “debt ceiling” is also a common usage and is likewise used in some statutory law. See 31 U.S.C. § 3101A (“Presidential modification of the debt ceiling”). ↑
-
. U.S. Const. art. I, § 8, cl. 2. ↑
-
. See generally The Least Unconstitutional Option, supra note 12. ↑
-
. See Act of Sept. 13, 1982, Pub. L. No. 97-258, 96 Stat. 877, 878 (codified at 31 U.S.C. § 3101). ↑
-
. See, e.g., Act of July 8, 1870, ch. 229, 16 Stat. 197, 198 (directing an issuance to refund interest paid by the State of Massachusetts for expenditures during the War of 1812); Act of Sept. 9, 1850, ch. 49, 9 Stat. 446, 447 (requiring the payment of $10 million in stock to Texas); Act of Aug. 10, 1846, ch. 175, 9 Stat. 85, 94 (mandating stock issuance “[f]or paying the principal and interest of the fourth and fifth instalments of the Mexican indemnities”). ↑
-
. See Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 472 (2001) (describing this nondelegation jurisprudence). This is a claim about present doctrine; there is an active scholarly and judicial debate about what delegation doctrine should look like and did look like at the Founding. See, e.g., Julian Davis Mortenson & Nicholas Bagley, Delegation at the Founding, 121 Colum. L. Rev. 277 (2021) (rebutting the claim that the Constitution was originally understood to contain a nondelegation doctrine); Larry Alexander & Saikrishna Prakash, Reports of the Nondelegation Doctrine’s Death Are Greatly Exaggerated, 70 U. Chi. L. Rev. 1297 (2003); see also West Virginia v. EPA, No. 20–1530, slip op. at 17 n.6 (U.S. June 30, 2022) (Gorsuch, J., concurring) (citing law review articles); id. at 29 (Kagan, J., dissenting) (same). ↑
-
. See generally 31 U.S.C. §§ 3101–3113 (describing borrowing authorities and limits). ↑
-
. 31 U.S.C. § 3104(b)–(c). ↑
-
. See, e.g., Act of Mar. 2, 1811, ch. 32, 2 Stat. 656 (“An Act authorizing a loan of money, for a sum not exceeding five millions of dollars.”); Act of Mar. 26, 1790, ch. 4, 1 Stat. 104, 105–06 (authorizing the President to empower the Secretary of Treasury to make loans). ↑
-
. See Act of Mar. 17, 1971, Pub. L. No. 92-5, § 3, 85 Stat. 5, 5 (allowing long-term borrowing with an interest rate of greater than 4.25%). ↑
-
. See S. Rep. No. 92-28, at 9 (1971) (“The Treasury Department proposed that the present 4¼ percent interest rate ceiling applicable to bonds with maturities of more than 7 years be eliminated on the grounds that the ceiling has, in recent years, impeded Treasury financing, interfered with prudent debt management and had an adverse impact on financial markets. . . . [The interest rate limit] forced the Treasury for the last several years to concentrate its financing entirely in the short end of the market where there is no interest rate limitation. . . . The consequence of this is that the Treasury must enter the market more and more frequently to refinance the maturing debt.”). ↑
-
. See, e.g., 31 U.S.C. § 3104(a). Substantially similar restrictions appear in each provision from § 3102 to § 3106. ↑
-
. See Kate Stith, Congress’ Power of the Purse, 97 Yale L.J. 1343, 1345 (1988). ↑
-
. See Gillian E. Metzger, Taking Appropriations Seriously, 121 Colum. L. Rev. 1075, 1085–86 n.35 (2021) (noting relationship between appropriations and revenue-raising). ↑
-
. The Treasury has attempted in recent decades to stick to a so-called “regular and predictable” strategy of debt sales. See Kenneth D. Garbade, The Emergence of “Regular and Predictable” as a Treasury Debt Management Strategy, 2007 FRBNY Econ. Pol’y Rev. 53. ↑
-
. For example, in March of 2020, at beginning of the COVID-19 pandemic, the Treasury had $1.6 trillion on hand as Congress contemplated the Coronavirus Aid, Relief, and Economic Security Act. See U.S. Gov’t Accountability Off., GAO-21-606, Federal Debt Management: Treasury Quickly Financed Historic Government Response to the Pandemic and Is Assessing Risks to Market Functioning (2021); see also Christine Desan, How to Spend a Trillion Dollars: Our Monetary Hardwiring, Why It Matters, and What We Should Do About It (Harv. Pub. L., Working Paper No. 22–04, 2022), https://ssrn.com/abstract=4056241 [https://perma.cc/4M73-UWGU]. ↑
-
. Because revenue is fungible, authorizing borrowing for a particular program will free up funds that can be used for other government spending. See Conor Clarke & Edward Fox, No New Tax Cuts? Examining the Rescue Plan’s New State Tax Limits, 103 Tax Notes St. 1361, 1364 & n.24 (2022) (making this point in the context of federal support grants to states). ↑
-
. See Act of Aug. 5, 1909, ch. 6, § 39, 36 Stat. 11, 117; Act of June 28, 1902, ch. 1302, 32 Stat. 481. The fact that these borrowing authorities closely preceded the 1917 Second Liberty Bond Act may also explain the common misconception that all borrowing authorities took this form. ↑
-
. See generally Tax Policy Center Briefing Book: The State of State (and Local) Tax Policy, Tax Pol’y Ctr. (May 2020), https://www.taxpolicycenter.org/briefing-book/what-are-state-balanced
-budget-requirements-and-how-do-they-work [https://perma.cc/U8UT-WNL2]. ↑ -
. For discussions of some international contrasts, see Sarah Love, How to Improve the Debt Ceiling to Fit a Partisan Government: A Global Examination of Which International Solutions Excel, 25 Ind. J. Glob. Legal Stud. 775 (2018); Michael D. Arena, Case Note, The Legal Frameworks Governing Sovereign Debt and Borrowing in the United States and European Union, 20 Colum. J. Eur. L. 283 (2014). ↑
-
. For some historical discussion, see Richard Sylla, Financial Foundations: Public Credit, the National Bank, and Securities Markets, in Founding Choices: American Economic Policy in the 1790s, at 59 (Douglas A. Irwin & Richard Sylla eds., 2010). See also Richard C. Schragger, Democracy and Debt, 121 Yale L.J. 860, 865–68 (2012) (describing structural mechanisms to control fiscal behavior). For an influential theoretical discussion, see Douglas W. Diamond, Reputation Acquisition in Debt Markets, 97 J. Pol. Econ. 828 (1989). ↑
-
. See Pauline Maier, Ratification: The People Debate the Constitution: 1787–1788, at 11 (2010); see also Max M. Edling, A Hercules in the Cradle: War, Money, and the American State, 1783–1867, at 81–107 (2014) (describing controversies over debt assumption in the 1790s). ↑
-
. See Articles of Confederation of 1777, art. IX, para. 6 (“The United States in Congress assembled shall never . . . borrow money on the credit of the United States . . . unless nine states assent to the same . . . .”). ↑
-
. The Federalist No. 30 (Alexander Hamilton). ↑
-
. Id. ↑
-
. U.S. Const. art. VI. ↑
-
. Act of Aug. 4, 1790, ch. 34, 1 Stat. 138. ↑
-
. See generally 2 The Records of the Federal Convention of 1787, at 412–14 (Max Farrand ed., 1911) (discussing the borrowing authority). The delegates considered adding to Article I the following: “That funds which shall be appropriated for payment of public Creditors shall not during the time of such appropriation be diverted or applied to any other purpose . . . .” Id. at 322. Madison’s notes reflect worry about this proposal: “Mr. Mason was much attached to the principle, but was afraid such a fetter might be dangerous in time of war.” Id. at 326. ↑
-
. Hamilton, supra note 45. ↑
-
. See, e.g., 2 Joseph Story, Commentaries on the Constitution of the United States § 1051, at 503 (1833) (“[T]he experience of all nations must convince us, that the burthen and expenses of one year, in time of war, may more than equal the ordinary revenue of ten years. Hence, a debt is almost unavoidable, when a nation is plunged into a state of war.”). ↑
-
. Brutus, Brutus VIII (1788); see also Luther Martin, Genuine Information VI (1788) (“doubting whether if a war should take place it would be possible for this country to defend itself without having recourse to paper credit”). ↑
-
. Brutus, supra note 52. ↑
-
. See Douglass C. North & Barry R. Weingast, Constitutions and Commitment: The Evolution of Institutions Governing Public Choice in Seventeenth-Century England, 49 J. Econ. Hist. 803, 810 (1989); see also David Stasavage, Partisan Politics and Public Debt: The Importance of the ‘Whig Supremacy’ for Britain’s Financial Revolution, 11 Eur. Rev. Econ. Hist. 123 (2007). ↑
-
. Edmund W. Kitch & Julia D. Mahoney, Restructuring United States Government Debt: Private Rights, Public Values, and the Constitution, 2019 Mich. St. L. Rev. 1283, 1291 (noting defaults in Spain as another influence). ↑
-
. See Barry Eichengreen, Asmaa El-Ganainy, Rui Esteves & Kris James Mitchener, In Defense of Public Debt 31 (2021) (noting the empirical regularity that “[a]bsolute monarchs with unchecked powers found borrowing more difficult”). ↑
-
. See Krishnakumar, supra note 15, at 141 (stating that, after 1789, “Congress did not cause the nation to incur debt again until 1812, when the still-fledgling United States once more went to war with England”). ↑
-
. There are two incomplete sources listing and describing legal authorities to borrow. The first: A.T. Huntington & Robert J. Mawhinney, Laws of the United States Concerning Money, Banking, and Loans, 1778–1909, S. Doc. No. 580 (2d Sess. 1910), which provides a compilation of many early borrowing authorities. The second is Rafael A. Bayley, The National Loans of the United States, from July 4, 1776, to June 30, 1980 (1881), which records all U.S. debt issuance between 1776 and 1880. For each issuance, Bayley records the date of the authorizing act, allowing relatively easy reference to the statutes at large. Both Bayley and Huntington and Mawhinney are somewhat underinclusive; Bayley does not summarize borrowing authorities that were never used; Huntington and Mawhinney provide a broad compilation of statutes related to national finance, but miss some borrowing authorities that are not so focused. I supplement the compilations in Bayley and Huntington and Mawhinney with organic searches of the statutes at large (reviewing for instance all statutes that refer to borrowing, loans, notes, bonds, etc.), finding a total of eighty-three statutes that grant a discretionary borrowing authority to the Executive Branch between 1790 and 1910. In cases where a borrowing authority allocates authority across several years, I allocate the authority accordingly. ↑
-
. The great majority of laws include the authorized borrowing amount on their face; an act “authorizing a loan of one million of dollars” is, indeed, an act authorizing a loan of $1 million dollars. See Act of Mar. 20, 1794, ch. 8, 1 Stat. 345, 345. In cases where the borrowing authority does not contain an amount, the amount can typically be calculated using other information from the statute. For example, the appropriations act of 1790 authorizes the President and the Secretary of the Treasury “to make such loans as may be requisite to carry into effect the foregoing appropriations”; the borrowing authority can be obtained by summing the appropriated amounts. See Act of Mar. 26, 1790, ch. 4, § 7, 1 Stat. 104, 105–06. In rare cases where that amount is less easily obtained (for example, a June 1898 authorization that the President may borrow “such sum or sums as, in his judgment, may be necessary to meet public expenditures”), I used the amount that was actually issued under the authority (in that case, $400 million). Act of June 13, 1898, ch. 448, 30 Stat. 448, 467. ↑
-
. The complete list of those authorities is as follows: Act of Mar. 14, 1812, ch. 41, 2 Stat. 694; Act of Mar. 2, 1811, ch. 32, 2 Stat. 656; Act of May 1, 1810, ch. 45, 2 Stat. 610; Act of Feb. 13, 1806, ch. 5, 2 Stat. 349; Act of Mar. 26, 1804, ch. 46, 2 Stat. 291; Act of Nov. 10, 1803, ch. 3, 2 Stat. 247; Act of Feb. 26, 1803, ch. 8, 2 Stat. 202; Act of Apr. 29, 1802, ch. 32, 2 Stat. 167; Act of May 7, 1800, ch. 42, 2 Stat. 60; Act of Mar. 2, 1799, ch. 31, 1 Stat. 725; Act of July 16, 1798, ch. 84, § 2, 1 Stat. 609, 609–10; Act of July 16, 1798, ch. 79, 1 Stat. 607; Act of July 8, 1797, ch. 16, 1 Stat. 534; Act of June 1, 1796, ch. 51, 1 Stat. 493; Act of May 31, 1796, ch. 44, 1 Stat. 488; Act of May 30, 1796, ch. 41, 1 Stat. 487; Act of May 6, 1796, ch. 21, 1 Stat. 461; Act of Mar. 3, 1795, ch. 46, § 6, 1 Stat. 438, 439; Act of Mar. 3, 1795, ch. 45, 1 Stat. 433; Act of Dec. 18, 1794, ch. 4, § 1, 1 Stat. 404, 404; Act of June 9, 1794, ch. 63, 1 Stat. 394; Act of June 5, 1794, ch. 46, 1 Stat. 376; Act of Mar. 20, 1794, ch. 8, 1 Stat. 345, 345; Act of Mar. 20, 1794, ch. 7, 1 Stat. 345; Act of Feb. 28, 1793, ch. 18, 1 Stat. 325; Act of May 8, 1792, ch. 41, 1 Stat. 284; Act of May 2, 1792, ch. 27, 1 Stat. 259; Act of Mar. 3, 1791, ch. 16, 1 Stat. 214; Act of Aug. 12, 1790, ch. 47, § 4, 1 Stat. 186, 187; Act of Aug. 4, 1790, ch. 34, § 2, 1 Stat. 138, 139; Act of Mar. 26, 1790, ch. 4, § 7, 1 Stat. 104, 105–06. ↑
-
. Act of Feb. 28, 1793, ch. 18, § 2, 1 Stat. 325, 328. ↑
-
. See, e.g., Act of July 8, 1797, ch. 16, § 1, 1 Stat. 534, 534; Act of Dec. 18, 1794, ch. 4, § 1, 1 Stat. 404, 404; Act of Mar. 20, 1794, ch. 8, 1 Stat. 345, 345. ↑
-
. See Bayley, supra note 58, at 104–30 (listing early debt issuance under early appropriations statutes). ↑
-
. Act of Mar. 24, 1814, ch. 29, § 1, 3 Stat. 111, 111 (authorizing a loan of up to $25 million “to defray any expenses which have been, or during the present year may be authorized by law, and for which appropriations have been, or during the present year may be made by law”); Act of Aug. 2, 1813, ch. 51, § 1, 3 Stat. 75, 75–76 (authorizing a loan of up to $17.5 million “to defray any expenses which have been or which may be authorized for the service of the years one thousand eight hundred and thirteen and one thousand eight hundred and fourteen, and for which appropriations have been or may be made by law during those years”); Act of Feb. 8, 1813, ch. 21, § 1, 2 Stat. 798, 798 (authorizing a loan of up to $16 million “to defray any of the expenses which have been, or, during the present session of Congress, may be authorized by law, and for which appropriations have been, or, during the present session of Congress, may be made by law”). ↑
-
. See, e.g., Act of Jan. 28, 1847, ch. 5, § 1, 9 Stat. 118, 118 (authorizing presidential borrowing of “such sum or sums as the exigencies of the government may require” up to $23 million). ↑
-
. See, e.g., Act of July 17, 1861, ch. 5, § 1, 12 Stat. 259, 259 (authorizing the Treasury Secretary to borrow “a sum not exceeding two hundred and fifty millions of dollars, or so much thereof as he may deem necessary for the public service”). ↑
-
. See Second Liberty Bond Act, ch. 56, § 1, 40 Stat. 288, 288 (1917) (authorizing borrowing for national defense but also for “other public purposes authorized by law”); Act of Apr. 24, 1917, ch. 4, § 1, 40 Stat. 35, 35 (same). ↑
-
. Act of June 13, 1898, ch. 448, 30 Stat. 448, 467 (“[T]he Secretary of the Treasury is hereby authorized to borrow on the credit of the United States from time to time as the proceeds may be required to defray expenditures authorized on account of the existing war (such proceeds when received to be used only for the purpose of meeting such war expenditures) the sum of four hundred million dollars, or so much thereof as may be necessary.”). ↑
-
. The “project finance” view is that debt issuance was “a vehicle for financing specific projects,” rather than revenue more generally. Kenneth D. Garbade, Birth of a Market: The U.S. Treasury Securities Market from the Great War to the Great Depression 29 (2012); Act of Aug. 5, 1909, ch. 6, § 39, 36 Stat. 11, 117 (authorizing borrowing of up to $295 million “as the proceeds may be required to defray expenditures on account of the Panama Canal and to reimburse the Treasury for such expenditures already made and not covered by previous issues of bonds”); Act of June 28, 1902, ch. 1302, § 8, 32 Stat. 481, 484 (providing for the construction of the canal and authorizing borrowing of up to $130 million “as the proceeds may be required to defray expenditures authorized by this Act (such proceeds when received to be used only for the purpose of meeting such expenditures)”). ↑
-
. Cf. Clarke & Fox, supra note 38, at 1364 (discussing these fungibility points in the context of state grants). ↑
-
. See Krishnakumar, supra note 15, at 141. ↑
-
. Act of May 7, 1800, ch. 42, § 1, 2 Stat. 60, 60. ↑
-
. Act of Mar. 31, 1840, ch. 5, § 2, 5 Stat. 370, 370 (“[U]nder the regulations and provisions contained in said [earlier] act, Treasury Notes may be issued in lieu of others hereafter or heretofore redeemed, but not to exceed in the amount of notes outstanding at any one time, the aggregate of five millions of dollars . . . .”). ↑
-
. See, e.g., Act of Mar. 3, 1843, ch. 81, 5 Stat. 614; Act of Feb. 8, 1861, ch. 29, 12 Stat. 129; Act of July 22, 1846, ch. 64, 9 Stat. 39. Another chain of referencing borrowing authorities occurred later. See Act of Mar. 3, 1917, ch. 159, § 401, 39 Stat. 1000, 1003 (amending Act of Aug. 5, 1909, ch. 5, 36 Stat. 11, 117–18, so the value of certificates of indebtedness that can be outstanding at one time can be $300 million rather than $200 million; Act of Aug. 5, 1909, ch. 5, 36 Stat. 11, 117–18, in turn amended Act of June 13, 1898, ch. 448, § 32, 30 Stat. 448, 466–67, so that the amount outstanding can be $200 million instead of $100 million). ↑
-
. To see this, compare two laws that have the same effect: “no more than X one-year securities may be outstanding” and “no more than X one-year securities may be issued each year.” ↑
-
. Act of Mar. 3, 1795, ch. 46, § 6, 1 Stat. 438, 439. ↑
-
. Act of May 7, 1800, ch. 42, § 1, 2 Stat. 60, 60. ↑
-
. See, e.g., Act of July 14, 1870, ch. 256, 16 Stat. 272. ↑
-
. Act of Mar. 17, 1971, Pub. L. No. 92-5, § 3, 85 Stat. 5. ↑
-
. Act of June 13, 1898, ch. 448, § 32, 30 Stat. 448, 466. ↑
-
. Act of May 1, 1810, ch. 45, § 1, 2 Stat. 610, 610. ↑
-
. 31 U.S.C. § 3104(a). ↑
-
. See, e.g., Act of July 16, 1798, ch. 84, § 2, 1 Stat. 609, 609–10. ↑
-
. See Act of Feb. 8, 1813, ch. 21, § 1, 2 Stat. 798, 798 (“[I]t shall be deemed a good execution of the said power to borrow, for the President of the United States to cause to be sold the whole or any part of the certificates of stock issued for the sums to be borrowed by virtue of this act.”); Act of Mar. 2, 1811, ch. 32, § 1, 2 Stat. 656, 656 (“[I]t shall be deemed a good execution of the said power to borrow, for the Secretary of the Treasury, with the approbation of the President of the United States, to cause to be constituted certificates of stock . . . .”). ↑
-
. See, e.g., Act of July 17, 1861, ch. 5, § 8, 12 Stat. 259, 261 (requiring a report to Congress on the exercise of the borrowing authority); Act of June 14, 1858, ch. 165, § 3, 11 Stat. 365, 365 (“[T]he said Secretary shall report to Congress, at the commencement of the next session, the amount of money borrowed under this act, and of whom, and on what terms . . . .”); Act of Apr. 15, 1842, ch. 26, § 6, 5 Stat. 473, 474 (requiring a report to Congress). ↑
-
. See Bayley, supra note 58, at 104–73 (listing all known early issuances since 1790 with statutory authority for each); see also Hall & Sargent, supra note 15, at 2943 (constructing an implicit debt limit from individual statutes and showing no years in which U.S. debt exceeds the limit). ↑
-
. Though in certain circumstances the temporary debt limit has been exceeded when it ended and dropped back to the permanent level. See Cong. Rsch. Serv., R45011, Clearing the Air on the Debt Limit i (2021) (“In October 1977 and April 1979, a lapse in temporary debt limit increases left federal debt above its limit for a few days.”). ↑
-
. This was often, though not always, true of early debt-authorities used to fund appropriations in the first decade. (Of debt-issuance authorities for appropriations in the first decade, I have found six cases in which the Executive Branch issued at the relevant maximum.). ↑
-
. Bayley, supra note 58, at 108. ↑
-
. Id. ↑
-
. See Alexander Hamilton, Report on Supplementary Appropriations for the Civil List for 1790, in 6 The Papers of Alexander Hamilton 280, 281 (Harold C. Syrett & Jacob E. Cooke eds., 1962). ↑
-
. Id. ↑
-
. See Act of Aug. 4, 1790, ch. 34, § 2, 1 Stat. 138, 139. ↑
-
. See Act of Aug. 12, 1790, ch. 47, § 4, 1 Stat. 186, 187. ↑
-
. See Bayley, supra note 58, at 104–07 (summarizing loans). ↑
-
. See Act of Mar. 3, 1791, ch. 25, 1 Stat. 218, 218–19. ↑
-
. Id. at 219. ↑
-
. Id. (emphasis omitted). ↑
-
. See Garbade, supra note 69, at 313. ↑
-
. See Second Liberty Bond Act, ch. 56, 40 Stat. 288 (1917). ↑
-
. Id. §§ 1, 6. ↑
-
. The first reference I can find in the congressional record with a reference to “the debt limit” (referring to the federal statutory limit) is at 79 Cong. Rec. 3843 (1935). ↑
-
. U.S. Dep’t of the Treasury, Annual Report of the Secretary of the Treasury on the State of the Finances, Doc. No. 3024, at 39 (1930). ↑
-
. Id. ↑
-
. Amendment to Second Liberty Bond Act, ch. 336, 53 Stat. 1071 (1939). ↑
-
. See Public Debt Act of 1941, ch. 7, 55 Stat. 7, 7. ↑
-
. See, e.g., 84 Cong. Rec. 6470, 6480, 6499–500 (1939) (statement of Sen. Pat Harrison) (“If the Treasury authorities . . . feel they can issue long-term Government bonds more advantageously to the Government than they can issue short-term paper, I think we ought to give them the authority. That is all this bill does.”). ↑
-
. Id. at 6480 (statement of Sen. Arthur Vandenberg); see also 84 Cong. Rec. 2684, 2714 (1939) (statement of Sen. William H. King) (“We should address ourselves, as I have indicated, to reducing expenses.”). ↑
-
. See Act of Mar. 17, 1971, Pub. L. No. 92-5, § 3, 85 Stat. 5 (allowing long-term borrowing with an interest rate of greater than 4.25%); see also S. Rep. No. 92-28, at 9 (1971) (describing the problems of the interest-rate limit). ↑
-
. The precise reasons for this approach are somewhat murky, but it appears to have been motivated by the view that federal spending would not continue to increase, and that temporary changes in the debt could provide a buffer to account for uncertainty about receipts and expenditures in a coming fiscal year. The legislative history of the first temporary debt limit suggests that it was motivated by “anticipation of a more even distribution of revenue collections and reduction of expenditures in future years.” S. Rep. No. 83-2225, at 2 (1954). Federal spending did not, of course, decrease, and the temporary debt created a number of problems. When federal spending increases faster than federal tax revenues such that the government is operating in the margin between the permanent and temporary debt limits—as it did during almost the entire period in which the temporary debt limit existed—the expiration of a temporary limit immediately restricted the government’s ability to borrow. ↑
-
. See Bipartisan Budget Act of 2019, Pub. L. No. 116-37, § 301(a), 133 Stat. 1049, 1057 (suspending application of 31 U.S.C. § 3101(b) until July 31, 2021); Bipartisan Budget Act of 2018, Pub. L. No. 115-123, § 30301, 132 Stat. 64, 132 (suspending application of 31 U.S.C. § 3101(b) until March 1, 2019); Continuing Appropriations Act, 2018 and Supplemental Appropriations for Disaster Relief Requirements Act, 2017, Pub. L. No. 115-56, § 101, 131 Stat. 1129, 1139 (suspending application of 31 U.S.C. § 3101(b) until December 8, 2017); Bipartisan Budget Act of 2015, Pub. L. No. 114-74, § 902, 129 Stat. 584, 621; Temporary Debt Limit Extension Act, Pub. L. No. 113-83, §§ 2–3, 128 Stat. 1011, 1011–12 (2014) (suspending application of 31 U.S.C. § 3101(b) until March 15, 2015); Default Prevention Act of 2013, Pub. L. No. 113-46, § 1002, 127 Stat. 558, 566–67; No Budget, No Pay Act of 2013, Pub. L. No. 113-3, § 2, 127 Stat. 51, 51 (suspending application of 31 U.S.C. § 3101(b) until May 18, 2013). ↑
-
. See Act of Sept. 29, 1979, Pub. L. No. 96-78, § 201(a), 93 Stat. 589, 589–91. ↑
-
. Id. at 589–90. ↑
-
. Austin, supra note 15, at 4; see also 165 Cong. Rec. H25 (daily ed. Jan. 3, 2019) (statement of Rep. Steve Womack) (noting that the 2019 rules package “includes a new iteration of the Gephardt rule . . . that makes increasing the debt limit even easier by automatically passing debt limit increases without separate debate and vote”). ↑
-
. House Rule XXVIII (116th Cong.). ↑
-
. House Rule XXVIII (117th Cong.). ↑
-
. There were forty-seven changes to the debt limit during that same period, and the Gephardt Rule allowed the House to avoid a separate and direct vote on ten of them. See Bill Heniff Jr., Cong. Rsch. Serv., RL 31913, Debt Limit Legislation: The House “Gephardt Rule” 5 (2023). ↑
-
. There is an enormous literature developing many interesting angles on these and related issues. For some theorizing on why the modern western welfare state has been financed in relatively greater part by borrowing rather than current revenues (and offering reasons including ageing populations, political fractionalization, and electoral incentives), see Eichengreen et al., supra note 56, at 108–10. For reflections on the general tradeoffs between debt-financing and financing through taxation, see Daniel Shaviro, Do Deficits Matter? (1997). For the political economy questions, see, for example, Yuki Uchida & Tetsuo Ono, Political Economy of Taxation, Debt Ceilings, and Growth, Eur. J. Pol. Econ., June 2021; Kowalcky & LeLoup, supra note 15. ↑
-
. See Hall & Sargent, supra note 15, at 2943. ↑
-
. See generally Donald R. Stabile & Jeffrey A. Cantor, The Public Debt of the United States: An Historical Perspective, 1775–1990 (1991) (describing the history of these fluctuations in the nineteenth century). ↑
-
. See Sylla, supra note 42, at 73 (describing the growth in early federal revenues); John A. James, Public Debt Management Policy and Nineteenth-Century American Economic Growth, 21 Expls. Econ. Hist. 192 (1984). ↑
-
. Kitch & Mahoney, supra note 55, at 1296 (describing fluctuations in debt in the nineteenth century); see also Carl Lane, A Nation Wholly Free: The Elimination of the National Debt in the Age of Jackson 161 (2014). ↑
-
. In the late eighteenth and nineteenth centuries, for example, the United States had sinking funds to redeem the entirety of the public debt. See Act of Mar. 3, 1795, ch. 45, 1 Stat. 433; see also Act of Apr. 29, 1802, ch. 32, 2 Stat. 167. ↑
-
. See, e.g., Kitch & Mahoney, supra note 55, at 1297 (“[T]he 1920s marks the final era of significant United States debt retirement.”). There have been brief periods of recent national surplus, such as the end of the Clinton Administration and the beginning of the Bush Administration. But only brief. ↑
-
. See, e.g., Act of Mar. 3, 1791, ch. 16, 1 Stat. 214. ↑
-
. See, e.g., Act of June 28, 1902, ch. 1302, 32 Stat. 481. ↑
-
. A related legal trend is that much of government spending does not come through the annual appropriations process but is instead “mandatory” spending. See, e.g., Christine Bogusz, Dan Ready & Jorge Salazar, Cong. Budget Off., The Federal Budget in Fiscal Year 2020 (2021), https://www.cbo.gov/system/files/2021-04/57171-mandatory-spending.pdf [https://perma.cc/9DAX
-AFGL]. ↑ -
. See Garbade, supra note 69, at 314–15 (summarizing efforts). ↑
-
. Id. ↑
-
. U.S. Dep’t of the Treasury, supra note 103, at 39. ↑
-
. Id. ↑
-
. Id. ↑
-
. Jennifer Epstein, Bill Clinton: I’d Use 14th Amendment, Politico (July 19, 2011, 12:28 PM), https://www.politico.com/story/2011/07/bill-clinton-id-use-14th-amendment-059331 [https://perma.cc
/YTB5-LUJY]; see also Victor Williams, Raze the Debt Ceiling: A Test Case for State-Sovereign and Institutional Bondholder Litigation to Void the Debt Limit Statute, 72 Wash. & Lee L. Rev. Online 96, 111–12 (2015) (“On its face, and as it is arbitrarily applied, the debt limit statute violates . . . the Fourteenth Amendment’s Public Debt Clause . . . .”). ↑ -
. See The Least Unconstitutional Option, supra note 12. ↑
-
. U.S. Const. amend. XIV, § 4. ↑
-
. See Complaint at 1, Williams v. Lew, 77 F. Supp. 3d 129 (D.D.C. 2015) (No. 14-00183) (quoting U.S. Const. amend. XIV, § 4). ↑
-
. See Williams v. Lew, 819 F.3d 466, 470–71 (D.C. Cir. 2016) (summarizing procedure). The government appears not have briefed the merits of Williams’s constitutional claim and instead relied entirely on jurisdictional arguments. ↑
-
. Williams, supra note 133, at 111–12 (“On its face, and as it is arbitrarily applied, the debt limit statute violates . . . the Fourteenth Amendment’s Public Debt Clause . . . .”); see also Jacob D. Charles, Note, The Debt Limit and the Constitution: How the Fourteenth Amendment Forbids Fiscal Obstructionism, 62 Duke L.J. 1227, 1232 (2013) (“[W]hen the Public Debt Clause is violated by congressional actions that place the debt’s validity in substantial doubt, the president can refuse to enforce . . . the debt limit and order the Treasury Secretary to continue borrowing funds to meet the government’s obligations.” (footnote omitted)). ↑
-
. See, e.g., Kitch & Mahoney, supra note 55, at 1287 (“[I]t is far from clear—at least to us—that payments on treasury obligations must be accorded the highest level of protection.”). ↑
-
. The debt had increased forty-fold and was broadly held by the northern public. See Charles W. Calomiris, The Motives of U.S. Debt-Management Policy, 1790–1880: Efficient Discrimination and Time Consistency, 13 Rsch. Econ. Hist. 67, 73 (1991) (describing growth of federal debt during the war). Many bondholders feared that, if southerners were to regain control of or substantial influence in Congress, they would repudiate the war debts—a fear on Congress’s mind when it established the Joint Committee on Reconstruction. See, e.g., Cong. Globe, 39th Cong., 1st Sess. 10 (1865). ↑
-
. U.S. Const. amend. XIV, § 4 (emphasis added). ↑
-
. Cong. Globe, 39th Cong., 1st Sess. 2938 (1866). There were three introduced versions of what became the Public Debt Clause. First, Senator Benjamin Wade proposed the following: “The public debt of the United States, including all debts or obligations which have been or may hereafter be incurred in suppressing insurrection or in carrying on war in defense of the Union, or for payment of bounties or pensions incident to such war and provided for by law, shall be inviolable.” Id. at 2768. Second, Senator Jacob Merritt Howard, who led a subcommittee that focused on the question of repudiation, introduced an alternative: “The obligations of the United States, incurred in suppressing insurrection, or in defense of the Union, or for payment of bounties or pensions incident thereto, shall remain inviolate.” Id. at 2938. Third, and finally, the Senator Daniel Clark proposed what was eventually ratified: “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.” Id. at 3040. Despite the differences in wording between the second and third versions, there is some contemporaneous commentary suggesting that they would produce the same “result.” After Clark proposed the third and final version, one Senator commented that the amendment did not “change[] at all the effect of” the provision from Howard’s language, and that “[t]he result is the same.” Id. Clark agreed: “The result is the same.” Id. Still, Michael Abramowicz has suggested that the exchange “may merely indicate that the versions would have the same result for the purposes of Reconstruction, since the generalization of the language would have impact only in future times.” Michael B. Abramowicz, Train Wrecks, Budget Deficits, and the Entitlements Explosion: Exploring the Implications of the Fourteenth Amendment’s Public Debt Clause 10 (Geo. Wash. L. Sch., Pub. L. Rsch. Paper No. 575, 2011) (emphasis added). An additional wrinkle is that the second version uses the word “remain”—“[t]he obligations . . . shall remain inviolate”—suggesting, perhaps, that they already were inviolate. Cong. Globe, 39th Cong., 1st Sess. 2938 (1866) (emphasis added). This is perhaps a reference to the fact that many such debts were already secured by statute. ↑
-
. Perry v. United States, 294 U.S. 330 (1935). ↑
-
. Id. at 354. ↑
-
. Id. ↑
-
. See Abramowicz, supra note 142, at 51 (“The language and history of the Clause indicate that it was not merely a prohibition on the repudiation of Civil War bonds.”). ↑
-
. Act of July 16, 1798, ch. 79, § 2, 1 Stat. 607, 608. ↑
-
. Act of July 17, 1861, ch. 5, § 9, 12 Stat. 259, 261 (“That the faith of the United States is hereby solemnly pledged for the payment of the interest and redemption of the principal of the loan authorized by this act.”). ↑
-
. Intriguingly, such statutory promises returned in 1982. See 31 U.S.C. § 3123(a) (“The faith of the United States Government is pledged to pay, in legal tender, principal and interest on the obligations of the Government issued under this chapter.”). While that does not affect the interpretive point about legislative actions around the ratification of the Fourteenth Amendment, why this statutory promise returned is a mystery that appears unanswered in the legislative history, though it followed major conflicts over the debt limit in the early 1980s. ↑
-
. Act of July 17, 1861, ch. 5, § 1, 12 Stat. 259, 259. ↑
-
. See, e.g., Act of Jan. 20, 1871, ch. 23, 16 Stat. 399; Act of July 14, 1870, ch. 256, 16 Stat. 272; Act of July 25, 1868, ch. 237, 15 Stat. 183; Act of Mar. 2, 1867, ch. 194, 14 Stat. 558. ↑
-
. See, e.g., Stuart McCommas, Note, Forgotten but Not Lost: The Original Public Meaning of Section 4 of the Fourteenth Amendment, 99 Va. L. Rev. 1291, 1298 (2013) (arguing that the phrase “question the validity” had “a unique legal meaning” at the time of the Fourteenth Amendment’s ratification: to stop lawsuits concerning an instrument); Abramowicz, supra note 142, at 24 (“The verb ‘to question’ would be an odd synonym for ‘to repudiate.’”). ↑
-
. Richard Sylla & John Joseph Wallis, The Anatomy of Sovereign Debt Crises: Lessons from the American State Defaults of the 1840s, 10 Japan & World Econ. 267, 290 (1998). ↑
-
. See Gerhard Peters & John T. Woolley, Republican Party Platform of 1868, Am. Presidency Project (May 20, 1868) (emphasis added), https://www.presidency.ucsb.edu/documents
/republican-party-platform-1868 [https://perma.cc/A7H2-QXDB]. ↑ -
. An additional wrinkle, which this paper does not address at any length, is what counts as “debt.” At some level of generality, any legal obligation of the government (such as entitlement spending) can be characterized as a “debt.” While not conceptually satisfying, legal authorities concerning the federal debt and debt limit tend to take a formalistic view of what counts: particular principal- and interest-bearing securities issued by the government. See 31 U.S.C. § 3101. But see Neil H. Buchanan & Michael C. Dorf, Borrowing by Any Other Name: Why Presidential “Spending Cuts” Would Still Exceed the Debt Ceiling, 114 Colum. L. Rev. Sidebar 26, 38–39 (2014) (arguing that “[t]he debt ceiling statute defines ‘debt’ to include more than the government securities that are periodically issued by the Treasury Department,” including unsatisfied spending obligations). One difficulty with this argument is that Chapter 31 uses the term “obligation” many times, each time referring to a more technical instrument. See, e.g., 31 U.S.C. § 3110 (entitled “[s]ale of obligations of governments of foreign countries”); id. § 3111 (entitled “[n]ew issue used to buy, redeem, or refund outstanding obligations”). ↑
-
. See, e.g., The Least Unconstitutional Option, supra note 12, at 1190 (describing the “the ‘nuclear option’”—which reads the debt limit statute itself as a violation of the Fourteenth Amendment—and noting that “the reading of Section 4 that underlies it is debatable”). ↑
-
. Act of July 25, 1868, ch. 237, 15 Stat. 183, 183. ↑
-
. Perry v. United States, 294 U.S. 330, 354 (1935) (emphasis omitted). ↑
-
. Brief for the United States at 62, Perry, 294 U.S. 330 (No. 532). ↑
-
. See also Great Lakes Higher Educ. Corp. v. Cavazos, 911 F.2d 10, 17 (7th Cir. 1990) (noting that the Perry Court “used strong language to confirm the binding obligations of the government when it incurs debts”). ↑
-
. One nice student note takes this position as well: “[T]he President must direct the Treasury to pay debt owed to bondholders first to avoid a constitutional violation.” Kelleigh Irwin Fagan, Note, The Best Choice Out of Poor Options: What the Government Should Do (or Not Do) If Congress Fails to Raise the Debt Ceiling, 46 Ind. L. Rev. 205, 207 (2013). ↑
-
. There may be other possible constitutionally required payments, which I do not address here. For example, it might be that there is a constitutional obligation to prioritize judicial salaries. See U.S. Const. art. III, § 1 (“The Judges, both of the supreme and inferior Courts, shall hold their Offices during good Behaviour, and shall, at stated Times, receive for their Services, a Compensation, which shall not be diminished during their Continuance in Office.”). ↑
-
. See The Least Unconstitutional Option, supra note 12, at 1197 (noting that the President “thus faces a ‘trilemma’: a choice between three bad options, all of which are unconstitutional”); see also Chad DeVeaux, The Fourth Zone of Presidential Power: Analyzing the Debt-Ceiling Standoffs Through the Prism of Youngstown Steel, 47 Conn. L. Rev. 395, 419 (2014) (making similar points). ↑
-
. U.S. Const. art. I, § 8, cl. 1–2. ↑
-
. See 2 U.S.C. §§ 681–688 (the Impoundment Control Act); 31 U.S.C. § 1341 (relevant provision of Anti-Deficiency Act). ↑
-
. See, e.g., U.S. Dep’t of the Navy v. FLRA, 665 F.3d 1339, 1347 (D.C. Cir. 2012) (“The Appropriations Clause is thus a bulwark of the Constitution’s separation of powers among the three branches of the National Government. It is particularly important as a restraint on Executive Branch officers . . . . The Appropriations Clause prevents Executive Branch officers from even inadvertently obligating the Government to pay money without statutory authority.” (citations omitted)); see also Stith, supra note 34, at 1386 (“The Executive violates the Principle of Appropriations Control and the Principle of the Public Fisc if it spends funds not appropriated by Congress.”). ↑
-
. See U.S. Gov’t Accountability Off., GAO-16-464SP, Principles of Federal Appropriations Law 2-36 (2016). ↑
-
. Id. ↑
-
. Morton v. Ruiz, 415 U.S. 199, 230 (1974); see also City of Los Angeles v. Adams, 556 F.2d 40, 49–50 (D.C. Cir. 1977); McCarey v. McNamara, 390 F.2d 601 (3d Cir. 1968). ↑
-
. See Comptroller Gen. of the U.S., B-138524, Question Concerning Secretary of the Treasury’s Authority (1985), https://www.gao.gov/products/b-138524 https://www.gao.gov
/products/b-138524-0 [https://perma.cc/VRV5-YFQS] (“It is our conclusion that the Secretary of the Treasury does have the authority to choose the order in which to pay obligations of the United States. . . . We are aware of no statute or any other basis for concluding that Treasury is required to pay outstanding obligations in the order in which they are presented for payment unless it chooses to do so.”). ↑ -
. 2 U.S.C. § 682(1). ↑
-
. U.S. Gov’t Accountability Off., GAO-04-261SP, 1 Principles of Federal Appropriations Law 1-34 (2004); see also Comptroller Gen. of the U.S., B-203057, Response to Request Concerning Deferral of Budget Authority (1981), https://www.gao.gov/products
/b-203057 [https://perma.cc/47AX-M2WX] (noting that deferrals generally occur “because the Executive decides to stop or slow a program that otherwise would proceed unhindered and in accordance with statutory prescriptions” (emphasis added)). ↑ -
. 2 U.S.C. § 682(1)(B). ↑
-
. Id. ↑
-
. If the ICA simply applied in every case in which budget authority went unused, the phrase “action or inaction” would have been unnecessary. ↑
-
. Comptroller Gen. of the U.S., B-290659, Obligation of Funds Appropriated for “International Organizations and Programs” (2002) (emphasis added), https://www.gao.gov
/products/b-290659 [https://perma.cc/MWJ8-5QWS]. ↑ -
. Comptroller Gen. of the U.S., B-200685 (1980), https://www.gao.gov/products/b-200685 [https://perma.cc/5TWQ-JQVC]; see also Comptroller Gen. of the U.S., B-221412, Question Concerning VA Delays in Obligating Funds (1986), https://www.gao.gov/assets/b
-221412.pdf [https://perma.cc/2ACZ-R8TF] (noting that an agency’s explanations to GAO “show no intention to refrain from using the funds, or to obligate less than was appropriated by the Congress”); Comptroller Gen. of the U.S., B-207374 (1982), https://www.gao.gov/products/b-207374-0 [https://perma.cc/CN2W-HSUM] (“[U]ncertainty as to the amount that ultimately will be available for a particular program may constitute a programmatic basis for a delay in obligating the funds particularly when the uncertainty arises in the context of continuing resolution-funding, where Congress has not yet spoken definitively.”). ↑ -
. Cf. Metzger, supra note 35 (describing various ways in which the importance of appropriations law may be underappreciated). ↑
-
. U.S. Const. art. I, § 8, cl. 1. ↑
-
. Id. § 9, cl. 7. ↑
-
. Id. ↑
-
. This basic insight has relevance for other areas in which legislative commands exceed available funds. See, e.g., Matthew B. Lawrence, Medicare “Bankruptcy”, 63 B.C. L. Rev. 1657, 1687 (2022) (discussing the possibility of spending triage in the context of a potential bankruptcy in the Medicare system). ↑
-
. Dames & Moore v. Regan, 453 U.S. 654, 678, 686 (1981). ↑
-
. See Trevor W. Morrison, Constitutional Avoidance in the Executive Branch, 106 Colum. L. Rev. 1189, 1218–19 (2006) (summarizing practice, albeit somewhat skeptically). ↑
-
. But see Daniel J. Hemel, The President’s Power to Tax, 102 Cornell L. Rev. 633 (2017) (making the argument that, at least in certain cases, the President has underutilized taxing powers). ↑
-
. See The Least Unconstitutional Option, supra note 12, at 1203; see also, e.g., Robert Hockett, This Is What Would Happen if Biden Ignores the Debt Ceiling and Calls McCarthy’s Bluff, N.Y. Times (May 9, 2023), https://www.nytimes.com/2023/05/09/opinion/biden-mccarthy-debt-ceiling.html [https://perma.cc/HE5L-RF5K] (citing the later-in-time rule, “which basically means that Congress’s most recent budget legislation trumps any earlier legislated ceiling,” as a “powerful” legal reason for rejecting the debt limit). ↑
-
. See, e.g., Consolidated Appropriations Act, 2022, Pub. L. No. 117-103, 136 Stat. 49, 52. ↑
-
. 31 U.S.C. § 3101(b). ↑
-
. The Least Unconstitutional Option, supra note 12, at 1203 (“[T]he two most useful doctrines [of timing and specificity] both point in the same direction, suggesting that the debt ceiling should give way when it is in conflict with the taxing and spending provisions of the government’s budget.”). ↑
-
. See, e.g., S.J. Res. 33, 117th Cong. (2021). The Office of Management and Budget typically publishes and updates a list of statutory limits on federal debt. See Table 7.3—Statutory Limits on Federal Debt: 1940–Current, White House, https://www.whitehouse.gov/omb/budget/historical
-tables/ [https://perma.cc/53T2-EDVW]. ↑ -
. Compare Extending Government Funding and Delivering Emergency Assistance Act, Pub. L. No. 117-43, 135 Stat. 344 (2021) (making appropriations for the fiscal year ending September 2022), with S.J. Res. 33, 117th Cong. (2021) (increasing the debt limit). ↑
-
. See U.S. Gov’t Accountability Off., supra note 167, at 2-52. ↑
-
. See Application of Anti-Nepotism Statute to a Presidential Appointment in the White House Office, 41 Op. O.L.C., slip op. at 11 (2017) (observing that “the canon about general and specific statutes seems of limited help” when “neither of the two relevant statutes can readily be characterized as more or less specific than the other”). ↑
-
. Act of Mar. 3, 1791, ch. 25, 1 Stat. 218, 218–19. ↑
-
. Act of May 7, 1800, ch. 42, § 1, 2 Stat. 60, 60 (authorizing borrowing in part “for making up any deficiency in any appropriation heretofore made by law”); Act of July 16, 1798, ch. 79, § 2, 1 Stat. 607, 608 (same). ↑
-
. See Gov’t Accountability Off., GAO-05-734SP, A Glossary of Terms Used in the Federal Budget Process 93 (2005). ↑
-
. The nineteenth century practice of coercive deficiencies is not to the contrary. See, e.g., Matthew B. Lawrence, Disappropriation, 120 Colum. L. Rev. 1, 19 (2020) (discussing history of coercive deficiencies and noting that Congress “asserted control over the executive’s commitment power” with the Anti-Deficiency Act). ↑
-
. The Least Unconstitutional Option, supra note 12, at 1208 (“The issuance of government debt is significantly less complicated than the determination of government spending levels, because debt is a relatively undifferentiated (and completely monetizable) asset.”). ↑
-
. U.S. Const. amend. XIV, § 4. ↑
-
. See Obama Urges Congress to Reopen Government, Avoid Default, Reuters (Oct. 7, 2013, 2:11 PM), https://www.reuters.com/article/idUSBRE9960LC [https://perma.cc/7V9A-5BSN%5D. ↑
-
. Lew, supra note 9, at 1. ↑
-
. See, e.g., Emily Cochrane, House Approves Bill to Avert U.S. Default, Sending It to Biden, N.Y. Times (Oct. 12, 2021), https://www.nytimes.com/2021/10/12/us/politics/debt-ceiling-house
-biden.html [https://perma.cc/7SJY-D3N2]; Sarah Binder, Congress Is Struggling to Raise the Nation’s Debt Cap. Here’s What You Need to Know, Wash. Post (Oct. 5, 2021, 7:45 AM), https://www
.washingtonpost.com/politics/2021/10/05/congress-is-struggling-raise-nations-debt-cap-heres-what
-you-need-know/ [https://perma.cc/2UJ7-Y9ZH]. ↑ -
. See Kitch & Mahoney, supra note 55, at 1293–98. ↑
-
. See Cong. Budget Off., Federal Net Interest Costs: A Primer 6 (2020), https://www
.cbo.gov/publication/56910 [https://perma.cc/7HST-TR2P]. ↑ -
. Id. ↑
-
. These figures are available in the Treasury’s Monthly Statement. See Bureau of the Fiscal Serv., U.S. Dept. of the Treasury, Monthly Treasury Statement: Receipts and Outlays of the United States Government for Fiscal Year 2022 Through February 28, 2021, and Other Periods (2021), https://fiscaldata.treasury.gov [https://perma.cc/H3JV-CWDB]. ↑
-
. Data aggregated from U.S. Treasury public fiscal data. ↑
-
. See Letter from Janet L. Yellen, supra note 6; Letter from Janet L. Yellen, Sec’y, U.S. Dep’t of the Treasury, to the Hon. Nancy Pelosi, Speaker, U.S. House of Reps. (Sept. 8, 2021), https://home.treasury.gov/system/files/136/Debt-Limit-Letter-to-Congress_20210908_FINAL-Pelosi
.pdf [https://perma.cc/3GBJ-G4T2]. ↑ -
. This raises a question: Why do so many people believe, or represent, otherwise? Perhaps the most obvious explanation is to take those representations at face value: Even if the argument of this paper is correct—default is both unlawful and avoidable—that is (alas) not an argument that all public actors need to understand and accept. But the public representations concerning the limit may also be a testament to what the limit has become: less a statute that anyone defends on the merits, and more a vessel for legislative bargaining and strategy. On that account, default talk is strategic: it emphasizes that the stakes of legislative inaction are steep (particularly for an administration eager to pass its budget priorities). ↑
-
. Act of Mar. 26, 1790, ch. 4, § 7, 1 Stat. 104, 105–06. ↑
-
. Second Liberty Bond Act, ch. 56, § 1, 40 Stat. 288, 288 (1917). ↑
-
. See Krishnakumar, supra note 15. ↑
-
. 31 U.S.C. § 3102(a). ↑
-
. According to press reports, the desire to increase the debt limit by a specific amount—rather than suspend it until a future date—was based on Senator McConnell’s perception that holding a vote on a specific dollar amount would provide a more politically meaningful datapoint—a “vote along party lines to raise our nation’s debt limit by trillions of dollars”—than a temporary suspension. Makini Brice & Susan Cornwell, U.S. Congress Approves Boosting Debt Limit to $31.4 Trillion, Reuters (Dec. 14, 2021, 11:30 PM), https://www.reuters.com/markets/rates-bonds/us-congress-vote-debt-limit-hike
-averting-default-risk-2021-12-14/ [https://perma.cc/W6GV-MKGE%5D. ↑ -
. See 31 U.S.C. § 3101A(a)(1), (b). ↑
-
. See H.R. Doc. No. 112-48 (2011); 158 Cong. Rec. 17 (2012); see also Austin, supra note 5, at 24–25. ↑
-
. They can be related. In 2013, for example, a legislative disagreement involving the debt limit resulted in a lapse that led to a sixteen-day shutdown of the federal government. See U.S. Gov’t Accountability Off., GAO-15-476, Debt Limit: Market Response to Recent Impasses Underscores Need to Consider Alternative Approaches 10 (2015). ↑
-
. One is simply the amount of money involved. While this varies from lapse to lapse, the scope of unappropriated government functions has in many recent shutdowns been a smaller percentage of total government spending than the percentage of government spending that is paid for by debt. ↑
-
. 31 U.S.C. § 1342. ↑
-
. Government Employee Fair Treatment Act of 2019, Pub. L. No. 116-1, 133 Stat. 3. ↑
-
. See, e.g., Authority for the Continuance of Government Functions During a Temporary Lapse in Appropriations, 5 Op. O.L.C., slip op. at 2 (1981); see also Government Operations in the Event of a Lapse in Appropriations, Op. O.L.C., slip op. at 1 (1995). ↑
-
. See The Economic Costs of Debt-Ceiling Brinkmanship Before the Joint Econ. Comm., 113th Cong. 113-127, at 2 (2013) (statement of Sen. Amy Klobuchar, Vice Chair, Joint Econ. Comm.) (“The delay in lifting the debt ceiling and the resulting uncertainty meant that the Treasury had to pay higher yields than otherwise would have been necessary, costing taxpayers $1.3 billion.”). ↑
-
. For example, the rule of prioritization could be that mandatory spending must precede discretionary spending, but the Executive Branch could construct a reasonable rule of prioritization within the category of discretionary spending. ↑
-
. U.S. Gov’t Accountability Off., supra note 167, at 2-36. ↑
-
. See Comptroller Gen. of the U.S., supra note 170. Such legislative prioritization rules also have something of an established pedigree. In early revenue laws, for example, Congress would occasionally prescribe the order in which funds would be applied. See, e.g., Act of May 7, 1800, ch. 42, § 3, 2 Stat. 60, 60 (authorizing borrowing to be spent in the following manner: “[f]irst, to make up any deficiency in any appropriation heretofore made by law”). ↑
-
. Letter from William French Smith, Att’y Gen., Off. of the Att’y Gen., to the Hon. Howard Baker, Sen., U.S. Senate (Nov. 11, 1983), https://www.google.com/books/edition/Effects_of_Potential
_Government_Shutdown/ujbFStj4vckC?hl=en&gbpv=1&dq=%22no+comparable+instance+in+which+the+President+has+been+required+to+determine+his+responsibilities%22&pg=PA95&printsec=frontcover [https://perma.cc/QS5A-3HKP]. ↑