Harvard Law Review Harvard Law Review Harvard Law Review

Harvard Law Review Forum

Historical Gloss: A Primer

Responding to Curtis A. Bradley and Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411 (2012)

In the separation of powers section of the constitutional law canon, no case dominates discussion more than Youngstown Sheet & Tube Co. v. Sawyer.1 The case is distinguished not only by its evocatively industrial, midcentury sobriquet, “the Steel Seizure Case,” but by the number of talismanic phrases that it injected into the constitutional law conversation. “Zone of twilight,” “imperatives of events and contemporary imponderables,” and the summing of executive and legislative powers through “plus” and “minus” operations all appear in the concurrence by Justice Robert Jackson that has come to be understood as the “law” of Youngstown.2

In contrast both to Justice Jackson’s scheme of assessing presidential power in light of specific instances of congressional action, and to the categorical distinction between executive and legislative power endorsed in Justice Hugo Black’s majority opinion, Justice Felix Frankfurter offered his own Delphic addition to the permissible sources of presidential power. Rejecting the “inadmissibly narrow conception of American constitutional law” that would hew only to “the words of the Constitution” and ignore “the gloss which life has written upon them,” Justice Frankfurter added a third source of presidential authority to Justice Black’s Constitution and Justice Jackson’s Constitution-plus-Congress.3 This “gloss on ‘executive Power’ vested in the President by § 1 of Art. II”4 stemmed from “a systematic, unbroken, executive practice, long pursued to the knowledge of the Congress and never before questioned.”5 Over time, such practices had become “part of the structure of our government.”6

A key element of this gloss originated in the actions of Congress: “long-continued acquiescence” from the legislative branch could add “decisive weight to a construction by the Executive of its powers.”7 To back up this theoretical assertion, Justice Frankfurter offered fourteen pages of historical appendices relevant to the specifics of the steel seizure: a “Synoptic Analysis of Legislation Authorizing Seizure of Industrial Property” and a “Summary of Seizures of Industrial Plants and Facilities by the President.”8 Justice Frankfurter thus posed a broader inquiry into congressional acquiescence in steel-seizure-like scenarios, in contrast to Justice Jackson’s analysis of Congress’s specific responses to President Truman’s gambit. Ultimately, Justice Frankfurter held that the history of congressional and presidential practice amounted not to acquiescence, but rather to a “conscious choice”9 not to grant the president seizure authority, a choice made manifest in the Taft-Hartley Act of 1947.10

In their provocative article, Professors Curtis Bradley and Trevor Morrison seek to fill a scholarly gap focusing on the Frankfurterian strand of separation of powers reasoning. They describe their project as examining the “role of historical practice in discerning the separation of powers,” with special attention to the “actual dynamics of congressional-executive relations.”11 From historical practice, or historical gloss (about which terms more infra), Bradley and Morrison turn to a related inquiry raised in Justice Frankfurter’s concurrence. According to the concurrence, evidence of the “gloss which life has written upon” the words of the Constitution” could be divined from a “systematic, unbroken executive practice” if that practice was accompanied by “long-continued acquiescence” by Congress.12 Bradley and Morrison follow Justice Frankfurter’s analysis by first considering the meaning of acquiescence and then suggesting that acquiescence is necessary in order for a given set of practices to constitute a sufficient gloss on executive power to become a source of presidential authority.

This short response addresses three central themes raised by Bradley and Morrison’s article: the meaning of historical practice; its shortcomings as a source of constitutional authority; and the significance of the “Madisonian model” for the separation-of-powers scheme offered by the article.


The title of Bradley and Morrison’s article, “Historical Gloss,” is itself a gloss on the words of Justice Frankfurter’s concurrence. This usage is consistent with that of much other constitutional law scholarship; the phrases “historical gloss” and “the gloss of history” are common in discussions of Youngstown.13 In the first sentence of the article, however, the terminology shifts from “historical gloss” to “historical practice.”14 Throughout the article, Bradley and Morrison appear to use the terms interchangeably.15 “[I]nvocations of the historical gloss method tend to emphasize long-term accretions of practice,” they note at one point, implying a subtle distinction between the two.16 For the most part, however, the article treats the phrases as synonymous.

But just what is “historical practice,” and is it in fact the same thing as the “historical gloss” that posterity has attributed to Justice Frankfurter? Bradley and Morrison are surprisingly reticent on the content and contours of their idea of historical practice. This omission is disappointing in an article that aims to remedy what the authors rightly describe as a dearth of “sustained academic attention to the proper role of historical practice.”17 The phrase “historical practice” remains largely unexamined throughout the article. Bradley and Morrison explore several possible meanings of Justice Frankfurter’s related notion of acquiescence (agreement, waiver, accumulated wisdom18 ), but the foundational concept of historical practice remains fuzzy. Is it custom, tradition, prescription, or something else? Or should we instead see it as something more theoretical — as, perhaps, an element of America’s unwritten constitution?19 Not all past activities or habits rise to the level of Frankfurterian historical gloss. But the reader is left uncertain as to what counts as relevant, actionable historical practice. If historical practice is synonymous with custom, the question then becomes how the relevant custom can be identified retrospectively from the disorderly mass of mere day-to-day actions by governmental actors. The diversity of possible historical practices that touch on a given constitutional dispute requires a practitioner of this mode of analysis to state very clearly what counts, and what does not, as relevant practice. Historical practice is a slippery, unhelpfully capacious notion masquerading as a mid-twentieth-century neutral principle. It was so in 1952 and has only become more problematic since then. Today, it defies translation into modern constitutional thought. Embedded within historical gloss analysis is a view of history as a social science providing objective, falsifiable data about the actions and reactions of governmental actors. To pose the question, “What is the historical practice?,” is to presume an artificial degree of unity and coherence within institutions, and from one action to another. Whose historical practice matters, and which moment encompasses the relevant distillation of that practice, are complex questions. Messiness, unspoken accommodation, and explicit disagreement abound.

To see the range of competing historical practices that a legal interpreter might consult, consider the example of a particular interbranch dispute: the early-nineteenth-century battles over the constitutionality of the Bank of the United States. A key element of Treasury Secretary Alexander Hamilton’s financial architecture, the First Bank was chartered in 1791 but met its demise in 1811 at the hands of the Jefferson¬ian Congress. Amidst the economic distress that accompanied the War of 1812, Congress passed legislation chartering a new incarnation of the Bank. In 1815, President James Madison, who had opposed the creation of the First Bank, took veto pen in hand against the Second Bank, only to sign a renewed charter into law in 1816, citing public support for the Bank and postwar exigency as the reasons for his reversal. Maryland fought the Bank with its famously destructive tax in 1818; in 1819, the Supreme Court upheld the constitutionality of the Bank in McCulloch v. Maryland.20 Even after that decision, Ohioans began to attack their local branches of the Bank, using taxes as well as force, in the form of raids on specie. In 1824, Chief Justice Marshall upheld federal jurisdiction on behalf of the Bank, against its Ohio opponents.21 In 1832, President Andrew Jackson declared his opposition to the Bank and vetoed its recharter; the veto stood, and over the next four years, President Jackson hastened the Bank’s demise by removing federal deposits. By 1836, the Bank had been reduced to a state-chartered Pennsylvania institution.22

If we could ask an observer in 1836 to describe the historical practice between the various branches of government on the subject of the Bank, what might she say? Strong presidential, and even stronger Cabinet-level, support for the Bank, followed by decisive congressional resistance; then presidential resistance but congressional support; then presidential support; then redoubled presidential resistance. Moreover, she might point to repeated endorsements of the Bank’s constitutionality by the Court, met with presidential disdain, opposition and self-help from some states, congressional grants of jurisdiction, and, eventually, the exercise of raw presidential power. The relevant moment of historical practice would matter: 1815, 1816, 1819, or 1832?

To be clear: the history is not indeterminate. We know a great deal about what these actors said, did, and wrote, and historians have methodologies to try to interrogate what political and legal actors thought they were doing, and why they thought they were doing it.23 But the “historical practice,” in the Frankfurterian sense in which Bradley and Morrison use that term, is indeterminate. Absent an account of which practices matter, Bradley and Morrison do not provide a satisfactory explanation for why historical gloss is worth salvaging.


Here lies a deeper confusion in the article. To mount an argument for the continued relevance of historical gloss to separation of powers disputes, as Bradley and Morrison do, is to make a broader claim about the proper sources of constitutional authority. They characterize the goal of their project as descriptive: “to specify what is entailed in historical practice¬—based arguments about executive and legislative power, and to identify the factors that are critical for evaluating such arguments.”24 Moreover, they take pains to disclaim normativity: for instance, while they advert to the growth of executive power since World War II, they note, “Unlike some, we do not intend this description to reflect any normative judgment.”25

On the contrary, however, the central premise of Bradley and Morrison’s argument is that historical practice continues to be, and ought to be, a valuable interpretive tool. Indeed, they state that “[t]o the extent past practice predicts the future actions of the branches, it should arguably inform legal analysis because descriptions of what the law is should have some correspondence to operational reality.”26 For Bradley and Morrison, then, defining the “operational reality” that currently obtains between institutions is a necessary predicate to determining what the future relationship of those institutions should look like. The article thus circles back to the question that Bradley and Morrison claim to be avoiding: how should historical practice be used in constitutional interpretation? And, more important, where does historical practice acquire its authority?

In order to understand whether a particular activity is constitutional, in both the descriptive and normative senses, one must have a clear sense of which elements one views as establishing constitutional authority: for example, popular understandings, judicial pronouncements, political theory, or constitutional text.27 Bradley and Morrison’s ostensible focus is the use of historical practice as a mode of legal decisionmaking. But their account fails to consider the theoretical justifications that legal and political actors use for employing arguments based on historical practice in some situations but not others. When Justice Frankfurter found that Congress’s decision in the Taft-Hartley Act not to grant seizure authority to the President meant that there was no basis to grant the analogous power to President Truman, he was selecting and weighing among historical practices.28 But do Bradley and Morrison agree with Justice Frankfurter that some congressional or presidential practices ought to be elevated to “part of the structure of our government”?29 Because Bradley and Morrison do not flesh out their conception of historical practice, the reader is left to wonder precisely which actions on the part of Congress or the Executive amount to a basis for deference by another branch. Moreover, it is not clear whether Bradley and Morrison view historical practice as a hermeneutic practice that is internal to the Constitution itself, or whether they regard it as extraconstitutional.30 This uncertainty in turn renders their account vulnerable to Professor Laurence Tribe’s caution (which they cite) that some historical practices may in fact be little more than “conscious end-run[s] around constitutional requirements.”31

At the end of Bradley and Morrison’s analysis, they return to the question of the interpretive status of historical practice. Discussing Myers v. United States,32 in which the Court invalidated the Senate’s attempt to limit the President’s power to remove a postmaster, Bradley and Morrison observe, “the more an interpreter deems nonpractice materials to provide a clear constitutional answer to the question at hand, the less inclined the interpreter will be to allow historical practice to change the Constitution’s meaning.”33 In Myers, the fact that “the weight given to historical practice varies inversely with the clarity of other sources”34 meant that the Court used a pertinent and apparently clear Founding-era textual source to set such a high bar for countervailing historical practice that it became virtually impossible for the Court to find that any such practice existed.35

In other words, historical practice is relevant only in situations where the interpreter has first determined that text is unclear or ambiguous. Or, to put the point another way, Bradley and Morrison fail to acknowledge that only an interpreter who believes that text is sometimes unclear or ambiguous will find practice to be a satisfying alternative source of constitutional meaning. Bradley and Morrison assert that historical practice and textual ambiguity vary inversely, but this analysis misses a vital prior step. Whether a text is ambiguous is itself determined by one’s chosen interpretive method. Interpreters who are strong textualists, structuralists, or originalists will rarely, if ever, reason from historical practice (at least explicitly). The inverse relationship is certainly accurate as a descriptive or functional matter, but it is external to most interpretive systems. To the extent that an interpreter actually believes in a particular mode of interpretation (textualism, certainly, but also common law constitutionalism, originalism, or welfarism), when she is operating within that system, she will typically not engage in a calculus weighing practice and nonpractice evidence. For such interpreters, historical practice is an inferior source of authority. The importance of differentiating between the theory and practice of historical practice lies in recognizing that although the connection between practice and nonpractice, or textual, evidence may be described in terms of an inverse relationship, few legal interpreters actually view the evidence in this way.

As Bradley and Morrison acknowledge, the difference in approach is “methodological.”36 Yet any examination of historical practice must reckon with the hierarchy of legal sources in these different methodologies, and with the fact that the calculus of simply weighing various types of constitutional source materials is not available within many systems of interpretation. The choice of methodology is also a choice among types of evidence, and the methodological choice typically comes first.


Acquiescence is a key indicium of historical practice in Bradley and Morrison’s account. They argue that acquiescence continues to serve as a crucial touchstone for arguments based on historical practice, even though some of the reasons that earlier generations of jurists and scholars might have given for its centrality are no longer as compelling as they once were. Moreover, Bradley and Morrison argue, acquiescence operates differently for the legislative and executive branches, leading them to conclude that the standard for finding acquiescence on the part of Congress should be higher than that for executive branch acquiescence.

Bradley and Morrison criticize previous invocations of historical practice for relying too heavily on a particular concept of acquiescence that they argue assumes a “Madisonian conception of interbranch competition” that is at best outmoded and at worst erroneous.37 According to this “Madisonian model,” Congress and the president are perpetually on “guard against encroachments” on their respective spheres of power.38 Consequently, in considering whether a particular policy or action by one of the political branches qualifies as historical gloss, a finding that the other branch has acquiesced to the prac-tice — and thus consented to the putative encroachment — can be dispositive.

In contrast to this view, Bradley and Morrison argue that acquiescence deserves more serious consideration in light of modern political science scholarship. The Madisonian model, they contend, no longer accurately represents the interactions between the executive and legislative branches. The concept of institutional acquiescence is still useful, but it “needs to be tied more closely to the reality of how the political branches actually interact.”39 In reality, there is no “consistent, robust, interbranch rivalry of the sort envisioned in The Federalist No. 51.”40 Particularly with respect to Congress, they argue, the Madisonian model fails because it does not grasp that members of Congress are not only unable to protect the interests of the legislative branch but are not motivated to do so.41 Moreover, acquiescence operates differently for each of the political branches. The President can act unilaterally; Congress cannot. The executive branch comprises many advisors on constitutional questions, such as the Office of Legal Counsel; Congress lacks such concentrated internal counsel. Congressional silence is notoriously difficult to interpret: is it acquiescence, the result of blockage, or just a sign of inattention? In short, for Bradley and Morrison, the Madisonian model applies more accurately to the President than to Congress, which makes historical-practice analysis difficult, insofar as it requires inquiry into congressional acquiescence.

But what is at stake for Bradley and Morrison in this critique of the Madisonian model? They are careful to point out that they “take no position on whether The Federalist No. 51 accurately or fully reflected Madison’s thoughts” on the separation of powers.42 Moreover, a key element of their argument is that historical-practice analysis is feasible and useful even in a post-Madisonian world. They argue that acquiescence is a useful marker of practice, as long as it is assessed carefully and with sensitivity to the specific dynamics of each branch.43 Yet the Madisonian model haunts Bradley and Morrison’s account. Much of their argument presumes an adversarial relationship between the political branches and largely overlooks the role of judiciary. Their Article for the most part treats the judiciary, especially the Supreme Court, as an arbiter of separation of powers disputes rather than as an active branch of the federal government. But, as the examples of judicial review and the political question doctrine demonstrate, the Court’s decisions on issues of justiciability are assertions of ultimate interpretive authority, however deferential and self-restraining their immediate impact.44 Just as legislative and executive practice generate a historical gloss, so too does judicial practice, insofar as the federal courts might also have a known practice of deferring (indeed, acquiescing) to one or the other of the political branches. A complete account of historical gloss, and certainly any account of its normative desirability, must address the question of courts as gloss producers.

Moreover, if the Madisonian model as presented here has become as much of a cartoonish stereotype as Bradley and Morrison (rightly) suggest, it is not clear why their account must devote so much attention to undermining that model. In The Federalist No. 51, the actual James Madison — writing as Publius — was not focused on giving the legislative branch tools to protect itself. The text of The Federalist No. 51 demonstrates not the highly schematized machinery of the Madisonian model, but a more pragmatic, and realistic, attempt to connect selfish individual motives with the jealousy of the branches.45 Indeed, as The Federalist No. 48 demonstrates, given the history of parliamentary overreaching in seventeenth-century England, Madison and his contemporaries were more concerned about the potential for legislatures to assume all the powers of government, including what we would now term the legislative and executive powers.46

Bradley and Morrison correctly note that the Madisonian model entails several problems. Indeed, implicit in Bradley and Morrison’s critique of Madisonian scholars is the point that to blame Madison for not creating a system in which members of Congress systematically check executive aggrandizement is willfully anachronistic.47 Such a critique faults James Madison for his ostensible failure to have anticipated the insights of late-twentieth- and early-twenty-first-century political science. Furthermore, in the 1830s, as he witnessed a series of conflicts between President Andrew Jackson and Congress, the elderly Madison demonstrated himself to be fully aware of the unmodelable, uncontrollable realities of executive-legislative interactions, as well as the shifting meanings of “practice.”48

But what is left after the rejection of the Madisonian model? The richness of Madisonian — as opposed to “Madisonian” — theory provides a potential source for a more nuanced separation of powers theory. But Bradley and Morrison fail to propose an alternative theory to replace the Madisonian model as a justification for the use of historical practice. Even if they were to give a more precise definition of historical practice and acquiescence, absent such a theory, their analysis is incomplete.


Many types of historical practice matter for constitutional law, but without an account of which practice and whose practice is most authoritative, appeals to the gloss of history risk becoming hopelessly open-ended or distressingly cynical. Even the most formalist Supreme Court decision can become part of the gloss, if we recognize that the Justices, like members of Congress and the President, are themselves producers of historical practice.

1. 343 U.S. 579 (1952).

2. Id. at 637, 635 (Jackson, J., concurring); see also, e.g., David Cole, Youngstown v. Curtiss-Wright, 99 Yale L.J. 2063, 2081 (1990) (reviewing Hongju Koh, The National Security Constitution (1990)); Mark D. Rosen, Revisiting Youngstown: Against the View That Jackson’s Concurrence Resolves the Relation Between Congress and the Commander-in-Chief, 54 UCLA L. Rev. 1703 (2007).

3. Youngstown, 343 U.S. at 610 (Frankfurter, J., concurring).

4. Id. at 611.

5. Id. at 610.

6. Id. at 610–11.

7. Id. at 613.

8. Id. at 615, 620; see also id. at 615–28.

9. Id. at 602.

10. Pub. L. No. 80-101, §§ 206–210, 61 Stat. 136, 155–56 (codified at 29 U.S.C. §§ 176–179 (2006)) (outlining procedures for “National Emergencies” but not providing the President with seizure authority).

11. Curtis A. Bradley & Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv. L. Rev. 411, 413–14 (2012).

12. Youngstown, 343 U.S. at 610, 613 (Frankfurter, J., concurring).

13. See, e.g., Am. Ins. Ass’n v. Garamendi, 539 U.S. 396, 414 (2003) (“historical gloss”); Stephen Dycus et al., National Security Law 27 (4th ed. 2007) (“Gloss of History”).

14. Bradley & Morrison, supra note 11, at 412.

15. See, e.g., id. at 418, 425, 428.

16. Id. at 427.

17. Id. at 413.

18. Id. at 433, 435, 434.

19. See generally Akhil Reed Amar, America’s Unwritten Constitution (2012); Thomas C. Grey, Do We Have an Unwritten Constitution?, 27 Stan. L. Rev. 703 (1975).

20. 17 U.S. (4 Wheat.) 316 (1819).

21. Osborn v. Bank of the U.S., 22 U.S. (9 Wheat.) 738 (1824); see also Alison L. LaCroix, Federalists, Federalism, and Federal Jurisdiction, 30 L. & Hist. Rev. 205 (2012).

22. See generally Ralph C.H. Catterall, The Second Bank of the United States (1903); Bray Hammond, Jackson, Biddle, and the Bank of the United States, 7 J. Econ. Hist. 1 (1947).

23. See Quentin Skinner, Meaning and Understanding in the History of Ideas, 8 Hist. & Theory 3 (1969).

24. Bradley & Morrison, supra note 11, at 485.

25. Id. at 447.

26. Id. at 456.

27. See generally Morton J. Horwitz, A Historiography of The People Themselves and Popular Constitutionalism, 81 Chi.-Kent L. Rev. 813 (2006) (distinguishing between descriptive or “small ‘c’” constitutions and normative or large-C constitutions); Alison L. LaCroix, Rhetoric and Reality in Early American Legal History: A Reply to Gordon Wood, 78 U. Chi. L. Rev. 733 (2011) (examining the theory and practice of popular sovereignty and judicial review during the founding period).

28. See Youngstown Sheet & Tube Co. v. Sawyer, 343 U.S. 579, 614 (1952) (Frankfurter, J., concurring).

29. Id. at 610–11.

30. See Amar, supra note 19, at xi (describing the Ninth Amendment as a “textual portal welcoming us to journey beyond the Constitution’s text”); Richard A. Posner, In Defense of Looseness, New Republic, Aug. 27, 2008, at 32, 33 (“Originalism without the interpretive theory that the Framers and the ratifiers of the Constitution expected the courts to use in construing constitutional provisions is faux originalism.”).

31. Laurence H. Tribe, Taking Text and Structure Seriously: Reflections on Free-Form Method in Constitutional Interpretation, 108 Harv. L. Rev. 1221, 1280 (1995).

32. 272 U.S. 52 (1926).

33. Bradley & Morrison, supra note 11, at 479–80.

34. Id. at 431.

35. Id. at 480.

36. Id. at 431.

37. Id. at 414.

38. Id.

39. Id.

40. Id. at 446.

41. Id. at 441.

42. Id. at 439 n.112.

43. One question raised by the suggestion that acquiescence might be a sign of a permissible “interbranch bargain,” id. at 457, is whether one branch may constitutionally consent to a change in its powers. In the related realm of interstate compacts, the Court has repeatedly held that a state may not consent to an infringement on its sovereignty and has made explicit the analogy between federalism and separation of powers concerns. See, e.g., New York v. United States, 505 U.S. 144, 182 (1992) (“The constitutional authority of Congress cannot be expanded by the ‘consent’ of the governmental unit whose domain is thereby narrowed, whether that unit is the Executive Branch or the States.”).

44. See Bradley & Morrison, supra note 11, at 428–30 (citing judicial review and the political question doctrine as “areas where the judiciary’s proper role is seen as particularly limited,” id. at 429).

45. See generally The Federalist No. 51, at 347 (Jacob E. Cooke ed., 1961); The Federalist No. 48 (James Madison), supra, at 332 (emphasizing that the three branches must “be so far connected and blended, as to give to each a constitutional controul over the others,” or else “the degree of separation [required] . . . can never in practice, be duly maintained”); see also Lance Banning, The Sacred Fire Of Liberty 224 (1995).

46. See The Federalist No. 48 (James Madison), supra note 45, at 333 (pointing to the legislature’s habit of “extending the sphere of its activity, and drawing all power into its impetuous vortex”); see also Banning, supra note 45, at 134 (noting Madison and Jefferson’s experience with the “overbearing power” of state legislatures during the Revolutionary period).

47. See Bradley & Morrison, supra note 11, at 446–47.

48. See, e.g., Letter from James Madison to Martin Van Buren (July 5, 1830), reprinted in 9 The Writings Of James Madison 376, 376 (Gaillard Hunt ed., 1910); Letter from James Madison to Martin Van Buren (June 3, 1830), reprinted in 9 The Writings Of James Madison, supra, at 375, 375–76 (explaining to Secretary of State Van Buren that President Jackson’s veto of the Maysville Road bill had misinterpreted Madison’s famed 1817 veto of the “Bonus Bill,” a controversial federal public-works bill).