Madison’s Waiver

Can Constitutional Liquidation Be Liquidated?

David S. Schwartz *



Non-judicial precedent is a thing of some importance in constitutional law, and the uncertainty of its parameters and status in constitutional interpretation seems to cry out for some sort of formalizing or organizing theory. 1 For recent efforts to do this, see Curtis A. Bradley & Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411, 415 (2012) (arguing that historical practice can generate interpretive meaning about separation of powers when reasons for “institutional acquiescence” are discernable in light of institutional realities); Richard H. Fallon, Jr., The Many and Varied Roles of History in Constitutional Adjudication, 90 Notre Dame L. Rev. 1753, 1773-78 (2015) (arguing that longstanding historical practice can sometimes constitute a “gloss” on constitutional meaning apart from a self-conscious effort to “liquidate” the Constitution’s meaning); Shalev Roisman, Constitutional Acquiescence, 84 Geo. Wash. L. Rev. 668, 674 (2016) (arguing that acquiescence in a longstanding practice lacks sufficient normative foundation to gloss constitutional meaning when not motivated primarily by constitutional analysis). Professor William Baude’s recent article Constitutional Liquidation outlines such a theory, by which indeterminate constitutional meaning can be “liquidated”—clarified and settled—through a “course of deliberate practice” by non-judicial public officials. 2 William Baude, Constitutional Liquidation, 71 Stan. L. Rev. 1, 12-13, 16 (2019). While the bulk of Baude’s scholarly and stimulating article focuses on description, Baude clearly believes that the theory holds great normative potential: Somebody, presumably judges, ought to consider liquidations to be binding, or at least very weighty, in constitutional interpretation. 3 See, e.g., id. at 35 (explaining that “[liquidation] serves several deep constitutional values”). Questions about the normative aspects of Baude’s version of liquidation have been raised elsewhere. 4 See, e.g., Curtis A. Bradley & Neil S. Siegel, Historical Gloss, Madisonian Liquidation, and the Originalism Debate, 106 Va. L. Rev. (forthcoming) (manuscript at 28-42). But philosophers tell us that “ought” implies “can,” so Baude’s theory must first clear the hurdle that it can be meaningfully used as a doctrine. Baude’s article makes a good start but leaves certain critical questions unaddressed. If Baude develops his theory further, he will have to analyze numerous examples of non-judicial precedent to define the contours and limits of liquidation. That is to say, the theory of constitutional liquidation must itself be liquidated—tested and explained by experience.

I. What Is Liquidation?

Baude builds his theory by piecing together fragments of James Madison’s statements about liquidation, which first appear in Federalist 37. There, Madison says that the proposed Constitution is “like all new laws,” which “are considered more or less obscure and equivocal, until their meaning be liquidated and ascertained by a series of particular discussions and adjudications.” 5 The Federalist No. 37, at 236 (James Madison) (Jacob E. Cooke ed., 1961). As Baude explains, “to ‘liquidate’ the meaning of a [constitutional] term means to ‘clarify’ or ‘settle’ it,” derived from the Latin word “‘liquidus’ which could mean . . . ‘clear’ or ‘evident.’” 6 Baude, supra note 2, at 12. According to Baude, as he interprets Madison, constitutional liquidation consists of three elements. Where (1) constitutional language is indeterminate, (2) a “course of deliberate practice” by the legislative or executive branches that (3) results in a “settlement” of the disputed constitutional interpretation should be taken as an authoritative exposition of the previously indeterminate constitutional provision. 7 Id. at 13-21. Baude defines indeterminacy straightforwardly as vagueness or ambiguity, 8 Id. at 14. to which we might add textual incompleteness. A “course of deliberate practice” involves both some significant repeated iterations of the interpretation by the political branches, as well as deliberateness: That is, there must have been some thoughtful debate over the relevant constitutional interpretation. 9 Id. at 16-17. “Settlement” consists of two further elements: “acquiescence,” the idea “that the losers in some sense gave up,” and “public sanction,” which means some manifestation of public approval 10 Id. at 18-20. —perhaps re-election of officials who openly embraced the interpretation in question.

Whether or not it is really attributable to Madison, 11 For arguments that it is not, see Bradley & Siegel, supra note 4, at 44-53; Rod Hills, Will Baude on Madison’s Constitutional Liquidation: A Triumph for Baude, A Failure for Madison?, PrawfsBlawg (July 22, 2018, 1:49 PM), (suggesting that Madison had no theory, but an ad hoc explanation deployed strategically and intermittently). Baude’s outline of liquidation has at least surface appeal. It candidly acknowledges the pervasive problem of constitutional indeterminacy and recognizes that experience can develop the Constitution’s meaning. It appears to show respect for constitutional interpretations of the political branches and “the people,” and relatedly to respect the intuitive idea that precedents need not and should not be limited to judicial decisions.

The initial published reactions to Baude’s theory may be unduly skeptical. 12 See Bradley & Siegel, supra note 4; Hills, supra note 11. This might be explained in part by an association between this theory and Baude’s scholarship on originalism, and in part by the fact that a theory of liquidation can be, and has been, cast in originalist terms. 13 Caleb Nelson has advanced such a theory. See Caleb Nelson, Originalism and Interpretive Conventions, 70 U. Chi. L. Rev. 519, 527-53 (2003). For a normative critique, see Bradley & Siegel, supra note 4, at 31-37. But the concept of “settlement” of the meaning of a constitutional provision, though consistent with, does not entail the originalist concept of “fixation.” Despite some casual suggestions that liquidation “fixes” constitutional meaning, Baude offers a looser theory in which a given liquidation is not necessarily permanent, and early practices are not necessarily “privileged” over later ones. 14 Baude, supra note 2, at 59-63. Having approached the article with a jaundiced eye myself, I wouldn’t be surprised if other non-originalist readers suspect a Trojan Horse lurking in it somewhere. But I find Baude’s nascent theory of liquidation attractive insofar as it tends to undermine several of the analytically and normatively problematic tenets of originalism. Conversely, I think the theory will be diminished if Baude attempts to bend it back to serve originalist commitments.

II. Liquidating Liquidation

Because the abstract concept of constitutional liquidation does not adequately explain itself, its contours must be developed through experience. Baude works through some examples of things that he says are and are not liquidation, but his account remains incomplete (as he candidly admits), and some of his examples could use rethinking. Three significant elements that may limit, or even undermine, a theory of liquidation are insufficiently addressed in Baude’s account: (1) reliance interests in constitutional settlements; (2) the role of judicial decisions in liquidations; and (3) the ability of liquidations to amend the Constitution.

A. Liquidation Versus Mere Reliance: The Texas Example

Numerous episodes in constitutional history involve actions of doubtful constitutionality that have nevertheless become firmly entrenched. Baude offers the example of the annexation of Texas. 15 Texas was annexed through congressional legislation in order to avoid the constitutionally mandated treaty procedure, because proponents knew they could not get the two-thirds majority required to ratify a treaty. Daniel Walker Howe, What Hath God Wrought: The Transformation of America, 1815-1848, at 698-99 (2007). The Louisiana Purchase and the ratification of the Reconstruction Amendments are prominent examples Baude does not mention. 16 Jefferson believed, and his Federalist opponents opportunistically agreed, that there was no enumerated power to acquire new territory. Jill Lepore, These Truths: A History of the United States 170-71 (2018). The Thirteenth and Fourteenth Amendments were ratified by three-fourths of the non-seceding states, not three-fourths of all the states, raising questions about whether the Article V ratification process had been complied with. These questions were compounded by the questionable legitimacy of some reconstructed state governments and the fact that some former rebel states were compelled to ratify as a condition of readmittance to the Union. See John Harrison, The Lawfulness of the Reconstruction Amendments, 68 U. Chi. L. Rev. 375, 377-78, 397-409, 419 (2001). Each one of these is an example of the type of “one off” that Baude suggests is not a true liquidation because it may not offer a usable precedent for anything other than its own acceptance. 17 See Baude, supra note 2, at 17, 51-52. But in fact, we can learn more about liquidation than this from the Texas example.

Crucially, liquidation must actually expound the Constitution. Baude implies as much by defining liquidation as a process that tells us the meaning of a term in the Constitution. 18 Id. at 10-11. This requirement is not merely stipulative: Without it, liquidation is hard to distinguish from the mere political settlement of a constitutional controversy. For liquidation to have any value as a theory, or even validity as a description of a distinct causal process, it must generate some sort of precedent. That means, the liquidating episode must tell us more than “that may have been unconstitutional, but we’re just going to grin and bear it because it was necessary/has been too heavily relied on/is water under the bridge.” Baude agrees that non-judicial precedents which generate a mere result are not “satisfying,” and that liquidation requires something analogous to a “holding.” 19 See id. at 52. For a similar argument, see Roisman, supra note 1, at 713 (arguing that normative considerations require that binding non-judicial precedent supply reasoned constitutional interpretations). Liquidation does not merely settle the controversy itself but expounds constitutional meaning for other cases.

This “expounding” requirement raises a serious question about whether liquidations really exist. To the extent that political settlements of disputes with constitutional implications, like the Texas annexation, are driven by reliance interests more than—or rather than—constitutional interpretations, they may not. To see the potential dominance of reliance interests, consider the disruptions that would be caused by disavowing the constitutionality of the Louisiana Purchase, the annexation of Texas, or the Thirteenth and Fourteenth Amendments. Reliance interests do not flow from constitutional interpretation, but—and this is the key point—the reverse is often true. Settlements that depend predominantly on reliance interests might well bring justifying constitutional interpretations in tow. They might even amend the Constitution rather than “clarify” it, as I will discuss later. Baude neglects reliance interests, but he must explain how they fit into his theory of liquidation.

B. Judicial Liquidation? The National Bank Controversy

A complete theory of liquidation would have to explain whether judicial confirmation and elaboration of a non-judicial precedent is an element of liquidation, whether a court decision could form part of or intertwine with a liquidating “course of practice,” and why such judicial intervention does not undermine the very concept of non-judicial liquidation. These questions pointedly arise in the National Bank controversy, an episode on which Baude places great emphasis. The twenty-five years of debate over the constitutionality of a national bank from 1791 to 1816 generated multiple constitutional arguments without clarifying which one(s) provided the basis for the Bank. House supporters of the bill to charter the First Bank of the United States in 1791, and Alexander Hamilton in his famous cabinet memo to President Washington, argued that the Bank was necessary and proper to the nation’s financial administration as well as to a handful of specified powers, and was also an implied power of sovereignty. 20 David S. Schwartz, The Spirit of the Constitution: John Marshall and the 200-Year Odyssey of McCulloch v. Maryland 39-41 (2019). There were also three distinct constitutional arguments in opposition. Eric Lomazoff, Reconstructing the National Bank Controversy: Politics and Law in the Early American Republic 18-26 (2018). When President Madison asked Congress for a new national bank in the aftermath of the War of 1812—the First Bank was killed in 1811 when Congress declined to renew its charter—Bank supporters added the argument that a national bank was necessary and proper to implement Congress’s power to create a uniform national currency. They never reached consensus over the source of this new-found currency power: whether from the Coinage Clause, or implied from multiple constitutional money clauses, or as an inherent power of sovereignty. 21 See David S. Schwartz, Coin, Currency, and Constitution: Reconsidering the National Bank Precedent, 117 Mich. L. Rev. (forthcoming 2020) (manuscript at 7-8) (available at

Ironically, Mr. Liquidation himself, Madison, was of no help in clarifying a consensus view of the power to charter the Second Bank. When he vetoed an earlier 1815 bank bill on policy grounds, he famously “waiv[ed] the question” of Congress’s constitutional authority to charter a bank. 22 James Madison, Veto Message, in 2 A Compilation of the Messages and Papers of the Presidents, 1789–1897, at 540 (James D. Richardson ed. 1898) [hereinafter “MPP”]. Constitutional objections were “precluded . . . by repeated recognitions under varied circumstances of the validity of such an institution in acts of the legislative, executive, and judicial branches of the Government, accompanied by indications, in different modes, of a concurrence of the general will of the nation.” 23 Id. Madison never offered a clear explanation of his own, 24 At most, he hinted vaguely in his 1815 and 1816 annual messages that a national bank might be constitutional as a means to address the nation’s currency crisis. 2 MPP, at 550-51, 563-64. but even if he had, nothing in liquidation theory would have privileged Madison’s individual views as the authoritative liquidation of the Constitution’s meaning. In any case, he was purporting merely to rely on a “liquidation” by others.

Moreover, Madison’s veto message and later correspondence indicate that he saw the Bank question as having been settled by its political acceptance per se. While Madison found it important that the constitutional issues were debated, he seemed unconcerned whether any particular constitutional argument gained a consensus. In 1831, responding to a query from the House Ways and Means Committee on the constitutionality of the Second Bank, Madison revealed the absence of a theory in which “liquidation” generates constitutional interpretation. Madison explained his “waiver” as the product “of the respect due to deliberate and reiterated precedents” after “the act originally establishing a bank had undergone ample discussions.” 25 Letter from Madison to Charles J. Ingersoll (Jun. 25, 1831), in Legislative and Documentary History of the Bank of the United States 780 (1832) [hereinafter “LDH”]. He did not identify what the arguments were. In explaining why the 1811 congressional rejection of the First Bank did not disrupt the “reiterated precedents,” Madison argued that the 1811 vote was determined by policy objections; but “[o]n a simple question of constitutionality, there was a decided majority in favor of it.” 26 Id. Madison’s idea of “liquidation” seemed to care only that members of Congress believed the Bank constitutional (the “simple question”), not why they did (a harder one). The Bank debates as of 1815, culminating in Madison’s waiver of constitutional objections, lacked the element of expounding the constitution.

The authoritative doctrinal explanation for the Bank’s constitutionality was instead set forth in Chief Justice Marshall’s opinion in McCulloch v. Maryland. 27 17 U.S. (4 Wheat.) 316 (1819). Baude’s suggestion that Chief Justice Marshall’s opinion embraced the elements of liquidation is a wishful reading that gets matters backwards. 28 See Baude, supra note 2, at 24. In upholding the Bank, McCulloch said only that the pro-Bank political consensus “ought not to be lightly disregarded.” 29 17 U.S. at 401. The opinion made no effort to uncover the doctrinal reasoning underlying that political consensus. Instead, Chief Justice Marshall’s opinion sets out reasons for constitutionality as though “the question [were] entirely new,” 30 Id. at 402. at least as to the constitutional basis if not the result. Marshall adopts Hamilton’s reasoning, but with subtle changes and without attribution to Hamilton or others, as if to emphasize the Court’s independent right to decide the question afresh.

The Bank Controversy thus raises the question whether courts can—or must—play a role in liquidating non-judicial constitutional precedents. Even a non-judicial precedent as long-running and “thick” with constitutional deliberation as the National Bank controversy did not explain itself. The problem here is not the uncertainty in determining whether a particular constitutional settlement meets Baude’s three elements of liquidation. For example: How long, how iterative, and how deliberative must the “course of practice” be in order to liquidate a question? 31 Similar questions include: How can we be sure the constitutional text was indeterminate? How much must disputing have died away before the question can be called “settled”? What if initial constitutional objections have been forgone but opponents of a practice have continued to dispute it on policy grounds—does that undermine a finding of a settlement of constitutional meaning? What if those opponents dress up their policy arguments as constitutional ones—does that prevent settlement? What if opponents continue the constitutional dispute with far-fetched or bad faith constitutional arguments? Such questions raise the sorts of uncertainty endemic to constitutional adjudication, and liquidation wouldn’t be undermined by the need to make judgment calls. If the elements are sufficiently coherent to guide the inquiry, as I believe they are, the doctrine as a whole can be viable.

But even assuming the elements have been met, it may remain indeterminate what constitutional principle has been expounded by the course of practice, or whether there was a principle at all. This problem will arise either because, as with the National Bank controversy, the deliberative record is overdetermined, meaning that there were too many constitutional justifications, without a clear consensus choice; or underdetermined, as in the Texas example, because the practice was undertaken with only thin or non-existent constitutional deliberation, or without any strong conviction that a constitutional provision supported the action. Baude must explain how judicial intervention factors into liquidations of this sort. And it may be that most or all liquidations are of this sort. 32 The spending power example offered by Baude similarly raises, without resolving, these questions. The spending power debate pitted the “Madisonian” position, that Congress can spend only on matters within its other enumerated powers, against the “Hamiltonian” position, that Congress can spend on any national purpose. President Monroe adopted the Hamiltonian view in the early 1820s, and the practice was consistently followed for the next century. See United States v. Butler, 297 U.S. 1, 66 (1936) (alluding to, but declining to “discuss the legislative practice” that “leads us to conclude that the [Hamiltonian] reading … is the correct one”); David S. Schwartz, An Error and an Evil: The Strange History of Implied Commerce Powers, 68 Am. U. L. Rev. 927, 962-63 (2019) (showing adoption by Monroe, Jackson, and Jacksonian Court of the Hamiltonian view, albeit without attribution to Hamilton, that the federal government could spend outside its regulatory authority); see generally Theodore Sky, To Provide for the General Welfare: A History of the Federal Spending Power 144-303 (2003) (providing examples of spending power legislation in the 1820-1936 period). When the Supreme Court ultimately endorsed Hamilton’s position in United States v. Butler, 297 U.S. 1 (1936), it relied more heavily on Story’s Commentaries than on the century of legislative practice, which the opinion barely mentioned. Id. at 66 (“We shall not review the writings of public men and commentators or discuss the legislative practice. Study of all these leads us to conclude that the reading advocated by Mr. Justice Story is the correct one.”). Most constitutional scholars assume that the question was settled by the Court, rather than by the century-long non-judicial precedent. See, e.g., Erwin Chemerinsky, Constitutional Law: Principles and Policies 274, 278-79 (2007) (basing spending power doctrine on Butler without reference to non-legislative precedent); Laurence Tribe, American Constitutional Law 834 (2000) (explaining that “the scope of the General Welfare Clause remained open through the first third of the twentieth century, since the Supreme Court did not have occasion to interpret the clause until the 1936 case of United States v. Butler”).

C. The New Deal and Constitutional Amendment

A closely related question is whether a liquidation can amend the Constitution. Baude defines Madisonian liquidation as precluding alteration or amendment of the Constitution. 33 Baude, supra note 2, at 20-21, 57, 65-66. But the theory would be strengthened if Baude were to back away from this unduly formalistic position. The first element of liquidation already incorporates the notion of indeterminacy. Therefore, by definition, a valid liquidation could not override clear constitutional text. I cannot think of any non-trivial, non-ludicrous sense in which a liquidation can be said to “amend” or “alter” determinate constitutional text.

The “no-amendment” rider (on top of an indeterminacy requirement) seems to serve no function other than to undermine liquidations that the interpreter dislikes, simply by asserting that the Constitution “really” means something other than the liquidated meaning. This back door would admit all sorts of (perhaps primarily, perhaps only) originalist arguments. For example: “Notwithstanding the indeterminacy of the Necessary and Proper Clause, the National Bank was really contrary to the original meaning of the Constitution, and therefore the liquidation can be rejected.”

Yet Madison himself seemed to abandon the idea that some purported original meaning could invalidate a subsequent liquidation. In one of his two 1831 letters to the House Ways and Means Committee on the constitutionality of the Second Bank, Madison made this revealing point. The 1815-1816 currency crisis, he wrote, “was not foreseen; and if it had been apprehended, it is questionable whether the constitution of the United States, which had so many obstacles to encounter, would have ventured to guard against it by an additional provision.” 34 Madison to Charles J. Ingersoll, Feb. 2, 1831, in LDH, at 778. Thus Madison at the end of his life stuck to his guns that the Bank was not constitutional according to the best interpretation of the original Constitution, but for him that original meaning was overridden by the non-judicial precedent—indeed, by the need for the Constitution “to be adapted to the various crises of human affairs,” as Chief Justice Marshall put it in McCulloch. 35 17 U.S. at 415 (emphasis in original).

What if the constitutional amendment arises, not in opposition to determinate constitutional text, but rather to Supreme Court precedent? This raises the instance of the “New Deal settlement,” the major shift in constitutional interpretation by the Supreme Court between 1937 and 1941, which overruled the Lochner-era jurisprudence that had sharply restricted the power of Congress and state legislatures to regulate economic affairs. Baude curiously ignores this most profound and enduring example of a possible liquidation, which is the elephant in the room for any discussion of liquidation. The New Deal settlement overrode judicial interpretations of the Commerce Clause and other constitutional provisions to determine that the federal government can regulate (virtually) all aspects of the national economy. We can debate the extent to which this constitutional liquidation was produced by the political branches or the Courts, and this circles back to the earlier question of whether liquidation theory should include constitutional interpretations produced by some intertwining of judicial and non-judicial precedent. The key point here is that Baude will hamstring his theory of liquidation if he insists that it cannot “amend” the Constitution. Among other things, this would drag him into unproductive definitional disputes such as whether the New Deal settlement amended the Constitution or “merely” corrected a mistaken (albeit longstanding) judicial interpretation of indeterminate constitutional provisions. He will also miss out on an intriguing popular constitutionalism twist: that in a democratic society, non-judicial liquidations might perhaps have preferred status as a means of overruling “wrong” judicial interpretations of the Constitution.

III. Liquidation and Originalism

Liquidation is fundamentally at odds with originalism, that unruly family of theories which maintain that correct constitutional interpretations are constrained by determinate meanings that are permanently fixed by the intentions or understandings of the Framers or Ratifiers. The idea of discerning the meaning of the Constitution through adaptations to unforeseen circumstances over the course of time certainly seems prima facie incompatible with originalism’s “constraint principle” and “fixation thesis.” 36 The “constraint principle” holds that courts and perhaps other constitutional decisionmakers are bound by original meaning. The “fixation thesis” stipulates that original meaning is permanently fixed as of the time of the pertinent original act—the signing of the Constitution at the end of the Convention, according to intentionalist originalists, or the Constitution’s ratification, according to original public meaning originalists. See Randy E. Barnett & Evan D. Bernick, The Letter and the Spirit: A Unified Theory of Originalism, 107 Geo. L.J. 1, 7-10 (2018) (differentiating intentionalist and public meaning originalist theories); Lawrence B. Solum, The Fixation Thesis: The Role of Historical Fact in Original Meaning, 91 Notre Dame L. Rev. 1, 1 (2015) (explaining constraint principle and fixation thesis). I use the term “Framers or Ratifiers” to encompass both “intentionalist” and “public meaning” originalists. It is insufficient for an originalist to claim to absorb liquidation simply by showing that the Constitution’s original meaning encompassed a methodology of liquidating indeterminate constitutional provisions by subsequent practice. They must also argue that the liquidated meaning becomes permanently fixed and binding for all future time, that such fixation occurs reasonably soon after ratification or at least bears a close connection with Ratification-era public meaning, and that the “liquidators” themselves applied an originalist methodology. A theory of liquidation lacking such elements is indistinguishable from the idea of living constitutionalism: It allows constitutional interpreters to violate the “constraint principle” and to give decisive weight to changed circumstances.

Significantly, Baude does not stipulate any of these originalist elements. Baude nowhere suggests that those who deliberated on a matter that ultimately produces a liquidation should themselves have used an originalist methodology. He remains agnostic about the time frame in which a liquidation must occur, and about whether a liquidation is subject to later revision or correction. 37 Baude, supra note 2, at 59-63. And rightly so. Such originalist strictures on liquidation are subject to many of the unrefuted objections made to originalism itself: that most significant constitutional problems fall into a broad “construction zone” of indeterminacy whose rules are stipulative and weakly connected to the values purportedly served by originalism; that “incorrect” interpretations of original meaning are subject to revision or correction and therefore hardly “fixed”; that early liquidations can produce a “dead hand” problem, and the like. 38 See, e.g., Eric J. Segall, Originalism as Faith 97-102, 114-15, 121 (2018) (summarizing objections to “construction zone” and lack of fixity); Andrew Coan, The Foundations of Constitutional Theory, 2017 Wis. L. Rev. 833, 873 (2017) (describing “dead hand” problem).


Baude’s outline of a theory of liquidation offers a promising start on formalizing a doctrine of an important type of non-judicial precedent. A course of deliberate non-judicial practice might indeed offer interpretations of vague, ambiguous, or underspecified constitutional provisions. With candor and scholarly integrity, Baude acknowledges that the theory must address some “hard[] questions” if it is “to become or remain a part of constitutional interpretation today.” 39 Baude, supra note 2, at 49. In addition to those he sets for himself, I think he must confront the questions set out in this essay and produce more examples to liquidate constitutional liquidation.

* Foley & Lardner-Bascom Professor of Law, University of Wisconsin Law School.