Bargaining in the Shadow of the Debt Ceiling: When Negotiating over Spending and Tax Laws, Congress and the President Should Consider the Debt Ceiling a Dead Letter

Neil H. Buchanan* and Michael C. Dorf**

If the debt ceiling is inconsistent with existing spending and taxing laws, what must the President do? In earlier work, we argued that when Congress creates a “trilemma”—making it impossible for the President to spend as much as Congress has ordered, to tax only as much as Congress has ordered, and to borrow no more than Congress has permitted—the Constitution requires the President to choose the least unconstitutional path. In particular, he must honor Congress’s decisions and priorities regarding spending and taxing, and he must issue enough debt to do so. Here, we extend the analysis in two ways. First, we rebut several recently-advanced arguments that purport to dissolve the trilemma. We explain that upon close inspection some of those arguments amount to non sequiturs, while one potentially promising solution would merely substitute a nondelegation violation for a separation-of-powers violation. Second, we ask whether our original analysis changes if both Congress and the President, when they pass appropriations measures, knowingly create a trilemma. We conclude that the answer does not change—that is, that spending and taxing laws must still take precedence over the debt ceiling. Accordingly, the debt ceiling is effectively a dead letter, and both Congress and the President should treat it as such.


I. Introduction


The debt ceiling statute1 was enacted nearly a century ago, but it did not play a significant role in U.S. fiscal policy until the spring and summer of 2011, when Republicans in the House of Representatives used the threat that the government would be forced to default on its existing obligations to pressure President Obama to reduce future federal spending. In its first go-round, that strategy was quite successful, as the President and the Democratic-controlled Senate agreed upon initial spending cuts and a mechanism for further cuts totaling more than two trillion dollars over a ten year period.2 The Budget Control Act of 2011 (BCA11), signed in August of that year after an acrimonious political standoff, paired those enactments with an increase in the debt ceiling sufficient to accommodate anticipated borrowing roughly through the end of 2012.3

In the aftermath of that political cliffhanger, we wrote How to Choose the Least Unconstitutional Option: Lessons for the President (and Others) from the Debt Ceiling Standoff [hereinafter How to Choose].4 We argued there that the debt ceiling statute was unconstitutional, on two grounds. First, we substantially agreed with the claim that the debt ceiling violates Section 4 of the Fourteenth Amendment. Second, even if there were no Fourteenth Amendment violation, we pointed out that, in combination with other laws, the debt ceiling could present the President with a “trilemma.” A failure of Congress to authorize borrowing sufficient to cover the gap between spending and taxes would force the President to choose how to violate the Constitution: by failing to spend the money that Congress has appropriated; by exceeding the tax collections that Congress has ordered; or by issuing debt beyond the level permitted by the debt ceiling that Congress has specified.


Faced with nothing but bad options, we argued that the President would be obliged to make the least unconstitutional choice. Under the criteria that, we argued, should guide such a choice, we concluded that issuing debt in excess of the debt ceiling would be least unconstitutional, and thus obligatory. Therefore, the premise underlying the debt ceiling-inspired political standoff—that congressional refusal to increase the debt ceiling would require the President to unilaterally cut spending (but not unilaterally increase taxes) to stay below the ceiling—is incorrect.


Shortly after How to Choose was published, in late 2012, and as the level of federal debt was (as expected) reaching the level to which the debt ceiling was set by BCA11, House Republicans announced that they would again insist on additional spending cuts (but no tax increases) as the “price” of increasing the debt ceiling as necessary to accommodate the borrowing that Congress’s own taxing and spending decisions will make necessary.


At that time, we published a follow-up essay, Nullifying the Debt Ceiling Threat Once and for All: Why the President Should Embrace the Least Unconstitutional Option [hereinafter Nullifying the Debt Ceiling Threat],5 clarifying and extending our points from How to Choose. We observed that “none of the participants in the negotiations has offered any public indication that they even understand the nature of the problem that the President would face, much less how to resolve that problem, should Congress refuse to raise the debt ceiling.”6 We thus argued that “the President should make it clear, as soon as possible, that the debt ceiling is not, and cannot legally be used as, a cudgel with which Congress can force him to renegotiate the federal budget,” and thus that “the President’s best course is to make clear that the debt ceiling must always give way to the wishes of Congress, as expressed through the budget of the United States.”7


On December 31, 2012, the Secretary of the Treasury notified Congress that the federal debt had reached the maximum level permitted by the debt ceiling, which required him to begin to undertake “extraordinary measures” (as he had done in the summer of 2011) to prevent default.8 Those measures would, however, extend the day of reckoning only until sometime between mid-February and early March of 2013. If we had reached that point, the debt ceiling would have become genuinely binding, and the President would have faced a trilemma.


Before the extraordinary measures were exhausted, however, Congress passed (and the President signed) the No Budget, No Pay Act of 2013, which temporarily suspends enforcement of the debt ceiling, through May 18, 2013, with the ceiling to increase automatically by the amount necessary to accommodate whatever borrowing will occur in the meantime.9 During the time that the debt ceiling is in suspension, the federal government’s current “continuing resolution” will expire. That resolution, which was passed by Congress in fall 2012 (in lieu of the standard full-fiscal-year appropriations bills), funds the government through March 27, 2013.10 After that date, unless a new set of appropriations bills is approved, there will be no authority for the President to spend, and the government will shut down, except for required spending on “emergency measures” that have been put in place in light of earlier shutdowns.


Because the threat of a government shutdown is not a threat to refuse to raise the debt ceiling, our analyses in How to Choose and Nullifying the Debt Ceiling Threat appear not to be directly applicable to the current situation. However, everyone knows that the debt ceiling looms in the background. Any budget deal in March that does not address what will happen in May, when the temporary suspension of the debt ceiling expires, is vulnerable to further revision in light of the possibility that House Republicans might again attempt to use the debt ceiling as leverage.11 Are our previous arguments relevant given the shadow that the debt ceiling casts over current negotiations?


Maybe not. Prior to 2013, it was at least plausible that the President and Congress could have found themselves in a trilemma essentially “by accident”—or, at least, that the President could have imagined that the debt ceiling would continue to be increased in ways that would avoid trilemmas.12 To date, therefore, our work could be read as addressing the following question: If the already existing taxing, spending, and borrowing laws create a trilemma, what must the President do? Now, however, the President is faced with a different question: If he agrees to a series of taxing and spending laws that everyone knows would, if executed faithfully, exceed the debt ceiling, would the resulting trilemma be different from one that no one realistically anticipated?


The short answer to that question is that an anticipated trilemma would not be constitutionally different from an unanticipated one, because our argument does not depend on the manner in which a trilemma comes into existence. As we explained in How to Choose, when faced with no plausible constitutional options, a President should choose the course that minimizes the usurpation of power (or other constitutional violation), that maximally preserves the ability of other actors to reverse his choice, and that minimizes sub-constitutional harm.13


Moreover, as we explain below, drawing any other conclusion would perversely allow Congress to abrogate its legislative duties, by passing spending and taxing laws that are politically popular but that would exceed the debt ceiling, thus putting the President in the position of making unpopular spending cuts—a duty that the Constitution rightly assigns to the legislative branch. Thus, the debt ceiling’s constitutional defect is in some respects more serious during budgetary negotiations than after the fact, because that is when Congress must commit itself to spending and taxing laws that the President can actually execute.


Before explaining more fully how the current posture amplifies our conclusion that the President should treat a threatened failure to raise the debt ceiling as hollow, we address preliminary objections. Although no one has offered a scholarly rejoinder to our work, some legal scholars (and others) have offered piecemeal responses to various elements of our analysis. Those responses raise legitimate questions about our approach. We address such questions in the next Part.


II. No Escaping the Trilemma


In the months leading up to the latest stopgap legislation, commentators offered various arguments against the view that the President should unilaterally violate the debt ceiling in the event that Congress authorizes spending in excess of the revenue it authorizes to be raised through a combination of taxation and borrowing under the debt ceiling. In this Part, we consider and reject three such arguments.14


A. The Fourteenth Amendment Non Sequitur

Both the Obama Administration and some commentators have taken the position that the President may not borrow money in excess of the debt ceiling under any circumstances. Yet one of the arguments that has been most prominently advanced in favor of this position is ultimately a non sequitur. According to this argument, it does not matter whether failure to pay the government’s bills would violate Section 4 of the Fourteenth Amendment because, as Professor Tribe wrote in January 2013, “nothing in the 14th Amendment suggests that the President may usurp the power ‘to borrow money on the credit of the United States,’ which Article I, Section 8 of the Constitution vests in Congress, in order to prevent a default.”15


We agree that the Fourteenth Amendment does not empower the President to borrow money without congressional authorization. Indeed, neither does any other constitutional provision. We nonetheless regard this line of argument as beside the point for two reasons.


First, the claim that the debt ceiling violates Section 4 of the Fourteenth Amendment does not depend on the Fourteenth Amendment itself or any other part of the Constitution as a source of authority for the President to borrow money beyond the debt ceiling. Rather, the claim rests on basic principles of severability.


In a provision of law that we shall call the “borrowing authorization,” Congress delegated to the President and the Secretary of the Treasury the power to “borrow on the credit of the United States Government amounts necessary for expenditures authorized by law. . . .”16 A series of other laws defines the debt ceiling, which places a cap on how much debt the federal government can incur.17 If particular circumstances arise under which the debt ceiling is unconstitutional, then the debt ceiling would be a nullity, at least as applied. But if the debt ceiling is a nullity, the President still has authority to borrow because the borrowing authorization remains valid. Congress was—and in the imagined crisis would remain—the source of that authorization.


Another way to put this last point would be to say that the borrowing authorization is severable from the debt ceiling. This conclusion seems unassailable because the borrowing authorization is codified in a separate Code section from the debt ceiling and has long been treated as separate and distinct from the debt ceiling. The debt ceiling was amended in the summer of 2011 and again in early 2012.18 By contrast, the borrowing authorization was last amended in any substantial way three decades ago.19 Although there may be some doubt about whether the courts will presume that sub-parts of a single statute are severable,20 there is generally no good reason to treat two already-separate statutory provisions as nonseverable, especially when Congress has repeatedly treated them as separate and when the valid portion can function effectively without the invalid portion.21


Might there be some reason to regard the debt ceiling as a special case? Perhaps the debt ceiling is the rare statutory provision that should be treated as nonseverable from another provision, namely, the borrowing authorization. Perhaps when Congress delegates power to the executive branch in one statute and then places limits on the exercise of that power in another statute, a finding of the invalidity of those limits has the consequence of invalidating the original delegation of power itself.


But what could be the basis for such a novel approach to (non)severability? Suppose that some statutory provision specified that federal bonds could only be sold to white people or to men. The conclusion that the provision was unenforceable as a denial of equal protection surely would not entail the invalidity of the separate borrowing authorization, even if that borrowing authorization was originally enacted as part of the same law that established the racist or sexist limit. So why should the fact that the debt ceiling violates Section Four of the Fourteenth Amendment—rather than the Equal Protection Clause of Section One of the Fourteenth Amendment (as “reverse incorporated” by the Fifth Amendment’s Due Process Clause)—lead to a different answer here? Congress undoubtedly has the power to specify that in the event that some limitation on an otherwise permissible enactment is found invalid, the enactment shall be treated as nonseverable from the invalid limitation.22 But Congress has enacted no legislation requiring that the debt ceiling statute be treated as nonseverable from the borrowing authorization. Accordingly, ordinary severability principles lead to the conclusion that if the debt ceiling is best read as violating Section Four of the Fourteenth Amendment, the severable borrowing authorization provides the President with ample authority to issue bonds beyond the debt ceiling.


The claim we have been examining—that the Fourteenth Amendment does not empower the President to borrow in excess of the debt ceiling—was evidently intended by those making it to operate conditionally: Even if the debt ceiling violates the Fourteenth Amendment, they appear to be saying, that does not entitle the President to borrow money beyond the debt ceiling. As we have just seen, this reasoning is flawed, because the President’s borrowing authority comes from the already-separate and thus surely severable statutory borrowing authorization, not from any assertion that the Fourteenth Amendment itself empowers the President to act unilaterally.


However, perhaps we can make more sense of the claim if we imagine that those making it do not really mean to assert that the Fourteenth Amendment does not independently empower the President to borrow money. Rather, their objection might simply be that the debt ceiling does not violate the Fourteenth Amendment in the first place. Regarding that argument, we are content to rest on the analysis that we laid out in our first Article.23 Here, we simply point out that those who argue against the Fourteenth Amendment as the basis for finding the debt ceiling unconstitutional stop there, ignoring entirely our logically separate argument based on the separation of powers (which was the focus of the vast majority of our analysis in both of our prior papers). This converts their objection into a different non sequitur: the claim that, because they disagree with us about the Fourteenth Amendment, our argument based on the trilemma must be wrong, too.


But if we are right that all of the President’s realistic options—unilaterally borrowing, unilaterally raising taxes or unilaterally cutting spending—would violate the Constitution, then the fact that the President cannot borrow without usurping congressional power is only the starting point. A President who faces only these unconstitutional options will usurp a power of Congress no matter what he does. The question is which usurpation is least unconstitutional. One does not establish that unilateral presidential borrowing is more unconstitutional than the other options by showing only that unilateral presidential borrowing is unconstitutional. One would also need to establish that one of the other options is less unconstitutional. No serious commentator has even tried to do so, much less succeeded.


B. Is Cutting Spending an Omission?

If one thought that unilateral presidential spending cuts were constitutionally valid, then the conclusion that presidential borrowing in excess of the debt ceiling is unconstitutional really would be sufficient to rule out the latter. And there is at least some intuitive appeal to the idea that when the President fails to spend money that Congress has appropriated, he violates the statutes requiring him to spend—both the specific appropriations measures and the Impoundment Control Act24—but not the Constitution. In this view, the congressional power set out in Article I, Section 8 is the power to decide to spend, and so the President only usurps that power by spending without authorization. A failure to spend may be illegal but not unconstitutional.


In Nullifying the Debt Ceiling Threat, we explained why this argument is flawed.25 Core principles of separation of powers and case law establish that a presidential decision not to spend appropriated funds would violate the Constitution, not just the federal statutes requiring that those funds be spent. To summarize our prior argument, the case law casts doubt on the ability of the President to cut spending even with congressional authorization. He surely has no such power without congressional authorization.26


To our knowledge, no one seriously disputes the general proposition that unilateral presidential spending cuts are unconstitutional, but we have encountered resistance to its application to a potential debt ceiling standoff. For example, in January 2013, one of us was a guest on an hour-long nationally broadcast radio show. For part of that time, Professor McConnell was also a guest on the show. He appeared to concede that unilateral presidential spending cuts are indeed unconstitutional in most circumstances but contended that there is a constitutionally relevant difference between a President simply declining to spend money that Congress has appropriated on the ground that he does not support the policy goals set by Congress—as President Nixon did—and a President not spending money because the federal Treasury has no money.27


We think that the position espoused by Professor McConnell is not entirely implausible, but that the case law is nonetheless best read to reject his view.28 The precedents do not speak to the precise question because Congress has never been so reckless as to order the President to spend more money than it authorizes him to collect through taxes, borrowing and other means. Thus, in a technical sense, it might be possible for a court to read the prior cases as applying only in ordinary times.


However, one must have some reason for so limiting the prior precedents, and here no such reason is apparent. The conclusion that the President may not unilaterally cut spending in ordinary times rests on the constitutional text and the structural principle of separation of powers: Article I, Section 8 assigns the power to choose how much money to spend, and for what programs and purposes, to Congress; thus, a presidential decision to spend less than Congress has ordered to be spent usurps the power of another branch. Nothing in this chain of reasoning depends on the Treasury being flush or empty, or on whether the President acts with evil intent or reluctantly.


To put the point in practical terms, the power of Congress to spend money is not simply the power to name a dollar figure that will leave the Treasury. It is fundamentally a power to set priorities, to allocate among different programs or at least to delegate some authority to make allocation decisions to the executive branch. At bottom, what makes a presidential decision to make unilateral spending cuts unconstitutional in extraordinary no less than in ordinary times is that it usurps Congress’s authority to say what programs get how much money.


Furthermore, a decision by the President to make unilateral spending cuts in the face of a debt ceiling standoff presents the same separation-of-powers problem as a Nixonian decision to impound funds for programs he opposes, because there is no neutral or natural spending-cut formula that a well-intentioned President can employ. Some essential government services and programs will almost certainly be prioritized, but once the President abandons the priorities set by Congress, the selection of any criteria will necessarily involve the exercise of quintessentially legislative judgment.


Nor can this problem be avoided by so-called “across-the-board” spending cuts, even if such an approach were not a policy disaster, because considerable judgment will be needed to define baselines. Do “across-the-board” cuts apply in the same manner to bills by government contractors for work already performed as for work under contract but not yet performed? Do past current-fiscal-year payments to Social Security recipients count as part of the denominator in figuring out what fraction of money owed to them should be paid in a month when the debt ceiling blocks government borrowing sufficient to pay all sums due? Even at the level of a particular agency facing, say, a budget cut of thirty percent, there will be innumerable questions about allocation of those cuts among salary, services, maintenance, and so on.


To be sure, we may feel greater sympathy for a President Obama making spending cuts because the federal till is empty than for a President Nixon making spending cuts because he disagrees with Congress about some policy goal, but from the constitutional perspective, their motives are irrelevant. The pure-of-heart Obama reluctantly arrogating legislative power to the executive violates the separation of powers no less than the tricky Nixon. Accordingly, Professor McConnell’s proposed distinction does not dissolve the trilemma.


C. Substituting a Nondelegation Problem for Presidential Usurpation

Our analysis to this point, both here and in our prior work, has treated the relevant statutes as posing mutually contradictory commands for the President: The crisis arises when Congress tells the President that he must spend more money than it authorizes him to collect through taxes, borrowing and other means. But suppose that this view is wrong. Is it possible that Congress has already established a priority rule?


In How to Choose, we considered but ultimately rejected the possibility that appropriations measures enacted after the most recent adjustment to the debt ceiling impliedly repealed the debt ceiling,29 much as we would have favored that reading on policy grounds. Perhaps, however, Congress has established the opposite priority rule. After we wrote Nullifying the Debt Ceiling Threat, Professor Tillman30 brought to our attention the following language, which typically appears in appropriations measures: “The following sums are appropriated, out of any money in the Treasury not otherwise appropriated . . . .”31 Might the inclusion of the qualifier “out of any money in the Treasury” imply that Congress has in fact already prioritized the limits contained in other statutes, including the debt ceiling, over spending?


We think this is a mistaken reading of “money in the Treasury,” but before we explain why, it will be useful to explicate the argument for reading this language as tacitly incorporating the debt ceiling into appropriations laws. Read in this way, the “money in the Treasury” language sets forth a principle that Congress only wants to spend money coming into the Treasury lawfully via authorized taxes, borrowing, fees and other measures. In this view, appropriations laws incorporate any other limitations on how Congress raises money, including the debt ceiling. Thus, there would be no conflict between the debt ceiling and the appropriations measures that include the “money in the Treasury” language. When the debt ceiling is hit, the appropriations stop.


We think the foregoing argument has three key flaws. First, if read as a qualification on what money can be spent for appropriations, the statutory language proves too much. The quoted appropriations language appears to say that appropriations can only be made from money that happens to be in the Treasury when the law is passed (or, perhaps, from money that will already be in the Treasury on the day that the appropriated amounts are to be spent). That reading, however, would make all deficit spending illegal, even when Congress has authorized borrowing and the debt ceiling is not binding. If the money is not already “in the Treasury,” under this reading, the appropriation would be invalid. Proponents of the argument now under consideration could only avoid that absurd result by construing the relevant language as though it permitted appropriations to be paid from any money in the Treasury when the appropriation is made or that may lawfully come into the Treasury in the future. That is, they would have to say that the phrase “money in the Treasury” must be implicitly amended to mean “money in the Treasury, and any money that can be added to the Treasury by borrowing that does not exceed the debt ceiling.”


But why read in that atextual expansion of the statutory language when there is a much more straightforward way to avoid the absurdity? We would deny that the statutory language contains any temporal element at all. The more natural reading of the “money in the Treasury” language is that it simply identifies where the government is supposed to pay money from—the Treasury, rather than any other reservoir of government (or private) funds. Consistent with their silence on the matter, the appropriations laws leave to other statutes the authorizations for and limitations on how money comes into the Treasury.


Second, reading the “money in the Treasury” language to incorporate the debt ceiling would create a bizarre prioritization rule. That is because other statutes that require appropriations do not contain the “money in the Treasury” language. For example, the obligation to pay Social Security and other so-called “entitlements” beneficiaries is stated in absolute terms.32 Hence, if “money in the Treasury” means that Congress may not breach the debt ceiling to pay appropriations made in laws containing that language, then the absence of such language in other statutes suggests that the debt ceiling may be breached to pay those other obligations. Thus, the argument under consideration would not even authorize the President to preserve and use federal resources for the most vital services, as its proponents claim.


Third, reading “money in the Treasury” as presidential authority for budget cutting would merely trade one constitutional infirmity for another. The Congressional Research Service has quite rightly concluded that neither the appropriations laws nor any other laws on the books provide the executive branch with any instructions about what bills to pay, in what proportion, or in what order.33 Accordingly, to say that the federal statutory appropriations measures instruct the President to “cut spending” is to say that they authorize the President to cut spending by employing a formula or procedure entirely of his own devising. Yet such an instruction is so broad as to violate the nondelegation doctrine.


As a formal matter, the nondelegation doctrine forbids Congress from delegating its legislative power to the executive branch34 (or anywhere else). As a practical matter, however, the federal government could not function if Congress had to write all of the particulars of the nation’s laws. Consequently, the case law permits Congress to assign to the executive a substantial gap-filling role, so long as Congress provides the executive with an “intelligible principle” to guide the exercise.35 Students of constitutional law learn that since the New Deal, the Supreme Court has upheld every law challenged on nondelegation grounds, even when the ostensible intelligible principle is quite vague.36 The resulting pattern of cases has led some commentators to declare the nondelegation doctrine a dead letter, even as those same commentators lament its demise.37


But the fears of these commentators are overblown.38 Although the nondelegation doctrine does indeed permit very broad language to serve as constitutionally adequate policy guidance, some limits remain. The Court’s description of those limits in Whitman v. American Trucking Ass’ns,39 is instructive. The Court described the two New Deal-era cases that did find violations of the nondelegation principle as still good law and characterized them in the following terms:


[W]e have found the requisite “intelligible principle” lacking in only two statutes, one of which provided literally no guidance for the exercise of discretion, and the other of which conferred authority to regulate the entire economy on the basis of no more precise a standard than stimulating the economy by assuring “fair competition.”40


Thus, we see two outer bounds. First, even very vague language will ordinarily satisfy the requirement of an “intelligible principle” but “literally no guidance” will not. Second, when Congress authorizes the executive to prescribe rules for the “entire economy,” something more specific than the usually permissible broad standard is required.


Accordingly, if there is anything at all left to the nondelegation doctrine—and the Court’s extended discussion of it in American Trucking Ass’ns pretty clearly indicates that there is—then a tacit instruction to the President to cut federal spending across the entire federal budget with literally no guidance about how to set priorities surely violates the doctrine. Indeed, in its standardlessness and breadth, it would appear to exceed both of the two outer bounds the Court has identified. At a bare minimum, the very substantial nondelegation question that would be raised by construing the “money in the Treasury” language in various appropriations measures as authorizing the President to make unilateral cuts to federal appropriations strongly reinforces the other considerations that lead us to reject the claim that this language incorporates the debt ceiling.41


Our consideration of the nondelegation problem that would be created by reading existing appropriations measures to authorize presidential discretion to rebalance spending priorities also points the way to a solution that would not require the President to choose among unconstitutional options: Congress could clearly stipulate in future appropriations measures that the authorization to spend is contingent on compliance with the debt ceiling and specify a constitutionally adequate “intelligible principle” to guide the President in the event that he must cut spending. So long as cutting spending under this procedure did not renege on any promises in violation of Section 4 of the Fourteenth Amendment, there would be no constitutional infirmity, because there would be no trilemma for the President.


To be sure, the inclusion of a provision prioritizing spending cuts over borrowing or taxing would be deeply controversial politically. Accordingly, we can expect the President and congressional Democrats to resist Republican efforts to provide formal authority for such prioritization. If such a prioritization rule were nonetheless enacted, it would amount to a kind of “Gephardt Rule,”42 because it would specify a contingent set of spending and taxing priorities that would fit within the debt ceiling. If Congress wished, it could specify (but would be neither presumed nor required to do so) that the President should regard the initial, non-prioritized spending and taxing provisions as setting forth part of the intelligible principle to guide the executive in making cuts.43


In other words, if we imagine that Congress is able to pass a set of laws that never results in a trilemma, then we have assumed away the problem. What we actually see today, by contrast, is a Congress with one house dominated by members who may well want to create genuine trilemmas, specifically to force the President to cut spending by amounts greater than they have been able to extract through normal negotiations.


In the next Part, we consider the resulting implications for the negotiating posture of the parties and what should happen if, with their eyes wide open, the parties, including the President, dig in their heels and leave the President with no constitutional options.


III. During Negotiations over Spending and Taxes, the Threat of Not Raising the Debt Ceiling Has No Constitutional Force

In both How to Choose and Nullifying the Debt Ceiling Threat, we argued that—viewed from the perspective of an already-existing trilemma—the debt ceiling is unconstitutional in the sense that the President would be constitutionally required to execute the spending and tax laws, rather than adhere to the debt ceiling. Importantly, we did not draw this conclusion as a matter of policy preference, based on any prior commitments to the level or content of the specific spending and tax provisions that Congress has enacted. Nor did we reach our conclusion based on the likely economic consequences of a President’s unilateral spending cuts.44


For example, even though we do happen to believe, as a matter of macroeconomic policy, that now is a particularly bad time to cut government spending, our analysis here is fully consistent with the idea that Congress—and Congress alone—could cut spending as much as it wishes (subject to the President’s veto power). Congress unquestionably has the power under the Constitution to make (what we would regard as) bad decisions about taxing and spending, but the President cannot change those decisions unilaterally, even if the debt ceiling would provide an excuse to do so.


Our argument, therefore, is not that the debt ceiling is bad policy, but that separation of powers is most threatened by the possibility of having the President arrogate to himself the ability to rebalance the spending decisions that Congress has expressed in its appropriations laws. The separation of powers suffers the least damage when the President honors the spending and tax laws, by issuing debt in excess of the debt ceiling.


In this Part, we extend our earlier arguments to say that the President’s constitutional duties are neither situational nor path dependent. No matter how we end up in a trilemma, the outcome is the same: The President must regard exceeding the debt ceiling as the least unconstitutional course, and he must therefore spend and tax as Congress has otherwise ordered him to do.


A. The Constitutional Criteria and Their Application

In our earlier work, we considered how a President should respond to an already-existing trilemma, proceeding in three steps: We articulated criteria based upon fundamental constitutional principles; we defended those criteria; and we then applied those criteria to the choices that a President would face if Congress were to confront him with a trilemma. In particular, we argued that the separation of powers draws its normative force from the idea that the President’s role in our system of government is appropriately circumscribed, especially in the process of making laws. Although the President holds the veto power and is certainly an important participant in all policy negotiations, the Constitution vests in the legislative branch the power and responsibility to make the often-difficult political choices that are the essence of democratic government.


Even when Congress gives the President mutually inconsistent commands, therefore, the President should not respond by exercising whatever policy discretion that he might be able to rationalize.45 Instead, he must minimize the damage to the constitutional balance of powers. In How to Choose, we discussed various canons of statutory interpretation that can be helpful in resolving conflicts between laws (especially the “last in time” doctrine, and the doctrine that specific laws supersede general laws),46 but the fundamental constitutional question is how a President can minimize his exercise of legislative authority.


The two specific criteria that best capture the minimalism that we endorse are, again, that the President should choose the path that affords him the fewest opportunities to make the types of policy choices that are normally made by legislatures, and that he should try to make decisions that Congress can later negate, if it disapproves of the President’s choices. Under both of these criteria, exceeding the debt ceiling is clearly the least unconstitutional option.47


Looking to our first criterion, exceeding the debt ceiling essentially involves only one choice—to issue additional debt securities48—whereas increasing taxes or cutting spending involves a multitude of different choices that are, as we noted in Part II above, quintessentially legislative in nature. Legislators who prefer military spending to social spending negotiate with legislators who have different priorities, and the resulting set of appropriations laws reflects the fragile political balance that is at the heart of the legislative process. Even across-the-board cuts by the President would not necessarily preserve Congress’s collective will because of the baseline issues discussed in Part II.B, and because some spending levels are almost certain to be more important to maintain than others, from the standpoint of those who voted for the compromises that are inherent to every collection of spending decisions. Moreover, some spending would, as a practical matter, never be reduced in the face of a trilemma—for example, every President would choose to maintain full funding for certain public safety functions, military operations, and so on. Accordingly, it would be impossible for the President to preserve Congress’s exact priorities while cutting spending in response to a trilemma.


Our second criterion—that the President should choose the route that is most reversible—also points toward violating the debt ceiling, and for similar reasons. If the debt ceiling expresses a congressional preference, it is that the aggregate national debt should not exceed a specific dollar amount. If the President in 2013 were to engage in borrowing in excess of Congress’s preferred amount, Congress could reverse that decision by passing a budget for 2014 that included a large enough annual surplus to retire the excess debt that the President created this year. That is not to say that timing means nothing, of course, but the macroeconomic consequences of violating the debt ceiling for less than a year are relatively minor.


No matter what the President chooses under a trilemma, he will of course make decisions that are irreversible in a literal sense. However, the consequences of failing to spend as Congress ordered—to provide funds to feed hungry children, to protect fragile ecological areas, or to prevent military setbacks that involve lost lives and strategic consequences that can stretch out over decades—are not even measurable by the same metrics as those used to measure the small amount of lost (or in light of the current underlying macroeconomics, more likely, gained) economic output from temporarily increasing the government’s debt.


This leads to the ultimate conclusion of this essay: Nothing in our argument changes, even if the negotiations over spending and taxing are carried out by parties who know that the debt ceiling might not later be increased as needed. That is, if at any time in the future Congress chooses to pass taxing and spending laws that will violate the debt ceiling, and the President signs those bills, both parties will do so knowing that a trilemma might actually come into existence. Even so, once that trilemma is actually facing the President, he would be obligated—just as he would have been, had he inadvertently found himself in a trilemma—to minimize the constitutional damage. The answer to the question of how to satisfy the two criteria—minimizing discretion and maximizing reversibility—simply does not depend on who created the trilemma, when it was created, or how.


B. Forcing Congress and the President to Fulfill Their Constitutional Roles

The prospective nature of negotiations over spending and taxation does, however, add an important additional consideration that further supports our conclusion. As noted earlier, our previous papers were limited to analyzing trilemmas that already exist, which means that they can only be resolved by an affirmative decision by Congress to increase the debt ceiling to accommodate its own spending and taxing decisions. In the broader perspective that we take in this essay, however, we must also ask how Congresses in the future might respond if their members believed that the debt ceiling would trump those spending and taxing decisions.


In that situation, any future Congress could evade its constitutional responsibilities by passing appropriations laws that purport to fund broadly popular spending programs, and to do so at generous levels. Congress could also then choose to reduce taxes, another popular decision. Finally, to show that Congress is “fiscally responsible,” it could set (or leave) the debt ceiling at a low level and thus create a trilemma for the President. If the President then (ignoring our analysis) were to engage in the most unconstitutional form of legislating by making spending cuts far below what Congress appropriated, he would not only be arrogating to himself powers that the Constitution bestowed upon the legislative branch, but he would be validating Congress’s attempt to shift political responsibility for unpopular decisions—decisions that are potentially unpopular precisely because they can have real, negative consequences in citizens’ lives.


The proper constitutional analysis, therefore, is not limited to asking which powers the President should be least reluctant to seize. It must also ask how to prevent Congress from shirking its own responsibilities. The power to make the quintessentially political decisions that we have discussed here is quite properly conferred upon Congress, but the importance of those decisions also creates the temptation to avoid the consequences of making difficult choices among competing priorities.


A similar concern arises when we look at the President’s possible strategic considerations during negotiations that might result in a trilemma. If a President harbored a secret desire to exercise extra-constitutional control over spending, he could agree to a combination of spending and taxing that would put him in a trilemma. If, when the trilemma inevitably arrived, Congress passed an increase in the debt ceiling, the President could then veto that increase, and then claim that he is required to start cutting spending, at his discretion. This would amount to a presidentially created line-item veto. As we noted above, the Supreme Court has already ruled that Congress itself cannot delegate item-by-item spending decisions to the President.49 Certainly, the President must not be permitted to achieve the same result on his own.


To be sure, one might think that the bad faith exhibited by a President who vetoed a debt-ceiling increase precisely for the purpose of grabbing legislative power would render his subsequent use of that power constitutionally more defective in virtue of his evil intent. But reducing that intuition to an administrable constitutional principle presents serious challenges. In general, constitutional law eschews motive tests because of difficulties of proof,50 and while the President could be said to know his own motives, they will typically be mixed. A presidential veto does not occur in a vacuum but in the context of a political battle between the President’s policy priorities and those of Congress. We doubt that any President would admit to himself, much less to the public, that he vetoed a bill raising the debt ceiling for the purpose of arrogating power to the executive branch. Accordingly, we do not think that the President’s motives—any more than the motives of Congress—can play a role in the constitutional analysis of a trilemma they necessarily jointly create.


Our analysis, requiring the President to honor Congress’s spending and taxing decisions, thus removes inappropriate strategic possibilities for both Congress and the President. Knowing in advance that the spending and taxing laws that they pass will actually be executed, notwithstanding any gamesmanship with the debt ceiling, does more to preserve the allocation of both the powers and the responsibilities of the separate branches than the widespread but erroneous assumption that the President can cut spending if more borrowing would breach the debt ceiling.


C. Hostage Taking and Political Strategies, in the Shadow of the Constitution

As a matter of negotiation strategy, of course, it might appear that this is all irrelevant. The Republicans in the House would not necessarily negotiate sequentially, saying, “First, we’ll agree on spending and taxes as if there is no debt ceiling; and second, we’ll renegotiate that agreement on more favorable terms for us, by threatening not to increase the debt ceiling.” And the President and his Democratic colleagues would have no reason to go along with that approach. To be sure, that appears to be the direction in which the current Congress and President are heading. With the continuing resolution expiring on March 27,51 and the debt ceiling suspended through May 18,52 it would at least be possible that the spending and taxing decisions will be made two months before a political deal will be struck on the debt ceiling.53 Given recent experience, we certainly see no reason to doubt that the decisions will be made in a chaotic, last-minute fashion—and that each agreement will address only the issues that must be addressed immediately.


If we are wrong, however, those two steps could be effectively condensed into one, with Republicans simply driving a harder bargain from the beginning, confident that they hold a trump card in the debt ceiling statute.54 Therefore, the spending and taxing decisions would not necessarily appear to have been separately affected specifically by the debt ceiling, but the constitutional status of the debt ceiling statute would be crucial in creating bargaining strength, and thus in determining the final outcomes of the negotiations.


Finally, we must acknowledge that the constitutional uncertainty surrounding our analysis is itself an important part of any real-world confrontation over the debt ceiling and a possible trilemma. Our experience in trying to explain our conclusions to journalists and the public leaves no doubt that there is a great deal of confusion about the debt ceiling, and there is strong visceral resistance to the idea that the President can simply, as one interviewer repeatedly put it in questions to one of the current authors, “blow past the debt ceiling.”55


Therefore, even if the Republicans knew, as a constitutional matter, that the debt ceiling is invalid, they might nonetheless threaten to force the issue, daring the President to risk being sued—or even impeached—by following the path that we have described in our work. Even short of those more extreme outcomes, the public’s general sense that the President cannot “just keep borrowing” might tempt Republicans to force the President to take an intensely unpopular stand.


Our analysis cannot, of course, provide an easy answer for any President who might face that kind of threat. When one side of a negotiation is willing to threaten to create either a first-ever default by the government, with terrible economic and social consequences, or force the President to risk taking a stand that would create a constitutional crisis, then the President will have no good options. We can only say that, as a matter of constitutional principles, exceeding the debt ceiling is the least bad option. Even constitutionally irresponsible tactics by one’s opponents, however, must sometimes be taken seriously.


Ultimately, the only way back from the brink is for all of the political actors to behave responsibly, because the courts cannot bail them out. Even if a party could be found with standing to challenge the President’s resolution of a debt-ceiling-induced trilemma, and even if the resulting case were not deemed to present a political question,56 judicial relief would come too late to undo all of the serious economic, political, and constitutional damage that pursuit of any of the President’s unconstitutional options would cause. To avert that damage, the political actors must either agree to repeal the debt ceiling, agree to increase it as necessary, to accommodate the consequences of the taxing and spending laws, or agree to painful spending cuts that would enable the government to pay its remaining bills without new net borrowing. If they follow the latter course, any such spending cuts must be properly enacted by Congress, not unilaterally imposed by the President. As we have emphasized, we believe that such cuts would be a bad idea on the merits; but Congress retains its power to set spending levels, even if it uses its powers to harm the economy by wielding its powers unwisely.


IV. Conclusion

Writing in the summer of 2011, Professor Tribe quoted the second Justice Harlan’s warning that “the Constitution is not a panacea for every blot upon the public welfare.”57 The choice of quotation is arresting. It comes from Justice Harlan’s dissent in Reynolds v. Sims, the case that established the constitutional principle of one person, one vote.58 Harlan was a great Justice, but he was on the wrong side of history in dissenting from the Court’s apportionment cases, for judicial review is easiest to justify when used to disentrench incumbents whose continuation in office depends on an electoral system that systematically denies popular majorities the power to elect legislative majorities.59


Harlan’s caution has even less purchase with respect to the debt ceiling crisis, because his chief concern was how to limit judicial enforcement of the Constitution. Yet as we explained in How to Choose, the President’s obligation to choose the least unconstitutional option may well be nonjusticiable.60 The issue is not what a court will tell the President to do, but what the Constitution requires him to do.


We nonetheless agree with Professor Tribe’s bottom line that it would be best for Congress and the President to reach a political settlement that allows the government to pay its bills. But if Congress, or to be more precise, Republicans in the House of Representatives, refuse to act responsibly, the President will find himself faced with no constitutional options. At that point, the Constitution will be no panacea, but it will remain relevant. The President’s duty will be to choose the least unconstitutional option, and here that means borrowing in excess of the debt ceiling, in exactly the way that Congress’s spending and taxing laws require.


Forewarned is forearmed, and thus all parties should understand that any budget deal they strike that requires borrowing beyond the debt ceiling to close the gap between mandated expenditures and authorized tax revenues will create a trilemma. But the separation-of-powers analysis remains the same regardless of whether the President stumbles upon a trilemma or knowingly signs or vetoes legislation to create a trilemma. Accordingly, those who think that the debt ceiling can be used as leverage in budget negotiations to force spending cuts should think again. Properly understood, the debt ceiling provides no leverage because, properly understood, the debt ceiling is a dead letter.


*Professor of Law, The George Washington University Law School, and Senior Fellow at the Taxation Law and Policy Research Institute, Monash University.


**Robert S. Stevens Professor of Law, Cornell University Law School. The authors thank John Manning for a very helpful discussion, Michael McConnell and Seth Barrett Tillman for usefully pointed questions, and the faculty of the University of Kentucky College of Law (in particular Scott Bauries and Nicole Huberfeld) for extremely helpful insights. Angela Buckner, Robert F. Lehman, and Sergio Rudin provided excellent research assistance.


131 U.S.C. § 3101(b) (2006).


2See Bill Heniff, Jr., et. al., Cong. Research Serv., R41965, The Budget Control Act of 2011 2–3 (2011). The first $917 billion of cuts (spread out over ten years) was approved immediately upon the bill’s signing. At least an additional $1.2 trillion in cuts was included in the bill, to begin to take effect automatically on January 1, 2013 if a “super-committee” did not agree to $1.5 trillion in additional cuts by late 2012. The super-committee failed to produce a plan, but the onset of the additional cuts—the so-called sequester—was later extended until March 1, 2013. As of this writing, the sequester is the focus of intense negotiations, and it is anyone’s guess how much (if any) of the cuts will ultimately be allowed to take effect, and when.


3See id. at 2.


4Neil 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 (2012) [hereinafter Buchanan & Dorf, How to Choose].


5Neil H. Buchanan and Michael C. Dorf, Nullifying the Debt Ceiling Threat Once and for All: Why the President Should Embrace the Least Unconstitutional Option, 112 Colum. L. Rev. Sidebar 237 (2012) [hereinafter Buchanan & Dorf, Nullifying the Debt Ceiling Threat].


6Id. at 238.


7Id. Note that we used the word “budget” in a colloquial sense, meaning essentially “the combination of the spending and taxing laws enacted by Congress.” The formal budget resolutions required of Congress do not, of themselves, obligate the President to spend or tax. Only appropriations laws require the President to spend money: in specific amounts, for specific purposes, at specific times. The Internal Revenue Code requires the President to collect tax revenues, with similar specificity.


8Such measures include temporary suspension of federal purchases of bonds for federal workers’ pension accounts. See Rich Barbieri and Jeanne Sahadi, It’s Official: U.S. Hits Debt Ceiling, CNNMoney (Jan. 2, 2013, 11:05 am), (on file with the Columbia Law Review) (noting Treasury Secretary Geithner’s imposition of “debt issuance suspension period”).


9Pub. L. No. 113-3, 127 Stat. 51.


10Continuing Appropriations Resolution, 2013, Pub. L. No. 112-175, § 106, 126 Stat. 1313, 1315 (2012).


11See, e.g., Ned Ryun, After Big Punt, Here’s How GOP Can Score, RedState (Jan. 24, 2013, 1:23 pm) (on file with the Columbia Law Review) (predicting “[t]he House will use the May debt ceiling to force the Senate to pass a budget which balances in the next 10 years”); Jonathan Weisman & Ashley Parker, Acceptance of Defense Cuts Signals Shift in G.O.P. Focus, N.Y. Times (Feb. 24, 2013), (on file with the Columbia Law Review) (“[M]ost Congressional Republicans are standing their ground [on the sequester], a position they say is strategic. . . .The pain of further cuts to discretionary programs could bring Mr. Obama to the negotiating table on them by the spring, if not by midsummer, when Congress must once again raise the government’s borrowing limit.”).


12Indeed, after the mid-term elections of 2010, President Obama was asked during a press conference why he had not included a debt ceiling increase as part of his year-end agreement with Republicans to extend a package of tax cuts. Press Briefing, President Barack Obama (Dec. 7, 2012, 2:25 pm), available at (on file with the Columbia Law Review). The President replied that he was confident that he and Speaker Boehner would deal with the debt ceiling appropriately when necessary. Id. At the time, the notion that the Republicans would create a genuine debt ceiling showdown was arguably implausible.


13See Buchanan & Dorf, How to Choose, supra note 4, at 1222–43.


14This essay does not address another argument aimed at showing that, even absent an increase in the debt ceiling, the President has a legal option for complying with all of his obligations: the suggestion that the President order the minting of platinum coins worth roughly $2 trillion, to be deposited with the Federal Reserve Board in the Treasury’s account. In How to Choose, we explained that even if this gambit would be technically authorized by statute, it would likely so unsettle the financial markets as to be impractical and for that reason might even count as a violation of Section Four of the Fourteenth Amendment. See Buchanan & Dorf, How to Choose, supra note 4, at 1231. Since then, in our popular writing, we have offered two additional arguments against the legality of “jumbo” platinum coins. First, that in light of the substance of the transaction, they should count against the debt ceiling, and thus violate it. See Neil H. Buchanan, Even After the Coin is Gone, the Legal Analysis is Instructive, Dorf on Law (Jan. 13, 2013, 10:33 am), (on file with the Columbia Law Review); Neil H. Buchanan, If You’re Explaining, Everyone’s Losing (Platinum Coin Edition), Dorf on Law (Jan. 11, 2013, 11:00 am), (on file with the Columbia Law Review); Neil H. Buchanan, Big Coins, Political Credibility, and Hatred of Lawyers, Dorf on Law (Jan. 10, 2013, 10:07 am), (on file with the Columbia Law Review). Our second argument is that applying ordinary principles of administrative law, the statute authorizing the minting of platinum coins cannot fairly be read to delegate to the Secretary of the Treasury the authority to mint coins for this purpose. Michael C. Dorf, Post-Mortem Part 2: Zombie Big Coins, Chevron and the Quadrilemma, Dorf on Law (Jan. 14, 2013, 7:30 am), (on file with the Columbia Law Review). Meanwhile, both the Treasury Department and the Federal Reserve Board (whose cooperation would be necessary for the jumbo coin option to work) have ruled the gambit out. See Ezra Klein, Treasury: We Won’t Mint a Platinum Coin to Sidestep the Debt Ceiling, Wonkblog, Wash. Post (Jan. 12, 2013 3:32 pm), (on file with the Columbia Law Review). Accordingly, we put jumbo platinum coins aside for present purposes.


15Laurence H. Tribe, The Priorities, Room for Debate, N.Y. Times (Jan. 13, 2013 6:31 pm), (on file with the Columbia Law Review). See also Press Briefing, White House Press Sec’y Jay Carney (Dec. 6, 2012, 11:58 am), available at (on file with the Columbia Law Review) (“[T]his administration does not believe that the 14th Amendment gives the President the power to ignore the debt ceiling—period.”).


1631 U.S.C. § 3102(a) (2006).


17Id. §§ 3101, 3101A.


18See Budget Control Act of 2011, Pub. L. No. 112-25, § 301, 125 Stat. 240, 251–55; H.R.J. Res. 98, 112th Cong. (2012); Robert Pear, Senate Vote Approves Rise in Debt Limit, N.Y. Times (Jan. 26, 2012), available at (on file with the Columbia Law Review) (discussing 2012 debt ceiling amendment).


19See, e.g., Act of Sept. 13, 1982, Pub. L. No. 97-258, 96 Stat. 938; Act of Jan. 12, 1983, Pub. L. No. 97-452, 96 Stat. 2467.


20See, e.g., Nat’l Fed’n of Indep. Bus. v. Sebelius, 132 S. Ct. 2566, 2668 (2012) (Scalia, Kennedy, Thomas, and Alito, JJ., dissenting) (stating that Court has sometimes but not always applied presumption of severability).


21The dissenters in the Affordable Care Act case summarized the Court’s severability precedents as requiring two inquiries: first, whether the valid portion of the statute can function as Congress designed it without the invalid portion; and second, whether Congress would have enacted the statutory scheme without the invalid portion. See id. at 2668–69. Here both requirements would be satisfied. Congress’s overall enactments can function better without the debt ceiling, because its elimination resolves the trilemma. As for legislative intent, the congressional pattern of periodically raising the debt ceiling without amending the borrowing authorization indicates that Congress has for decades understood these statutory provisions as separate and thus severable.


22Congress did just that with respect to a sex-based line for Social Security survivor benefits. Although the Supreme Court upheld the classification against an equal protection challenge in Heckler v. Mathews, 465 U.S. 728, 745–51 (1984), the Court’s discussion of the plaintiff’s legal standing made clear that it would have respected Congress’s judgment of nonseverability if it had found the classification invalid. See id. at 738–40 (discussing power of Congress to remedy inequality by leveling down).


23See Buchanan & Dorf, How to Choose, supra note 4, at 1188–94.


24Congressional Budget and Impoundment Control Act of 1974, Pub. L. No. 93-344, 88 Stat. 297 (codified at 2 U.S.C. §§ 681– 688).


25Buchanan & Dorf, Nullifying the Debt Ceiling Threat, supra note 5, at 244–47.


26See id. (discussing Train v. City of New York, 420 U.S. 35 (1975) and Clinton v. City of New York, 524 U.S. 417 (1998)).


27The Debt Ceiling and the Constitution, On Point with Tom Ashbrook, 90.9 WBUR (Jan. 7, 2013, 10:00 am), [hereinafter On Point].


28See Buchanan & Dorf, How to Choose, supra note 4, at 1199–1201, and cases cited therein.


29See Buchanan & Dorf, How to Choose, supra note 4, at 1202–03 (discussing this and other methods of statutory interpretation that might avoid constitutional infirmity, but rejecting them).


30E-mail from Seth Barrett Tillman to authors (Jan. 20, 2013, 09:41 EST) (on file with the Columbia Law Review).


31See, e.g., Department of Defense and Full-Year Continuing Appropriations Act, 2011, Pub. L. No. 112-10, Div. A, 125 Stat. 38, 38; Disaster Relief Appropriations Act, 2012, Pub. L. No. 112-77, 125 Stat. 1277, 1277 (2011); Department of Homeland Security Appropriations Act, 2004, Pub. L. No. 108-90, 117 Stat. 1137, 1137 (2003).


32See, e.g., 42 U.S.C. § 301 (2006) (“For the purpose of enabling each State . . . to furnish financial assistance to aged needy individuals, there is hereby authorized to be appropriated for each fiscal year a sum sufficient to carry out the purposes of this subchapter.”); 42 U.S.C. § 701 (“To improve the health of all mothers and children . . . there are authorized to be appropriated $850,000,000 for fiscal year 2001 and each fiscal year thereafter”); 42 U.S.C. § 1396-1 (“For the purpose of enabling each State, . . . to furnish (1) medical assistance on behalf of families with dependent children and of aged, blind, or disabled individuals . . . and (2) rehabilitation and other services . . . , there is hereby authorized to be appropriated for each fiscal year a sum sufficient to carry out the purposes of this subchapter.”).


33Mindy R. Levit et al., Cong. Research Serv., R41633, Reaching the Debt Limit: Background and Potential Effects on Government Operations 8–10 (2013).


34See Whitman v. American Trucking Ass’ns, 531 U.S. 457, 472 (2001) (noting that “[i]n a delegation challenge, the constitutional question is whether the statute has delegated legislative power to the agency” because the constitutional “text permits no delegation of those powers”).




36See id. at 474–75 (collecting examples).


37See John Hart Ely, Democracy and Distrust 133 (1980) (describing death of nondelegation doctrine, but characterizing that demise as unwarranted “death by association” with doctrines of old Court that were deservedly interred); Gary Lawson, The Rise and Rise of the Administrative State, 107 Harv. L. Rev. 1231, 1237–41 (1994) (lamenting death of nondelegation doctrine).


38See Larry Alexander and Saikrishna Prakash, Reports of the Nondelegation Doctrine’s Death are Greatly Exaggerated, 70 U. Chi. L. Rev. 1297 (2003) (advancing historical and normative arguments for limits on delegation of legislative power).


39531 U.S. 457 (2001).


40Id. at 474 (citing Panama Refining Co. v. Ryan, 293 U.S. 388 (1935); A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935)).


41In How to Choose, we explained that our analysis should apply even if the courts would deem any challenges to the President’s actions nonjusticiable. See Buchanan & Dorf, How to Choose, supra note 4, at 1241. Hence, our reasoning here can only rest on judicial precedents if they apply within the executive no less than within the courts. Yet Professor Morrison has argued that the principle of constitutional avoidance should not apply with the same force in the executive branch as in the judicial branch when the executive has means of inferring congressional purpose unavailable to the judiciary. See Trevor W. Morrison, Constitutional Avoidance in the Executive Branch, 106 Colum. L. Rev. 1189, 1240­­–45 (2006) (explaining that frequent congressional-executive interactions provide executive with superior access to intended meaning of ambiguous statutory language). However, the case under discussion falls outside of Professor Morrison’s class of cases in which avoidance would not apply inside the executive branch. Total silence is not ambiguity and extra-statutory communications between Congress and the President cannot, consistent with the Article I, Section 7 lawmaking procedures, provide the intelligible principle that the nondelegation doctrine requires. Thus, constitutional avoidance would be properly invoked within the executive branch in these circumstances.


42Under former Democratic House Speaker Richard Gephardt, the so-called “Gephardt Rule” required that all annual spending and taxing negotiations include lockstep changes in the debt ceiling. See Joshua Green, How Dick Gephardt Fixed the Debt-Ceiling Problem, The Atlantic (May 9, 2011 11:36 am), (discussing Gephardt rule). The virtue of the rule was that it neutralized the debt ceiling as a source of post-hoc adjustments to appropriated government spending. We have that virtue in mind when we say that any rule specifying prioritization would be “a kind of” Gephardt Rule. Permanently repealing the debt ceiling would be another example. A law authorizing the President to raise taxes to cover a shortfall would be yet another.


43Why would Congress pass a law appropriating more money than it intends for the executive to spend, in light of the debt ceiling? We think the most likely reason in this context would be an unseemly effort by Congress to win favor with the public, while forcing the President to make the hard choices to cut popular spending programs. See infra Part III.B. However, the more general practice by which Congress writes laws that can be waived by the executive pursuant to an intelligible principle can be defended as consistent with the nondelegation doctrine where formal characteristics of the waiver authority indicate that Congress only set initial policy guidelines so that the executive could weigh their utility in light of new information. See David J. Barron & Todd D. Rakoff, In Defense of Big Waiver, 113 Colum. L. Rev. ____ (forthcoming Mar. 2013).


44To be sure, in How to Choose, we did argue that sub-constitutional harm is a factor in determining the least unconstitutional option, but we did not rely on the harm that would result from spending cuts for our ultimate conclusion that overriding the debt ceiling would be less unconstitutional than such cuts. See Buchanan & Dorf, How to Choose, supra note 4, at 1229–31.


45For an argument in favor of policy discretion, see Eric Posner, The President Has the Power to Raise the Debt Ceiling on His Own, Slate (Jan. 4, 2013, 5:23 pm), (on file with the Columbia Law Review). But see Michael C. Dorf, Debt Ceiling Redux (and Why Merely Calling an Argument “Unpersuasive” or “Unconvincing” is Not in Itself an Argument), Dorf on Law (Jan. 4, 2013, 9:05 pm), (on file with the Columbia Law Review).


46See Buchanan & Dorf, How to Choose, supra note 4, at 1202–05.


47Our third criterion—minimizing sub-constitutional harm—also supports our conclusion. We focus on the first two criteria here simply because we believe that they most clearly capture the constitutional stakes of the crisis that a trilemma would create.


48Strictly speaking, as we acknowledged in How to Choose, there are further choices that must be made when issuing debt, including the size and maturity of the securities that will be sold to the public. Although such choices involve some discretion, they are the types of choices that are regularly delegated to the Treasury in any event. They are thus categorically different from the types of decisions that Congress makes when deciding which potential spending programs to fund, and at what levels. See Buchanan & Dorf, How to Choose, supra note 4, at 1211.


49See supra text accompanying note 26.


50See, e.g., Whren v. United States, 517 U.S. 806 (1996) (rejecting motive test under Fourth Amendment); United States v. O’Brien, 391 U.S. 367 (1968) (rejecting motive test under First Amendment); but see Village of Arlington Heights v. Metro. Housing Dev. Corp., 429 U.S. 252 (1977) (applying motive test under Fourteenth Amendment).


51Continuing Appropriations Resolution, 2013, Pub. L. No. 112-175, § 106, 126 Stat. 1313, 1315 (2012).


52See No Budget, No Pay Act of 2013, Pub. L. No. 113-3, 127 Stat. 51, § 2(a) (suspending debt ceiling through May 18, 2013).


53It might, in fact, be more than two months before the debt ceiling next becomes binding. When the debt ceiling comes back into effect, further “extraordinary measures” by the Treasury will allow the government to operate for a short time after the debt ceiling is technically reached, as it did in mid-2011 and early 2013. One reliable estimate puts the final default date in August 2013. See Steve Bell et al., When Will the Next Debt Limit X Date Be?, Bipartisan Policy Center (Jan. 30, 2013), (on file with the Columbia Law Review) (forecasting “that the next X Date would most likely fall in August, but with a realistic chance of coming even later”). Therefore, the deadline for a debt ceiling deal might extend into the summer.


54See supra note 11 and accompanying text.


55On Point, supra note 27.


56See Buchanan & Dorf, How to Choose, supra note 4, at 1240–41 (discussing justiciability of challenges to various presidential responses to trilemma).


57Laurence H. Tribe, A Ceiling We Can’t Wish Away, N.Y. Times (July 7, 2011), (on file with the Columbia Law Review).


58377 U.S. 533, 624–25 (Harlan, J., dissenting).


59See Ely, supra note 37, at 121.


60See Buchanan & Dorf, How to Choose, supra note 4, at 1242.



Preferred Citation: Neil H. Buchanan & Michael C. Dorf, Bargaining in the Shadow of the Debt Ceiling: When Negotiating over Spending and Tax Laws, Congress and the President Should Consider the Debt Ceiling a Dead Letter, 113 Colum. L. Rev. Sidebar 32 (2013),