In a classic book on constitutional interpretation, Professor Philip Bobbitt identified six “modalities” of constitutional argument: historical, textual, structural, doctrinal, ethical, and prudential.Footnote 1 In Historical Gloss and Foreign Affairs, Professor Curtis Bradley, of the University of Chicago Law Faculty, argues convincingly that Bobbitt’s list is incomplete. Historical gloss is a distinct, seventh modality that courts and other government officials utilize when they engage in constitutional construction. Moreover, Bradley demonstrates, historical gloss is especially important as a tool of constitutional construction in the field of foreign affairs law.
Bobbitt’s “historical” modality focuses on the intentions of the Constitution’s framers. His “doctrinal” modality focuses on judicial precedent. Historical gloss, says Bradley, is a type of “nonjudicial precedent” (p. 15). It is “a different type of historical modality, focused on post-Founding governmental practice” (p. 6). Bradley shows that “[the] constitutional law governing the United States’ conduct of foreign affairs … has been shaped and defined by the accretion of governmental practice” (p. 1). More specifically, “constitutional understandings relating to foreign affairs authority have often been developed through the practices of Congress and the executive branch” (id.). One of Bradley’s main goals in the book “is to confirm gloss’s status as one of the important modalities of constitutional interpretation” (p. 6). In this reviewer’s judgment, Bradley has succeeded admirably in accomplishing that goal, insofar as he is advancing a descriptive claim that gloss is a distinct, seventh modality utilized by various actors in making constitutional arguments.
However, at times Bradley goes beyond description to suggest that historical gloss can be used to settle disputes about contested constitutional issues. Whereas Bobbitt talks about modalities of “constitutional argument,” Bradley speaks about modalities of “constitutional interpretation” in a manner that suggests that these modalities, including historical gloss, can provide definitive answers to disputes about constitutional meaning. On this point, I am more skeptical.
Bradley identifies three requirements for “crediting gloss”—that is, for treating gloss as an authoritative guide to constitutional meaning. There must be “(1) governmental practice (2) that is longstanding and (3) concerning which the affected branch of government has acquiesced” (pp. 25–26). Although practice must be longstanding, it “need not date to or near the Founding period, and modern practice can potentially qualify as gloss even if it differs from earlier practice” (p. 27). Bradley acknowledges that “[t]he most debated element of the three requirements for gloss concerns acquiescence” (p. 28). To meet the third requirement, he says, “the practice must have operated for a significant amount of time without generating continued interbranch contestation” (p. 29). I will say more later about the “acquiescence” requirement. At this point, I note that there is no agreed standard for determining how much acquiescence is enough, or how much interbranch contestation is too much. Absent such a standard, the acquiescence element provides fertile ground for argumentation, but casts doubt on the use of gloss as a tool for achieving a definitive settlement of disputes about constitutional meaning. (Of course, in this respect, one could argue that gloss is no different than several of Bobbitt’s other modalities.)
Chapter One presents and defends the book’s central thesis: that historical gloss is a distinct and important modality of constitutional argument. Chapter Two notes that the text of the Constitution is under-determinative with respect to many important foreign affairs issues. For example, Article I grants Congress the power to declare war, and Article II specifies that the president “shall be Commander in Chief,” but the Constitution does not specify how those two powers interact with each other. Bradley contends that gloss helps fill gaps where the text is indeterminate.
Chapters Three through Six focus on concrete issues to further illuminate the role of gloss in constitutional interpretation. These chapters, respectively, address the president’s power to recognize foreign governments (ch. 3), the president’s power to make international agreements (ch. 4), the president’s power to terminate treaties (ch. 5), and the president’s power to utilize military force (ch. 6). All four chapters present case studies that show, in detail, how historical gloss has contributed to the development of current constitutional understandings related to the distribution of power between the president and Congress over discrete foreign affairs issues. For anyone interested in the history of U.S. foreign affairs law, Chapters Three through Six present a veritable gold mine of historical information and analysis. Both scholars and practitioners will undoubtedly refer to these chapters for many years to come to shed light on the historical evolution of constitutional understandings in these four doctrinal areas.
In an important law review article published more than two decades ago, Professor G. Edward White documented the rise of executive power in foreign affairs, and the corresponding decline in congressional power, during the period between World War I and World War II.Footnote 2 Bradley’s narrative in Chapters Four and Five, addressing the president’s power to make and terminate international agreements, is broadly consistent with White’s claim that the period after World War I was a critical period in the rise of executive power over foreign affairs. However, Bradley’s analysis challenges White’s chronology in certain respects. For example, Chapter Three shows that the president’s power to recognize foreign governments has deep historical roots dating back to the late eighteenth century (see pp. 52–53).
Moreover, Bradley’s analysis of the president’s power to use military force (ch. 6) challenges the conventional wisdom that “Truman’s commitment of troops to the Korean War in 1950 without congressional authorization marks a sharp break from earlier understandings of presidential war powers” (p. 144). In his view, “if there was a turning point in presidential war powers, it was more likely the period following the Spanish-American War, a period that began fifty years before the war in Korea” (p. 145). Bradley cites with approval a memorandum written by William Rehnquist, when he served as head of the Justice Department’s Office of Legal Counsel, that described the Korean War as the “high water mark of executive action without express congressional approval” (p. 139). Bradley contends that the Korean War “is an outlier in modern practice” (p. 144) because the “most significant military conflicts since the enactment of the War Powers Resolution … have been authorized by Congress” (p. 140).
Bradley notes that “the most common objection to relying on gloss … is that it will tend to favor expansions of executive authority” (p. 30). He concedes that, “when interpreters credit gloss, they should be attentive to the danger of executive aggrandizement” (p. 32). Even so, the main purpose of Chapter Seven is to show that concerns about the danger of executive aggrandizement are overstated. Chapter Seven describes several “examples of congressional (and senatorial) powers relating to foreign affairs” (p. 147) where reliance on historical gloss as a tool of constitutional construction tends to favor congressional power over executive power. Bradley is undoubtedly correct to note that historical gloss does not always operate as a one-way ratchet in favor of executive power. However, it appears to this reviewer that gloss usually enhances executive power at the expense of congressional power. Indeed, in the three doctrinal areas that are addressed in Chapters Four through Six—making international agreements, terminating treaties, and use of force—Bradley’s analysis confirms that historical gloss has expanded executive power relative to legislative power since the beginning of the twentieth century. Given recent expansive claims of executive power—such as President Trump’s claim that the United States will “take over the Gaza strip”—the concern that gloss frequently contributes to executive aggrandizement cannot easily be brushed aside.
Chapter Eight—the book’s final chapter—addresses congressional delegations of foreign affairs authority to the president. Bradley rightly notes that “nondelegation concerns have not played much of a constraining role in the foreign affairs area” (p. 169). His analysis suggests that, even if the Court decides to give more “teeth” to the non-delegation doctrine in purely domestic matters,Footnote 3 the Court will likely continue to acquiesce in broad delegations of authority related to foreign affairs. That conclusion is consistent with the Supreme Court’s analysis in Curtiss-Wright,Footnote 4 which suggested that courts should be more deferential to broad delegations from Congress to the president in foreign affairs, compared to domestic affairs. Even so, I have one minor quibble with Bradley’s analysis.
Bradley claims that “there is a long-standing and bipartisan practice of broad congressional delegations of foreign affairs authority to the executive branch” (p. 189). The claim that this is a bipartisan practice ignores the “separation of parties,” as opposed to the separation of powers. In an important law review article, Professors Daryl Levinson and Richard Pildes argued that our constitutional system of checks and balances, in practice, relies heavily on the division of power between political parties, as opposed to (or in addition to) the formal division of powers among the three branches.Footnote 5 In light of their analysis, one might hypothesize that Congress is more likely to delegate broad authority to the president when a single party controls the presidency and both Houses of Congress. In contrast, one would expect Congress to be more reluctant to delegate broad authority in periods of divided government. Bradley’s analysis does not account for this separation-of-parties framing. The implication for historical gloss is clear. Subsequent interpreters should be more reluctant to credit historical gloss as a tool for constitutional construction if that gloss is based on single-party control of the government, especially if that gloss-based argument supports executive aggrandizement.
The remainder of this review focuses on Chapter Four, which addresses the president’s power to make international agreements without resorting to the Article II treaty process. Although Bradley contends that historical gloss firmly supports the president’s power to bypass the requirements of Article II by using executive agreements to make binding international commitments, the following analysis suggests that Congress may not have acquiesced in this practice to the degree that is commonly assumed. This analysis relates back to the distinction introduced earlier in this essay between gloss as a modality of constitutional argument and gloss as a method for settling disputes about constitutional meaning. The following analysis of the president’s power to make international agreements illustrates the point that historical gloss is very useful as a modality of argumentation, but much less useful for providing definitive, right answers. This should not be surprising, because disputes about constitutional meaning are inherently normative, and historical analysis rarely provides “right” answers to normative questions.
Making International Agreements
International agreements can be divided between “treaties”—which require a two-thirds majority vote in the Senate under Article II—and “executive agreements,” which create binding international obligations for the United States, but do not require a two-thirds vote in the Senate. Between 1789 and 1889, the United States entered into 275 treaties and 265 executive agreements.Footnote 6 Between 1889 and 1939, the ratio of executive agreements to treaties was about 1.75 to 1. However, after 1939, reliance on executive agreements grew dramatically. Between 1939 and 1989, the United States concluded 11,698 executive agreements and 702 treaties, a ratio of about 16.7 to 1.Footnote 7 Between 1990 and 1999, the United States entered into 2,857 executive agreements and 249 treaties, a ratio of about 11.5 to 1.Footnote 8 (The slight decrease from 16.7 to 1 to 11.5 to 1 does not appear to be significant.) The increasing reliance on executive agreements as a substitute for Article II treaties transferred substantial power over the creation of international agreements from the Senate to the president.
The U.S. Supreme Court decisions in United States v. Belmont Footnote 9 (1937) and United States v. Pink Footnote 10 (1942)—both of which approved the use of executive agreements as a substitute for Article II treaties—undoubtedly contributed to the dramatic change in the ratio of treaties to executive agreements after 1939. Bradley’s account understates the importance of Belmont and Pink in shifting the path of treaty-making away from the Article II process and toward greater reliance on executive agreements. He maintains that gloss develops from the interactions between Congress and the executive branch, but the Court’s decisions in Belmont and Pink demonstrate that the Supreme Court can have a significant impact on those interactions when it intervenes in foreign affairs issues. (This point is not limited to the topic of treaty making. For example, the Supreme Court’s decision in Zivotofsky substantially strengthened presidential control over recognition of foreign governments, the subject of Chapter Three.Footnote 11 )
Bradley contends that historical gloss supports the growth in presidential power to make international agreements without obtaining a two-thirds majority in the Senate. Referring to his three factors for “crediting gloss,” we undoubtedly have “(1) governmental practice (2) that is longstanding” (pp. 25–26). However, the story of congressional acquiescence (the third factor) is more complicated. To appreciate this point, we must first distinguish among the different types of executive agreements.
Bradley’s account relies on the traditional distinction between “sole executive agreements” and “congressional-executive agreements.” Sole executive agreements (SEAs) are binding international agreements concluded on the basis of the president’s independent constitutional authority, without congressional approval. In contrast, congressional-executive agreements (CEAs) depend on either ex post or ex ante congressional authorization. Ex post CEAs are approved by a majority vote in both Houses of Congress after the text of the international agreement has been finalized. For ex ante CEAs, the president claims authority to enter into the international agreement based upon a statute enacted by Congress that pre-dated negotiation of the agreement.
Bradley claims that “the vast majority” of executive agreements “are of the ex ante variety” (p. 75). However, it is virtually impossible to know whether this claim is true. The proper classification of a particular agreement as either an SEA or an ex ante CEA depends largely on how broadly or narrowly one construes the statute (or statutes) that the president invokes as a source of ex ante authority for entering into the international agreement. As Professors Daniel Bodansky and Peter Spiro have noted: “Since the President often does not specify the legal basis for joining an executive agreement, whether the basis is ex ante congressional authorization or independent presidential power is often unclear, making classification impossible.”Footnote 12 Similarly, Professor Harold Hongju Koh has concluded that “the always imaginary ‘triptych’ of Article II treaties, congressional-executive agreements, and sole executive agreements … is dying or dead.”Footnote 13
In short, many executive agreements occupy the shadowy netherworld between SEAs and ex ante CEAs, where reasonable people could disagree about whether a particular agreement is properly classified as an SEA, a CEA, or neither. The “neither” category includes agreements that are arguably beyond the scope of the president’s independent authority, and that might not be authorized by statute under a narrow construction of the relevant statute. I will refer to executive agreements that occupy the grey area between SEAs and ex ante CEAs as “unidentified executive agreements,” or UEAs. I estimate that approximately 49 to 62 percent of executive agreements are UEAs. Congress has clearly acquiesced in the use of SEAs and CEAs, but it is debatable whether Congress has acquiesced in the executive branch practice of making thousands of international agreements that do not fit neatly into either category.
Evidence compiled by Professors Oona Hathaway, Curtis Bradley, and Jack Goldsmith, which they summarized in an important law review article, reinforces the point that the purported distinction between SEAs and ex ante CEAs is very problematic.Footnote 14 The Case Act, originally enacted in 1972, requires the secretary of state to transmit to Congress all international agreements concluded by the United States, other than Article II treaties, within sixty days after entry into force. Regulations subsequently adopted by the State Department specify that “the Department will transmit to Congress a background statement to accompany each agreement reported under the [Case] Act, and that this background statement will include ‘a precise citation of legal authority.’”Footnote 15 Professors Hathaway, Bradley, and Goldsmith used the Freedom of Information Act to obtain 5689 “cover memos” for international agreements that the State Department transmitted to Congress to accompany agreements reported under the Case Act between 1989 and 2017. They present a detailed analysis of those cover memos to help clarify the sources of the president’s authority to enter into executive agreements.Footnote 16
Notably, 3,915 of the cover memos (69 percent of all cover memos) cite the U.S. Constitution as a basis for the president’s legal authority to make a binding international agreement. However, only 374 cover memos (fewer than 7 percent) rely exclusively on the Constitution as a source of legal authority.Footnote 17 Subtracting 374 from 3,915, it follows that 3,541 cover memos (about 62 percent) cite both Article II and one or more statutes as a basis for the president’s authority to conclude an executive agreement. For those agreements, one can fairly assume that the executive branch lawyers who drafted the cover memos were uncomfortable relying exclusively on Article II to classify the agreement as an SEA, or relying exclusively on a statute to classify the agreement as an ex ante CEA. Thus, by this metric, approximately 62 percent of all executive agreements are UEAs—that is, executive agreements that are not easily classified as either SEAs or ex ante CEAs.
Professors Hathaway, Bradley, and Goldsmith coded the statutes cited in the cover memos on a scale from 1 to 5. Statutes coded as “1” grant the president “express authorization to conclude [international] agreements.” Statutes coded as “5” provide “no arguable delegation of agreement-making authority.”Footnote 18 According to their analysis, a shocking 37.5 percent of the authorities cited in the cover memos as a statutory basis for entering into an ex ante CEA “provide no arguable delegation of agreement-making authority.”Footnote 19 Only about 44 percent of the cover memos “rely on a legal authority with the highest code of 1 (express authority to enter an agreement).”Footnote 20 In sum, in more than half of the cases where the president claims ex ante authority to enter into an international agreement, it is debatable whether the statute or statutes cited in the cover memos actually grant such authority.
Let us assume that the 44 percent of cover memos coded as “1” are properly classified as ex ante CEAs, and the 7 percent of cover memos that rely exclusively on the Constitution as a source of authority are properly classified as SEAs. Adding those two figures together suggests that about 51 percent of all executive agreements fit neatly into either the SEA or ex ante CEA category. Thus, by this metric, approximately 49 percent of all executive agreements are UEAs. (The very small number of ex post CEAs does not materially alter this quantitative analysis.) To reiterate, the UEA category includes many agreements that are arguably beyond the scope of the president’s independent Article II authority, and that arguably exceed the scope of delegated authority under relevant statutes. In my judgment, Professor Bradley has not demonstrated that Congress has acquiesced in the executive branch practice of making thousands of unidentified executive agreements that do not fit squarely into either the SEA or ex ante CEA category.
Moreover, the history of repeated congressional efforts to constrain the president’s use of executive agreements to bypass the Article II treaty process belies the claim that Congress has acquiesced in the practice of making unidentified executive agreements. In February 1952, Senator John Bricker proposed a constitutional amendment that said: “Executive agreements shall not be made in lieu of treaties.”Footnote 21 After two years of heated Senate debates, Senator Walter George introduced a substitute amendment, which stated: “An international agreement other than a treaty shall become effective as internal law in the United States only by an act of Congress.”Footnote 22 The George substitute would have overruled Belmont and Pink by making clear that executive agreements do not have domestic legal effects, absent congressional legislation. “The Eisenhower administration strenuously opposed limits on the domestic effects of sole executive agreements. On this point, though, the president almost lost.”Footnote 23 The Senate voted 60–31 in favor of the George substitute, just one vote short of the two-thirds majority needed for a constitutional amendment.Footnote 24
Since the failed effort to enact a constitutional amendment in the 1950s, Congress has engaged in repeated, episodic efforts to constrain the president’s use of executive agreements. Congress first imposed reporting requirements on the executive branch in the Case Act, enacted in 1972.Footnote 25 It amended the Case Act in 1977, 1978, 1994, 2004, 2020, and most recently in 2022. Before the Supreme Court’s 1983 decision in INS v. Chadha,Footnote 26 the Case Act included a legislative veto provision that authorized Congress to review the text of executive agreements, prior to entry into force, “and decide whether to disapprove [a particular agreement] through one-house or two-house vetoes.”Footnote 27 (Chadha held that such legislative veto provisions are unconstitutional.) Congress first imposed publication requirements on the executive branch in 1950.Footnote 28 It amended those publication requirements in 1994, 2004, and 2022. (The legislation enacted in 2022 covers both publication and reporting requirements.Footnote 29 Passage of that legislation is at least partially attributable to advocacy efforts by Professors Hathaway, Bradley, and Goldsmith following publication of their co-authored articles.) Detailed analysis of those amendments is beyond the scope of this essay. Suffice it to say that repeated congressional efforts to strengthen the publication and reporting requirements cast doubt on the claim that Congress has acquiesced in the president’s use of executive agreements as a substitute for Article II treaties, except insofar as particular agreements are expressly authorized by statutes or fit squarely within the scope of the president’s independent Article II authority.
Conclusion
Professor Bradley’s book on historical gloss makes an important contribution to the existing scholarly literature on constitutional foreign affairs law. His analysis demonstrates persuasively that historical gloss is an important modality of constitutional argument, which is distinct from Bobbitt’s canonical list of six other modalities. However, the preceding analysis of executive agreements suggests that at least some gloss-based constitutional understandings will likely continue to be sources of ongoing political contestation. In an era of sharp partisan polarization, when seemingly settled judicial precedents are being overturned on a regular basis, it would be unrealistic to expect gloss-based understandings to provide a greater degree of stability in constitutional construction.
Open access
