Against Historical Practice: Facing Up to the Challenge of Informal Constitutional Change

35 Const. Comment. 79

Stephen M. Griffin*

Contentious disputes over war powers and judicial nominations in the Obama and Trump administrations as well as recent Supreme Court cases have drawn increased attention to the use of historical “practice” in American constitutional law. The use of governmental practice to inform legal analysis has a long pedigree in the American constitutional tradition.[1] In this Essay I will argue that it is nonetheless fundamentally flawed in multiple ways that suggest it should be replaced or, at least, reconstructed. Practice-based accounts of constitutional law should be understood as raising the crucial question of how to understand the phenomenon of informal constitutional change. This is change that is acknowledged by legal authorities as part of the content of constitutional law, perhaps even equivalent in significance to a formal amendment, but has occurred outside the Article V process. To replace the use of practice, I advocate an approach which I call “constitutional change as state building.”[2]

My argument has several steps. In Part I, I describe the current use of practice in constitutional law. Two types of appeals are suggested by the relevant scholarship. First, there is the “comity” version associated with Justice Frankfurter’s influential test in Youngstown.[3] This version holds that repeated actions by one of the political branches of government along with “acquiescence” on the part of the responding branch can create an authoritative interpretation or “gloss” with respect to the meaning of the Constitution—new law in fact.[4] Second, there is the “invitation to struggle” version suggested by Edward Corwin’s famous remark concerning the locus of constitutional power in foreign policy.[5] This version also stresses institutional relationships but sets them within a context of contestation between the branches of government for power and control. This latter version is perhaps best exemplified by the controversy over presidential war powers.[6]

In general terms, I argue that both of these versions are underdeveloped. They ignore the relationship of the various incidents of the use of governmental power to the kind of contextual analysis used by historians. In addition, they fail to answer the practical question of how lawyers and officials are to know when enough incidents of “practice” are sufficient to generate law. Arguably, they should be treated as proto-theories of constitutional change of the kind offered, for example, by Bruce Ackerman and Jack Balkin.[7] Yet they have not been treated as such, perhaps because practice-based accounts are mostly confined to separation of powers doctrine.[8]

To elaborate this argument, I specify seven objections to practice-based accounts of constitutional law in Part II. I illustrate the force of these objections in Part III by discussing several examples drawn from the dispute over presidential war powers. Part IV presents my alternative theory, constitutional change as state building, which requires incidents of governmental action to be institutionalized in a constitutional order before they can be regarded as valid law.

  1. The Role of Practice in Constitutional Law

With respect to the use of governmental practice to inform the content of constitutional law, there would seem to be no shortage of recent examples, thanks in part to the unusual presidency of Donald Trump.[9] Indeed, we are seemingly experiencing a never-ending blizzard of new “precedents.” At least until President Trump was impeached by the House of Representatives, perhaps the most striking was his declaration of a national emergency with respect to the situation at the U.S.-Mexico border, made to unlock funds he used to build the border wall that Congress refused to provide.[10] Informed commentators have argued that Trump’s response to the investigation into Russian influence on the 2016 election set new precedents contrary to those established in the earlier Watergate scandal in terms of the involuntary removal of the FBI Director and the Attorney General.[11] With Trump, there are always more examples.[12] In response to requests for records from the House Democratic majority, the White House signaled it would not cooperate in any way, another break with the past.[13] Trump has also bypassed the Senate confirmation process by purporting to appoint “acting” heads of cabinet departments in a way arguably not allowed by federal law.[14]

The legislative branch has been active as well in the Trump era. The Republican-controlled Senate used the “nuclear option” in order to cut down the time presidential nominees to certain offices are considered on the floor.[15] In addition, the Senate confirmed a nominee from California to the U.S. Court of Appeals for the Ninth Circuit without the consent of either senator from that state. This was a definite departure from the so-called “blue-slip” procedure and thus a break from past practice.[16]

These examples are departures from prior practice and perhaps are “precedents” in some sense. But are they legal precedents? Do they establish new law? That is what is asserted in practice-based accounts of constitutional law.[17] These assertions take inspiration as well as borrow authority from the use of practice by the Supreme Court. The Court’s resort to the use of practice or “gloss” has been reasonably common in constitutional law across the years, especially in separation of powers doctrine.[18]

In the landmark Steel Seizure case,[19] Justice Frankfurter laid down an influential marker with respect to practice-based arguments. Following up on the use of such arguments in cases such as United States v. Midwest Oil[20] and The Pocket Veto Case,[21] he stated:

The Constitution is a framework for government. Therefore the way the framework has consistently operated fairly establishes that it has operated according to its true nature. Deeply embedded traditional ways of conducting government cannot supplant the Constitution or legislation, but they give meaning to the words of a text or supply them. It is an inadmissibly narrow conception of American constitutional law to confine it to the words of the Constitution and to disregard the gloss which life has written upon them. In short, a systematic, unbroken, executive practice, long pursued to the knowledge of the Congress and never before questioned, engaged in by Presidents who have also sworn to uphold the Constitution, making as it were such exercise of power part of the structure of our government, may be treated as a gloss on “executive power” vested in the President by sec. 1 of Art. II.[22]

Several features of Frankfurter’s endorsement of “gloss” are worthy of note. Frankfurter advocates an “operational” approach to constitutional change. Studying how government operates in practice can contribute to our understanding of constitutional meaning—albeit with the important proviso that it cannot contradict the Constitution or statutory law. At the same time, Frankfurter signals that the textualist approach associated with Justice Black[23] is inadequate, a point that perhaps has special force in the context of separation of powers. These features of Frankfurter’s discussion have influenced my own approach to informal constitutional change, discussed in Part IV.

There are more recent examples of the use of practice by the Supreme Court. In NLRB v. Noel Canning,[24] the Court decided for the first time the meaning of the recess appointments clause, which gives the president the power to fill vacancies in executive offices that arise “during the recess of the Senate.”[25] In determining whether it was constitutional for President Obama to appoint members of the National Labor Relations Board (NLRB) during a period when the Senate was holding short pro forma sessions, the Court emphasized the role of historical practice,[26] especially given that the case concerned “the allocation of power between two elected branches of Government.”[27] Citing cases that included McCulloch v. Maryland[28] and The Pocket Veto Case,[29] the Court through Justice Breyer stated they showed “that this Court has treated practice as an important interpretive factor even when the nature or longevity of that practice is subject to dispute, and even when that practice began after the founding era.”[30] He concluded his discussion of practice by saying that where the Court had not interpreted the constitutional clause in question,[31] “we must hesitate to upset the compromises and working arrangements that the elected branches of Government themselves have reached.”[32]

Noel Canning and Zivotofsky v. Kerry,[33] which is something of a companion case, can be seen as embodying what I shall call a “comity” version of the value of examining “accepted understandings and practice”[34] between the political branches. In Zivotofsky, the Court examined practice in determining whether the president’s power to recognize foreign governments is exclusive. It discussed six “historical incidents”[35] at some length,[36] while referring to many more.[37] In general, these incidents happened in full view of Congress and usually with its support. In fact, the Court remarked that “[a]t times, Congress itself has defended the President’s constitutional prerogative.”[38] It concluded that “[t]he weight of historical evidence indicates Congress has accepted that the power to recognize foreign states and governments and their territorial bounds is exclusive to the Presidency.”[39]

The discussion in Noel Canning and Zivotofsky suggests that the comity version of practice depends on there being an overlapping area of agreement between the executive and legislative branches. The historical incidents seem to be well-known to both branches. Hence the responding branch, typically Congress, has the opportunity to object and thus defend its own constitutional position against encroachment or aggrandizement by the initiating branch.[40] In these cases, harmony rather than discord prevails.

In contrast, the controversies engendered by the use of war powers and unilateral executive orders in the Bush, Obama, and Trump administrations represent a different version of the use of practice. In these disputes, there is little agreement between the branches, at least so long as they are controlled by different political parties. It seems that the parties are using separation of powers doctrine and the checks and balances afforded to them by the Constitution as invitations to struggle for the control of policy and political advantage, with the next election cycle always in view. How to combine or reconcile the comity version of the relevance of practice with the invitation to struggle version is not obvious. This suggests an inconsistency within practice-based accounts of constitutional law.

Before moving on, I should make one important distinction. In focusing on practice-based accounts of law, I am separating out analyses of constitutional norms or “conventions” that are not asserted or regarded as having the status of law.[41] For example, Josh Chafetz and David Pozen define constitutional norms specifically as those that regulate the “public behavior”[42] of officials—but without being law. These norms or conventions are probably too numerous and diffuse to be profitably included in legal analysis. So, for example, the fact that President Trump uses Twitter to make official announcements may have created a new norm, but no one has claimed that it set a new “precedent,” legally binding future presidents. In any event, I do not attempt to analyze non-legal constitutional norms in this essay.

  1. The Problems with Practice

I will identify and discuss briefly a series of related problems that cast doubt on the uncritical use of government practice to inform the content of constitutional law. To get the critique off the ground, I do not distinguish sharply among the various kinds of practice-based accounts put forward by the Supreme Court, the political branches, or commentators. So these problems may not exist with respect to all such accounts. It is also likely that there is some degree of overlap among the problems identified and therefore I do not claim they are independent. I first list the problems and then discuss them together in more detail. In Part III, I use the example of presidential war powers to demonstrate their plausibility and applicability to practice-based accounts of constitutional law.

With this in mind, consider these points:

  1. “Practice” is simply a label for an assemblage of supposedly related historical incidents asserted to be relevant to the operation of government. But lawyers and scholars assemble these lists for particular purposes. With respect to interpreting any part of the Constitution, there are usually alternative ways to build the list. That means the choice of incidents requires justification.
  2. The incidents chosen do not necessarily have a single valence with respect to judging the constitutionality of official action in the present. This raises the risk that the attribution of legal meaning to the incidents is being influenced by contemporary values, rather than being extracted in an unproblematic way from the past.
  3. Historical incidents used to support the constitutionality of executive action often leave the contribution or relevance of Congress unclear. As a possible consequence, the idea of practice seems to be used solely to support increases in executive power, a one-way ratchet in effect.
  4. The legal status of past government actions that are morally questionable or worse by contemporary standards is unclear. How are we to regard actions that are connected to “dark times,”[43] such as those that occurred in the antebellum “slaveholding republic,”[44] for example? Relatedly, historical incidents never seem to possess expiration dates, regardless of later legal enactments that may be inconsistent with them.
  5. Accounts of practice leave unclear the relevance of broad legal and constitutional changes spurred by social movements (such as the civil rights movement) that may undermine the normative reasons supporting those accounts.
  6. There is no apparent place in practice-based accounts for taking into consideration the lessons of experience with respect to the excesses of executive and legislative power, including those memorialized in legal enactments.
  7. The foregoing problems suggest that practice-based accounts are in substance rather thin theories of informal constitutional change. Yet they are not helpful in addressing the questions that such theories take on explicitly, including how to determine whether an informal legal change has genuinely occurred.[45]

When judges, lawyers, and scholars appeal to practice or “gloss” in American constitutional law, it is typically because the constitutional provision at issue is unclear. This is often with respect to an action by the executive or legislative branches. A series of historical incidents or events is then invoked to determine the meaning of the provision in question. But what attaches these incidents to the provision? Perhaps they were moments when constitutional actors faced literally the same interpretive problem we do today and resolved it in a way that was repeated across time. Again, however, we might inquire what assures us that these are the only relevant instances. With respect to the federal judiciary we can look to authoritative case reports as a closed set of official actions. But there is no parallel to case reports with respect to the political branches. So the question remains, and it is an important one: how are we to know that these are the only relevant (or most relevant) incidents?

In addition, any historical event is usually embedded in a complex web or pattern of other events. To be sure, some of these events may have little or no significance to constitutional analysis. Yet it is hard to know for sure unless we investigate further, something I suggest rarely happens. Practice-based accounts typically ask us to take it for granted that the events can be detached from the larger pattern without analyzing the significance of the latter. How then can we determine whether the larger pattern of events supports or undermines the relevance of the invoked incidents?

Although it sometimes seems that the executive is the only branch that relies on practice, any informed review of congressional history will show that its members appeal to cameral “precedents” as well. The case of judicial nominations is noteworthy in this regard.[46] Yet the relationship of congressional action to executive branch precedents is often unclear. As argued by Professors Bradley, Morrison, and Siegel in a series of leading articles, the influential test articulated by Justice Frankfurter in Youngstown has problems when applied to Congress.[47] Frankfurter’s test rests on determining whether Congress has “acquiesced” to the executive action. These scholars observed that this test assumes the validity of the familiar Madisonian model of interbranch rivalry, with each branch vigorously acting to check the others. But they argued that, given the difficulty of collective action, political asymmetries between the executive and legislative branches, and the growth of extensive delegations of power to the executive branch, Frankfurter’s criterion leaves Congress at a systemic disadvantage relative to the presidency.[48]

A more general problem is that instances of executive action never seem to indicate the propriety of a correction or reduction in the power of the executive branch. Somehow, the use of practice has morphed into a one-way rachet that favors continual increases in executive power.[49] Given that there have been significant efforts by Congress to curb executive power in recent decades, including in foreign affairs, this seems an unlikely and therefore questionable result. Perhaps practice-based accounts are inherently biased in some way with respect to executive power.

Another issue with the use of practice is the failure to take into consideration the tremendous shift in normative perspective between the past and the present. The founding generation, for example, consisted largely of people who understood themselves to be gentlemen of leisure, many of them slaveholders, who among other things valued republicanism, democracy within distinct limits, and maintaining a deferential society grounded in a sense of personal honor.[50] A party-based mass democracy was unknown and still in the future.[51] These and other normative shifts are well-known to scholars, but their significance is rarely assessed in practice-based accounts. If normative perspectives shift, due to both formal and informal legal changes, should this not affect the significance we attribute to past practices? If this is the case, then we should regard more skeptically the use of practices influenced by the different normative standards that prevailed in the eighteenth and nineteenth centuries.

Certainly some of these normative shifts are memorialized in constitutional amendments. We can assume, for example, that past practices such as slavery are outlawed and therefore cannot be relied on in any way as precedents in the future. But what of official action driven by objectives such as the maintenance of white supremacy and imperialism? I suggest these objectives were not as clearly outlawed compared to specific practices such as slavery. If, on the other hand, we regard them as equally beyond the legal pale, that could have the effect of eliminating reliance on a wide swath of past practices that were intimately connected to these objectives. Surely we should wish to take this into consideration, but it is not always clear that practice-based accounts do this.

Suppose, however, that the normative shift is contained not in constitutional amendments, but rather in important judicial decisions and legislative enactments. With the New Deal and the civil rights movement in mind, we might want to review practices that predate these developments to ensure we do not use those that are inconsistent with the new normative orientation of government. In other words, practice-based accounts are good at spotting the potential relevance of discrete government actions. But they are not built around assessing the relevance of broad informal constitutional changes such as those that occurred in the New Deal and civil rights movement. This again suggests that practice-based accounts should be evaluated as proto-theories of constitutional change. But they have usually not been advanced with this in mind and thus lack the kind of descriptive richness and normative justifications these theories bring to bear.[52]

If there is a common theme that animates this critique, it is that practice-based accounts ask us to accept that practices used long ago can be applied uncontroversially in the present even though the historical context in which they were adopted is gone. Perhaps not surprisingly, practice-based accounts seem to assume, indeed require, reinforcement by a continuous constitutional tradition that has not changed in any material respect since the practice in question began. Isn’t this implausible?

To further develop these points, I will examine the case of presidential war powers.

III. Illustrating the Problems with Practice: Presidential War Powers

The use of government practice in cases such as Noel Canning and Zivotofsky can seem unexceptionable. No matter how controversial President Obama’s recess appointments to the NLRB were as a constitutional matter, in policy terms they could be defended as responsible actions necessary to keep government agencies running. Making appointments, even recess appointments, is not terribly consequential for the operation of the government and interbranch relations. And executive branch lawyers felt that they could justify Obama’s actions in terms of the many incidents in which presidents had made recess appointments.

When we turn to presidential war powers and judicial nominations, however, the use of practice is far more contentious and readily illustrates the problems I identified above. Consider President Obama’s 2011 military operation in Libya, which Professors Bradley, Morrison, and Siegel deploy as a central example.[53] The Office of Legal Counsel (OLC) opinion justifying the Libya operation under the Constitution appealed directly to practice:

This understanding of the President’s constitutional authority reflects not only the express assignment of powers and responsibilities to the President and Congress in the Constitution, but also, as noted, the “historical gloss” placed on the Constitution by two centuries of practice. . . . “Our history,” this Office observed in 1980, “is replete with instances of presidential uses of military force abroad in the absence of prior congressional approval.”[54]

How replete it is could be contested, but there is no doubt there are many examples of the presidential use of military force taken in the absence of authorization by Congress. The OLC opinion featured a report by the Congressional Research Service that is a successor to earlier lists of U.S. military interventions assembled by the State Department.[55] These lists have been used for decades to justify presidential actions in disputes over war powers with Congress, beginning with President Truman’s decision to intervene in Korea.[56] The use of such lists justifying a broad reading of presidential war powers is perhaps the most prominent and consequential appeal to a practice-based account in all of American constitutional law.

Yet these lists are highly problematic and raise many difficult questions that have never received proper answers. What sort of legal authority did such a list constitute? How would we know whether these somewhat random incidents were really evidence of long-standing “practice”? Had they been claimed or recognized at the time by any competent authority? Or had they been recognized only at some later time? Moreover, even if a president had claimed any of the instances as precedent for unilateral military action, was this claim ratified by Congress?[57] The status of the incidents was unclear relative not only to the declarations of war approved by Congress, but to the repeated use of the congressional power to authorize war without making a formal declaration. When Congress acted to declare war or otherwise authorize military action, did this count as an incident worthy of inclusion? How should we assess the relevance of presidential uses of power without Congress against these repeated congressional actions?

One way to make analytical progress would be to provide a historical context for the incidents on the list—to situate the list within the conduct of U.S. foreign policy. As stated in the State Department memorandum justifying Truman’s Korea intervention, the president is understood to have the “authority to conduct the foreign relations of the United States.”[58] The OLC Libya opinion maintained the same position, which is quite plausible.[59] As far as I can tell, however, no one has ever looked closely at the incidents on the list in this light.[60] Once we do, additional troubling questions arise. Consider two seemingly innocuous incidents on the 1950 (and subsequent) lists:

Island of Sumatra…..1832…..To punish natives for attack and seizure of American ship and murder of crew.

Fiji Islands………….1840…..To punish natives for an attack upon Americans.[61]

The Sumatra incident began with an attack by pirates on the crew of a U.S. ship.[62] In response, President Andrew Jackson dispatched a naval vessel with orders to demand compensation and to use armed force if necessary.[63] Instead of pursuing the pirates, the captain decided to attack the nearest port.[64] According to historian George Herring, he “plundered the port, and burned the town, killing as many as two hundred Malays, women and children included.”[65] Once they returned Jackson punished the captain but defended his actions.[66] Meanwhile, as Herring notes, Jackson’s opponents “denounced him for being trigger-happy and bloodthirsty, and for usurping a war-making power rightly belonging to Congress.”[67] The incident was simultaneously an instance of gunboat diplomacy, the use of armed force against a people seen as inferior, as well as an example of American imperialism.[68]

The incident in Fiji is most probably a reference to a famous Pacific exploring expedition conducted by ships of the U.S. Navy.[69] As recounted by Nathaniel Philbrick, two sailors bartering for food were attacked and killed by the native population as a result of a misunderstanding.[70] The crew demanded “immediate and crushing action be taken” in response and the captain declared “war” on the island.[71] He ordered his men to attack a village, which was burned and destroyed as a result, killing many natives including a young girl.[72] When the natives begged him to stop, the captain “lectured them about the power of the white man, insisting that if anything like this should ever occur again, he would return to the island and exterminate them.”[73]

While these reprisals were and are morally objectionable, what should interest us in assessing the coherence of practice-based accounts is how to know which aspects of these incidents are relevant. If these incidents are to be regarded as practice or “precedents” in the common law style for a presidential power to engage in military action without the consent of Congress, they are surely also precedents for gunboat diplomacy, imperialism, and the indiscriminate killing of an indigenous population.[74] Further, given that members of Congress objected at the time, no one can claim them as precedents for presidential action without acknowledging that precedents had been set in favor of congressional power as well, a constitutional check that would seem all the more justified in light of the circumstances.

Once we dig deeper into the context of such incidents, we might be moved to shift the entire inquiry to ask how they relate to the history of American foreign policy. Has it really been under the sole control of the president without any role for Congress? If these exercises of power were typical of the nineteenth century, how relevant are they to the twentieth? Gunboat diplomacy, as represented by these instances as well as the notorious Greytown incident in what is now Nicaragua, is out of style and may well be inconsistent with current international law and U.S. treaty commitments.[75] But if this is true, how can we rely on these incidents today as legal authority? Indeed, should we not conclude that they have been rendered illegal?[76]

What is missing in appeals to practice that rely on incidents scattered across centuries is any sense that the relevant legal background may have changed. Presumably there are incidents where the federal government facilitated chattel slavery in the nineteenth century or racial discrimination in the twentieth in terms of promoting the “red-lining” of neighborhoods.[77] No one argues we should rely on those practices, presumably because their legal relevance was cut off by obviously relevant later enactments such as the Thirteenth and Fourteenth Amendments. Yet there are arguably similar legal cutoffs that pertain to presidential war powers.

In the war powers debate that occurred during the Vietnam era, for example, some scholars argued for the continued legal relevance of the many interventions the U.S. had made earlier in the twentieth century in the Caribbean and Latin America.[78] However, these interventions were emphatically repudiated by President Franklin Roosevelt as part of his “Good Neighbor” policy.[79] Yet the repudiation of these practices by the chief executive does not appear on these lists. If our concern is with an accurate constitutional history of U.S. foreign policy rather than simply listing military actions, how is this justified?

In addition, there is no mention in these lists or in executive branch arguments on war powers of the significance of the fight over the ratification of the Treaty of Versailles.[80] That President Woodrow Wilson failed to obtain Senate ratification of the treaty is well known. What many accounts fail to reflect is that a large part of the reason for Wilson’s failure was a dispute over war powers. During his personal participation in the Paris Peace Conference, Wilson drafted Article X of the proposed treaty as part of his effort to establish the League of Nations.[81] According to historian John Milton Cooper, the Article “guaranteed the political independence and territorial integrity of League members against external aggression, and it required members to take action, even to the extent of using military force, against violators of this guarantee.”[82] During the conference, the French delegation proposed that the League have an independent enforcement arm, an international military force with the ability to intervene anywhere in the world.[83] Wilson objected that assigning U.S. troops to such a force was incompatible with the Constitution, specifically, the control Congress had over whether the U.S. initiated a war.[84]

Once Wilson returned to the United States, Article X became a principal focus. One argument used by opponents was that the Article undermined Congress’s authority to decide whether the U.S. went to war. According to Cooper, Wilson responded that “the president would have to seek legislative authority in order to furnish ‘the necessary means of action,’ and he scoffed at the notion that Congress’s constitutional power to declare war might be impaired in this process.”[85] When the treaty was presented to the Senate, Republican Henry Cabot Lodge, Chair of the Foreign Relations Committee, ensured a reservation was attached to Article X that stated Congress “has the sole power to declare war or authorize the employment of military and naval forces of the United States.”[86] Wilson did not disagree with this point in principle and thought it was important to say that Article X left Congress’s power to initiate war unchanged.[87] The Treaty was nonetheless defeated in the Senate partly because of the worry that the role of Congress in authorizing war would be usurped.[88]

I do not mean to suggest that no one at the time believed that the president had some authority either to defend the territory of the U.S. (in the manner of repelling sudden attacks) or to protect American lives or property abroad without explicit authority from Congress. But there were many dogs that did not bark in the Treaty debate. If there were already preexisting precedents for presidents to commit significant U.S. forces to military interventions abroad without authorization by Congress, as is suggested by the lists cited by the OLC Libya opinion, no one noticed at the time. Article X could have been easily defended by Wilson on the grounds that it contemplated purely defensive actions against international aggression and the president already had such authority founded on practice. Both Wilson and Lodge made no sign that they were even aware of this possibility. This is fundamentally inconsistent with the justification for presidential war powers presented by the State Department and the OLC when they invoke their long lists of military incidents.

What the Treaty debate showed was that earlier presidential interventions were isolated episodes that were not regarded at the time as having the status of legal precedents. We must distinguish carefully between the significance given to such episodes after 1950 and what presidents and congressional leaders thought at the time. Presidents such as Wilson did not in fact claim such episodes as precedent for the kind of unilateral authority asserted by presidents after 1950. As a somewhat more conservative alternative, such interventions could have been regarded as part of a practice of unilateral presidential actions commanding troops to solve problems in foreign affairs. If so, however, the practice did not add up to what the State Department claimed in 1950—the authority to send large forces to other countries to meet the challenge of foreign aggression. If such had been the case, Wilson would have had a far more secure basis for arguing that the Treaty was not a departure from the text of the Constitution or its unwritten traditions.

None of these complex considerations surrounding the ratification of the Treaty of Versailles are so much as hinted at in the accounts of past executive branch actions relied on by the OLC. This illustrates the significance of the points made in Part II—there are always many chains of historical incidents operating at the same time, they often point in no single direction with respect to presidential power, and practice-based accounts tend to leave Congress out while at the same time endorsing past actions that raise serious unacknowledged moral and legal issues.

Another indication that something is awry with these lists is that there is no clear way for them to take account of the role of Congress.[89] Some members of Congress objected to uses of military force at the time they occurred, although the lists do not reflect this fact.[90] But there is a much bigger problem with respect to how these lists treat congressional declarations and authorizations for war. These authorizations, whether made through formal declarations of war or otherwise, have legally grounded many of the most consequential uses of military force in American history, including both world wars and the “9/11 War” that followed the September 2001 terrorist attacks.[91] Yet the original lists and prominent presidential defenses of the use of force such as the OLC Libya opinion accorded them no legal weight. If they are not precedents or “practice” relevant to assessing the role of Congress, what are they?

In addition, there is no obvious way a list of incidents can take proper account of a contextual shift such as the kind of historical sea change with respect to executive power that occurred after the Vietnam War and Watergate.[92] The early Cold War was the apotheosis of executive power, commonly referred to as the “imperial presidency,”[93] typified by the Truman, Kennedy, Johnson, and Nixon administrations (with Eisenhower as something of a special case).[94] As a consequence, in the 1970s Congress undertook a wide-ranging reevaluation of the costs of excessive presidential power on a bipartisan basis and resolved to do things differently henceforth. To be sure, not everything Congress tried worked as planned,[95] but the legacy of “no more Vietnams” proved compelling, at least through the Clinton presidency.[96]

In particular, the passage of the 1973 War Powers Resolution (WPR) should be viewed as cutting off any further reliance on practice-based accounts of presidential war power.[97] Practice-based arguments in favor of broad presidential power, including the use of long lists of military actions, were deployed to convince Congress that the Constitution had been changed informally.[98] Yet Congress deliberated and made a considered judgment to reject such practice-based accounts.[99] Indeed, in the WPR itself, Congress embraced what would now be called an originalist approach by appealing to the design of the framers of the Constitution.[100] Originalist commentary, as well as analyses based on other methods of constitutional interpretation, have repeatedly confirmed the soundness of Congress’s constitutional conclusions.[101]

Members of Congress continue to use the WPR as a means of opposing unilateral presidential action.[102] The steady invocation of a law passed over a presidential veto and backed by overwhelming popular support should surely figure somewhere in executive analyses, such as the OLC Libya opinion. Instead, it is relegated to the background. This is legally questionable.[103] It is at least relevant that in our system of law, statutes typically take precedence over conflicting common law.[104] The same rule should apply to practice-based appeals to presidential war powers. That this has never happened speaks to a basic confusion over the status of practice-based accounts and, by extension, to the nature and role of informal constitutional change in American constitutionalism. Addressing that confusion should be a signal priority for scholars.[105]

To sum up, the executive branch has consistently advanced practice-based accounts of presidential war powers that are not credible. On close inspection, they were never based on any systematic assessment of the historical record. At least partly for that reason, they neglect the role of Congress almost entirely. They ignore the repeated assertions, in matters both large and small in the course of American foreign policy, of Congress’s sole power to declare war (understood as the power to authorize or initiate a major military commitment).[106] Indeed until the advent of the Cold War, Congress was quite firm on this point.[107] After the Vietnam War ended, Congress reassessed presidential war power in the WPR, but this historical incident or point of “practice” has been largely ignored by the executive branch.[108] To put it another way, the executive branch argument based on practice is implausible once we place the exercise of the war power in historical context. This supports the points made in Part II and demonstrates the need to reconstruct practice-based accounts of constitutional law.



  1. Facing Up to Informal
    Constitutional Change

In Part I, I described two types of arguments from practice—the “comity” version in which there is interbranch agreement on the legal content of the interpretation or “gloss,” and the version of interbranch struggle in which there seems an unending argument over the boundaries of constitutional authority. The comity version is exemplified by the practical test advocated by Justice Frankfurter—repeated assertions of authority by the initiating branch validated by the acknowledgment and acquiescence of the responding branch. However, no such plausible approach is available for the arena of interbranch struggle.

Although it may seem we are at a dead end, I suggest progress can be made if we realize that the problems with using practice identified in Parts II and III are related directly to the theoretical difficulty we face in understanding how informal constitutional change occurs legitimately.[109] Broadly speaking, I believe these problems stem from an insufficiently historical approach to constitutional change. We need a model that helps us avoid these problems. At a minimum, this model should be designed to encourage an intensive study of the historical contexts in which governmental action occurs, including the changing relationship among all three branches of government. In addition, the model should help us understand how historical change alters the constitutional framework in ways that go beyond the Article V amendment process.

Happily, such models are readily available in terms of how scholars have been using the concept of a constitutional order in a historically sensitive way.[110] In addition, the Frankfurter test can provide inspiration once we notice that it resembles asking whether the practice in question has been successfully institutionalized. I pick up on this point by emphasizing the role of state building in constitutional change.

The kernel of this approach is that constitutional lawyers and scholars can reliably track informal alterations to the Constitution by focusing on state building or how constitutional orders are implemented through institutions and then reproduced (or not) over time. This is similar to the comity version in that it focuses on whether an official action such as a judicial decision has validated the practice in question. This means that the practice receives institutional support from a branch that is capable of reproducing it.  So understood, it becomes a “precedent.” What we should strive to avoid is what has happened in the debate over presidential war powers in terms of treating bare historical events as “practice.” Such practice-based accounts link incidents like beads on a string, assuming implicitly that each is of equal weight. This kind of analysis fails to engage meaningfully with any relevant historical context.

We can frame the historical context for understanding practice around the process of state building or institutionalization of forms of governance. State building refers to several developments—changes in the structure of government institutions, expansion of their administrative capacities, and increases in the experience, competence, and resources of branches of government, including agencies and bureaucracies. We look in particular for changes that enable institutions to reproduce their institutional abilities and capacities (“practice”) across time.[111] It is possible for this process to go into reverse, as it did with respect to the enforcement of civil rights after the end of Reconstruction.[112] If the expertise and resources of government institutions are demobilized or demolished rather than reproduced, this has implications for how we regard the enforceability and supremacy of constitutional developments that depended on such expertise and resources. “State unbuilding,” that is, can hinder the enforcement of the Constitution and arguably affect its content.

The idea of constitutional change as state building may seem unfamiliar. Consider, however, that the most intense conflicts over changing the Constitution such as those that took place in the antebellum era over issues such as internal improvements and the national bank, Reconstruction, the Progressive era, the New Deal, and the civil rights movement all implicated state building. These conflicts all featured debates and proposals over the powers, organization, and resources of government institutions. This suggests there is a close relationship between state building and constitutional change, which means we can draw from the existing rich literature on political and constitutional development.[113] Furthermore, the intuition I have pursued in my prior work is that a focus on state building can help us sort out which informal changes to the Constitution count as truly fundamental and thus have the status of supreme law. This avoids the problem, common to the British model of the unwritten constitution, of dissolving the fundamentality and supremacy characteristic of the U.S. Constitution in an undifferentiated soup of political events.[114]

To organize the inquiry into the relevance of state building to constitutional change, I suggest the concept of a constitutional order is useful. Whether or not scholars agree with the British model, it has become commonplace to assert that to understand how the Constitution operates today in a practical sense, we should take account of norms, practices, and institutions that go beyond its spare text.[115] The concept of a constitutional order is a useful way of summarizing the reality that, to be effective as a supreme law, the Constitution must be implemented and enforced through institutions. Some of these institutions might be justly designated “formal” because they are created by the text, such as three branches of government. Others are “informal,” such as political parties. But all of them have mediated constitutional meaning throughout American history.

Focusing on constitutional orders also serves as a useful heuristic in terms of enabling us to explore how a Constitution formed against an eighteenth-century institutional baseline changed mostly informally outside the Article V amendment process. My version of this concept features two inquiries. First, we should attend to how the structure and resources available to these institutions influence how the Constitution is enforced over time. In particular, we should keep in view the variable of state capacity, the ability of the state to govern competently and authoritatively.[116] Second, we should assess whether constitutional orders succeed at reproducing themselves across time.

An approach that focuses on state building, institutionalization, and constitutional orders can certainly appear remote from traditional legal argument. But I contend it is highly useful. Once we attend to the arena of interbranch struggle and separation of powers controversies such as war powers, we find that conventional legal argument has limits when evaluating the constitutionality of official action in an area not strongly governed by judicial decisions. In other words, the standard approach to issues posed by the “Constitution outside the courts” wrongly assumes that the methods that characterize judicial adjudication function equally well outside that sphere. Yet the argument presented here shows that there is no strong analogy between judicial cases and the “precedents” set by historical practice. This calls for not only a fresh approach but a recognition that the reason the standard approach fails is that it does not face squarely the challenge posed by informal constitutional change. Informal constitutional change outside the judicial sphere cannot be understood on an adjudicatory model.

With respect to evaluating the exercise of executive power, for example, without a sound approach to informal constitutional change we are stuck in a perpetual blizzard of presidential pseudo-events. We have no way to assess the oft-heard claim that these are “precedents” without a systematic method to separate those that are truly permanent in a legal sense from those that are not. This is where standard practice-based accounts leave us. Fortunately, an approach rooted in describing constitutional orders and appreciating the role of informal constitutional change using a historicist perspective—that is, “constitutional change as state building”—can provide a way forward.

How, then, should we assess executive or legislative claims that their actions have established legal “precedents”? Constitutional change as state building involves a shift in perspective. We no longer ask whether discrete official actions constitute practice, gloss, or precedents. Instead, we assess these claims against the preexisting baseline provided by the constitutional order that exists with respect to war powers, judicial nominations, congressional appropriations, and so on.[117] We search for a relevant historical context in which to embed the claim and assess its validity relative to the existing constitutional order. Suppose, however, we evaluate it as an attempt to change the current order. Such changes must be justified normatively. Mere assertions or claims do not provide such justifications.[118] Constitutional orders are and should be not that easy to change, as they involve not simply adopting new rules and standards but institutionalizing them so they can be reproduced across time, a process analogous to the formal constitutional mechanism for change in Article V.

With respect to the examples discussed in Part III of claims of a presidential power to use military force without the consent of Congress, this should be assessed against a baseline that consists of continued use of congressional authorizations for significant uses of military force along with the repeated invocation of the WPR by members of both parties. So presidentialist claims cannot be supported by merely listing past incidents that occurred during earlier constitutional orders whose baselines have long since been rendered obsolete by later developments. To be sure, constitutional orders can change. Perhaps in the future it will be widely acknowledged, as it was in the early Cold War, that the president should be given broad deference with respect to the use of military force and officials will regard the WPR (if it is still formally on the books) as defunct. But despite arguments to the contrary by publicists for the executive branch, such is not the case today.[119]

A salutary illustration of the shift in perspective this approach involves is helpfully provided in a recent article on judicial nominations by Josh Chafetz.[120] In analyzing this example of the invitation to struggle version of practice, Chafetz eschews reliance on supposed prior Senate precedents. After all, as he demonstrates, since 2004 the Senate has unceremoniously torched multiple prior “precedents” that it supposedly had been committed to previously.[121] As he remarks, “Indeed, the term ‘unprecedented,’ along with a few nearly synonymous phrases (‘never before in the nation’s history,’ ‘defies centuries of practice,’ ‘violates a two-hundred-year-old tradition,’ etc.), arises with such frequency that it takes on an almost ritualistic character.”[122] This does not describe a situation in which it makes sense to assume, as practice-based accounts invariably do, that Congress establishes precedents for itself in the same manner as the judiciary supervising case law.

Yet if we do not speak of practice and precedents, how should we conceptualize informal constitutional change outside the courts? Chafetz instructively substitutes a narrative that explains interbranch interaction on judicial nomination controversies—that is, he describes the structure of the constitutional order of which they are a part. He discusses two such structures—legislative obstruction by minority political parties and the politics of deferring to presidents in the federal appointments process.[123] He finds legislative obstruction to be a strong constant in Congress across time and that presidents run into more difficulty on appointments when their party does not control the Senate or when their own public support is low. Chafetz’s argument has the specific payoff of undermining the argument that there was a preexisting solid base of precedent that Senate Republicans departed from when they refused to consider President Obama’s nomination of Judge Merrick Garland.[124]

Chafetz’s discussion helps point up one feature of studying constitutional change as a succession of institutionalized constitutional orders. Once we conceptualize the task this way, it is difficult to deny that fundamental features of our political universe like political parties are part of the constitutional order.[125] Indeed, put this way, most would concede they have been at least since the early nineteenth century. So political parties, the dynamics of the two-party system, and the basic structure of party politics are endogenous to the constitutional order. This means judicial nominations are a series of moves within a constitutional order constituted by the actions of both parties. If we dislike the outcomes produced by this order, this kind of analysis focuses our attention in the right place—on changing the structure of the party system rather than bemoaning the violation of a supposed lasting precedent.

As a final comment, I should address how approaching constitutional change in this way addresses the issues with practice I identified in Part II. The theme of the seven problems is that practice-based accounts are unjustifiably thin theories of constitutional change. We should aim at producing robust, historicist theories of informal constitutional change. Once we do so, the problems I specified fall away. Rather than stringing together isolated and possibly unrelated incidents of government action, we instead build credible historical narratives describing the evolution of constitutional orders.[126] The structure of government, party politics, and American political and moral values, to name only three examples, are elements in these narratives. They of course tend to move with the times. So thus do constitutional orders and what they are understood to permit or deny.


Appeals to practice in American constitutional law come in two varieties—the “comity” version in which there is interbranch agreement and the version of interbranch struggle in which there seems to be an unending argument over the boundaries of constitutional authority. Although there are doctrinal tests for the comity version that allow analytical process to be made, no such helpful tests exist to solve the issues created by the arena of interbranch struggle. I have argued that we can make progress by realizing that practice-based accounts of constitutional law are actually proto-theories of informal constitutional change. As theories, however, they are not robust and do not lead in productive directions. Reconceiving practice-based accounts in terms of the historicist study of constitutional orders will enable us to better understand the process of informal constitutional change.




*     W.R. Irby Chair and Rutledge C. Clement, Jr. Professor in Constitutional Law, Tulane Law School. Copyright 2019 by Stephen M. Griffin. Email: An earlier version of this essay was presented at a May 2019 conference at Boston College on “Amending America’s Unwritten Constitution.” I am grateful for the comments I received from Carolyn Shapiro, Josh Braver, and the other participants.

         [1].    See, e.g., Curtis A. Bradley & Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411 (2012); Curtis A. Bradley & Neil S. Siegel, Historical Gloss, Constitutional Conventions, and the Judicial Separation of Powers, 105 Geo. L.J. 255 (2017); Michael J. Glennon, The Use of Custom in Resolving Separation of Powers Disputes, 64 B.U.L. Rev. 109 (1984); Tara Leigh Grove, Article III in the Political Branches, 90 Notre Dame L. Rev. 1835 (2015).

         [2].    See, e.g., Stephen M. Griffin, Understanding Informal Constitutional Change (Tulane Public Law Research Paper No. 16-1),

         [3].    Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 610–11 (1952) (Frankfurter, J., concurring).

         [4].    Id.

         [5].    Corwin stated: “[T]he Constitution, considered only for its affirmative grants of power capable of affecting the issue, is an invitation to struggle for the privilege of directing American foreign policy.” Edward S. Corwin, The President: Office and Powers, 1787–1984, at 201 (5th rev. ed. 1984) (originally published 1940).

         [6].    For an extended treatment of the modern controversy over war powers as a case study of informal constitutional change, see Stephen M. Griffin, Long Wars and the Constitution (2013).

         [7].    See, e.g., Bruce Ackerman, We the People: The Civil Rights Revolution (2014); Jack M. Balkin, Living Originalism (2011).

         [8].    Bradley and Morrison note the resemblance of theories of constitutional change, such as Ackerman’s, to the use of historical practice. Bradley & Morrison, supra note 1, at 426–27. They distinguish theories of change by saying they focus on “critical turning points” rather than “long-term accretions of practice.” Id. at 427. This distinction, however, raises one of the points I discuss in this Essay—whether such long-term accretions are plausible, given changing historical contexts.

         [9].    For a useful sizing up of the unusual features of the Trump administration, see Neil S. Siegel, Political Norms, Constitutional Conventions, and President Donald Trump, 93 Ind. L.J. 177 (2018).

       [10].    See, e.g., Anita Kumar & Caitlin Oprysko, Frustrated Trump Lashes Out After Border Defeat, Politico (Feb. 15, 2019),

       [11].    Peter Baker, Mueller’s Investigation Erases a Line Drawn After Watergate, N.Y. Times (Mar. 26, 2019),

       [12].    See, e.g., Nancy Cook, Trump Bulldozes Across the Presidency’s Red Lines, Politico (Apr. 12, 2019),

       [13].    Anita Kumar, Trump Officials Prepared to Stonewall Democratic Oversight Demands, Politico (Mar. 19, 2019),

       [14].    Michael Tackett, Another Day, Another ‘Acting’ Cabinet Secretary as Trump Skirts Senate, N.Y. Times (Apr. 8, 2019),

       [15].    Glenn Thrush, Senate Republicans Go ‘Nuclear’ To Speed Up Trump Confirmations, N.Y. Times (Apr. 3, 2019),

       [16].    Deanna Paul, ‘Damaging Precedent’: Conservative Federal Judge Installed Without Consent of Home-state Senators, Wash. Post (Feb. 28, 2019),

       [17].    See, e.g., Bradley & Morrison, supra note 1, at 417–24.

       [18].    Id.

       [19].    Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579 (1952).

       [20].    236 U.S. 459 (1915).

       [21].    279 U.S. 655 (1929).

       [22].    Youngstown, 343 U.S. at 610–11 (Frankfurter, J., concurring).

       [23].    In these remarks, Frankfurter appears to be referencing Black’s opinion for the Court, which he plainly regards as inadequate.

       [24].    573 U.S. 513 (2014).

       [25].    U.S. Const. art. II, § 2, cl. 3.

       [26].    Canning, 573 U.S. at 543–44.

       [27].    Id.

       [28].    17 U.S. 316 (1819).

       [29].    279 U.S. 655 (1929).

       [30].    Canning, 573 U.S. at 525.

       [31].    Id.

       [32].    Id.

       [33].    135 S. Ct. 2076 (2015).

       [34].    Id. at 2091.

       [35].    Id.

       [36].    Id. at 2091–94.

       [37].    Id. at 2093 (referring to “over 50 recognition decisions made by the Executive”).

       [38].    Id. at 2094.

       [39].    Id.

       [40].    Keeping in mind the problems for this view discussed in Bradley & Morrison, supra note 1.

       [41].    See, e.g., Josh Chafetz & David E. Pozen, How Constitutional Norms Break Down, 65 UCLA L. Rev. 1430 (2018); Keith E. Whittington, The Status of Unwritten Constitutional Conventions in the United States, 2013 U. Ill. L. Rev. 1847 (2013).

       [42].    Chafetz & Pozen, supra note 41, at 1433.

       [43].    I borrow this term from Jack Balkin. See Balkin, supra note 7, at 93–99.

       [44].    See Don E. Fehrenbacher, The Slaveholding Republic (2001).

       [45].    See Ackerman, supra note 7.

       [46].    See, e.g., Josh Chafetz, Unprecedented? Judicial Confirmation Battles and the Search for a Usable Past, 131 Harv. L. Rev. 96 (2017).

       [47].    See Bradley & Morrison, supra note 1; Bradley & Siegel, supra note 1.

       [48].    See Bradley & Morrison, supra note 1, at 438–47.

       [49].    The Noel Canning case might be considered a counterexample, but it in fact points in both directions.

       [50].    These points are suggested by Michael J. Klarman, The Framers’ Coup: The Making of the United States Constitution (2016); Gordon S. Wood, The Radicalism of the American Revolution (1991); Gordon S. Wood, Revolutionary Characters: What Made the Founders Different 14-15, 22-23 (2006). The founding generation were—or aspired to be—“gentlemen of leisure” in the sense that they were not dependent on others for employment or income. Id. at 16-17. On the importance of honor to the founding generation, see Joanne B. Freeman, Affairs of Honor: National Politics in the New Republic (2001).

       [51].    See, e.g., Daryl J. Levinson & Richard H. Pildes, Separation of Parties, Not Powers, 119 Harv. L. Rev. 2312 (2006).

       [52].    Ackerman, supra note 7; Balkin, supra note 7.

       [53].    Bradley & Morrison, supra note 1, at 453–54, 458–59, 465–66; Bradley & Siegel, supra note 1, at 261–62.

       [54].    Authority to Use Military Force in Libya, 35 Op. O.L.C. 1, 7 (2011) [hereinafter OLC Libya opinion].

       [55].    See id. (citing Richard F. Grimmett, Cong. Research Serv., R41677, Instances of Use of United States Armed Forces Abroad, 1798–2010 (2011)).

       [56].    For discussion, see Griffin, supra note 6, at 77–85.

       [57].    Interestingly, this point was noted by future Chief Justice Rehnquist in 1970. See William H. Rehnquist, Assistant Att’y Gen., Office of Legal Counsel, “The President and the War Power: South Vietnam and the Cambodian Sanctuaries,” Memorandum to Charles W. Colson, Special Counsel to the President, May 22, 1970, at 8.

       [58].    U.S. Dep’t of State, Authority of the President To Repel the Attack in Korea (July 3, 1950), in 23 Dep’t St. Bull. 173, 173 (1950) [hereinafter Bulletin].

       [59].    OLC Libya opinion, supra note 54.

       [60].    Scholars like John Hart Ely and Michael Glennon have suggested the incidents on the list fall into categories not consistent with a broad understanding of presidential war powers such as gunboat diplomacy, encounters with pirates, and the occasional rescue of American citizens. See John Hart Ely, War and Responsibility 10 n.54 (1993); Michael J. Glennon, The Cost of “‘Empty Words”: A Comment on the Justice Department’s Libya Opinion, Harv. Nat’l Sec. J.F. 3–4 (2011).

       [61].    Bulletin, supra note 58, at 177. The lists provided in the later war powers debates were consistent in retaining these incidents. See, e.g., War Powers Legislation: Hearing Before the S. Comm. on Foreign Relations, 92d Cong. 359–75 (1971) [hereinafter 1971 Hearings]. They remain in the latest compilation by the Congressional Research Service. See supra note 55.

       [62].    See George C. Herring, From Colony to Superpower: U.S. Foreign Relations Since 1776, at 170 (2008).

       [63].    Id.

       [64].    Id.

       [65].    Id.

       [66].    Id. at 170–71.

       [67].    Id. at 171.

       [68].    Id.

       [69].    Nathaniel Philbrick, Sea of Glory: America’s Voyage of Discovery, The U.S. Exploring Expedition, 1838–1842 (2003).

       [70].    Id. at 215–21.

       [71].    Id. at 222–23.

       [72].    Id. at 223–27.

       [73].    Id. at 230.

       [74].    See Herring, supra note 62, at 171.

       [75].    This incident featured a wildly disproportionate use of force by the Navy against the port of Greytown in 1854. Id. at 219–20; Francis D. Wormuth & Edwin B. Firmage, To Chain the Dog of War 37–41 (1986). This lamentable incident nonetheless resulted in a case that has been cited in favor of broad executive power in foreign affairs to protect the lives of American citizens. See Durand v. Hollins, 8 F. Cas. 111 (C.C.S.D.N.Y. 1860) (No. 4,186).

       [76].    I have been advised by international law scholars that such incidents today might violate the law of war and entitle the victims to compensation by the U.S. government.

       [77].    On the latter, see, e.g., Richard Rothstein, The Color of Law: A Forgotten History of How Our Government Segregated America (2017).

       [78].    See Griffin, supra note 6, at 83.

       [79].    Id. at 83–84.

       [80].    The material here and in the next three paragraphs is adapted from id.

       [81].    John Milton Cooper, Jr., Breaking the Heart of the World: Woodrow Wilson and the Fight for the League of Nations 10–11 (2001).

       [82].    Id. at 11.

       [83].    Id. at 52.

       [84].    Margaret MacMillan, Peacemakers: The Paris Conference of 1919 and Its Attempt to End War 102–03 (2001). See also Thomas J. Knock, To End All Wars: Woodrow Wilson and the Quest for a New World Order 205 (1992) (stating Secretary of State Lansing’s objections to proposed treaty along similar lines).

       [85].    Cooper, supra note 81, at 118.

       [86].    Id. at 194, 226.

       [87].    Id. at 142.

       [88].    Herring, supra note 62 at 429–30.

       [89].    The latest version of this list from the Congressional Research Service fixes this problem by including wars authorized by Congress. See supra note 55.

       [90].    Harold Bruff, whose work I greatly respect, remarks in discussing passage of the WPR that “[b]y 1973, there had been hundreds of large and small presidential uses of the military that had not previously been questioned.” Harold H. Bruff, Untrodden Ground: How Presidents Interpret the Constitution 337 (2015). My point is that some of them were questioned when they occurred, but the fact is that for lack investigation, we do not know for sure. No scholar has ever looked closely at the historical context of these various events.

       [91].    See Griffin, supra note 6, at 45–51.

       [92].    These events relevant to judging the appropriate exercise of presidential power were related, as I detail in id. at 165–74.

       [93].    See Arthur M. Schlesinger Jr., The Imperial Presidency (1973).

       [94].    Griffin, supra note 6, at 104–09, 120–52.

       [95].    See Eric A. Posner & Adrian Vermeule, Jr., The Executive Unbound: After the Madisonian Republic 85–89 (2010).

       [96].    Griffin, supra note 6, at 174–81.

       [97].    For an admirably straightforward argument that is similar, see Bradley & Morrison, supra note 1, at 467.

       [98].    See 1971 Hearings, supra note 61.

       [99].    It is surely relevant that President Nixon cited practice in vetoing the WPR. Congress’s overriding of Nixon’s veto should be viewed as a rejection of the way he used practice-based arguments with respect to war powers. See Veto of the War Powers Resolution, 1973 Pub. Papers 893, 893–95 (Oct. 24, 1973).

     [100].    See Pub. L. No. 93-148, 87 Stat. 555 (1973) (codified at 50 U.S.C. §§ 1541–1548 (2000)).

     [101].    See, e.g., Griffin, supra note 6, at 35–40 (including sources cited).

     [102].    Most recently with respect to Yemen. See S.J. Res. 7, 116th Cong. (2019).

     [103].    Bradley & Morrison, supra note 1, at 467.

     [104].    See, e.g., Mohamad v. Palestinian Auth., 566 U.S. 449, 457 (2012).

     [105].    See, e.g., Bradley & Morrison, supra note 1.

     [106].    For this reading, see Griffin, supra note 6, at 45–51.

     [107].    See id.

     [108].    See Bradley & Morrison, supra note 1.

     [109].    Griffin, supra note 2.

     [110].    There are many examples. See, e.g., Ackerman, supra note 7; Balkin, supra note 7; Philip Bobbitt, The Shield of Achilles: War, Peace, and the Course of History (2002); Mark A. Graber, A New Introduction to American Constitutionalism 18–23 (2013); Mark Tushnet, The New Constitutional Order (2003).

     [111].    See Griffin, supra note 6, at 23–25.

     [112].    See Eric Foner, Reconstruction: America’s Unfinished Revolution, 1863–1877, at xxvi (1988).

     [113].    See, e.g., Brian Balogh, A Government Out of Sight: The Mystery of National Authority in Nineteenth-Century America (2009); Michele Landis Dauber, The Sympathetic State: Disaster Relief and the Origins of the American Welfare State (2013); Max M. Edling, A Revolution in Favor of Government: Origins of the U.S. Constitution and the Making of the American State (2003); Daniel R. Ernst, Tocqueville’s Nightmare: The Administrative State Emerges in America, 1900–1940 (2014); Jerry L. Mashaw, Creating the Administrative Constitution: The Lost One Hundred Years of American Administrative Law (2012); William J. Novak, The People’s Welfare: Law and Regulation in Nineteenth-Century America (1996); Karen Orren & Stephen Skowronek, The Search for American Political Development (2004); Stephen Skowronek, Building a New American State: The Expansion of National Administrative Capacities, 1877–1920 (1982).

     [114].    See Griffin, supra note 2.

     [115].    See, e.g., Ernest A. Young, The Constitution Outside the Constitution, 117 Yale L.J. 408 (2007).

     [116].    See Griffin, supra note 6, at 14–15.

     [117].    Perhaps it will turn out that they are all related to a master, overarching constitutional order such as the “New Deal constitutional order” or the “Reagan constitutional order.” This is a matter for descriptive investigation.

     [118].    For a similar point, see Glennon, supra note 60, at 134–35.

     [119].    See Bradley & Morrison, supra note 1; supra note 102.

     [120].    Chafetz, supra note 46.

     [121].    Id. at 97–110.

     [122].    Id. at 110.

     [123].    Id. at 110–27.

     [124].    See id. at 128–30.

     [125].    See, e.g., Levinson & Pildes, supra note 51.

     [126].    For an excellent example of providing narratives rather than stringing together incidents of practice, see Tara Leigh Grove, The Origins (and Fragility) of Judicial Independence, 71 Vand. L. Rev. 465 (2018).