ARTICLE CONTENTS INTRODUCTION I. U.S. INTERNATIONAL LAWMAKING AND THE DEBATE OVER INTERCHANGEABILITY A. The Interchangeability Debate B. How International Law Is Made in the United States 1. Legislative and Regulatory Guidelines 2. Article II Treaties vs. Congressional-Executive Agreements: The Empirical Evidence C. U.S. Practice in Comparative Perspective II. A BRIEF HISTORY OF INTERNATIONAL LAWMAKING IN THE UNITED STATES A. The Treaty Clause: A Compromise To Save the Union 1. The Senate as a "Council of Advice" to the President 2. Protecting Regional Interests: The Mississippi River and the Origins of the Treaty Clause B. The Rise of the Congressional-Executive Agreement 1. The First Hundred Years: A Modest Tool 2. The Second Hundred Years: Reversal of Fortunes C. Divergent Paths: The Bricker Amendment Controversy and Fast Track III. THE CASE FOR (ALMOST) ABANDONING THE TREATY CLAUSE A. Stronger Democratic Legitimacy B. A Less Cumbersome and Politically Vulnerable Process C. More Reliable Commitments 1. Enforcement of Treaties and Congressional-Executive Agreements 2. Withdrawal from Treaties and Congressional-Executive Agreements IV. THE END OF THE TREATY AND ITS CONSEQUENCES A. Constitutional Consequences B. The International Legal Consequences C. Treaties' End CONCLUSION APPENDIX A. Data Sources for Treaties and Executive Agreements B. Constitutional Requirements for Domestic and International Lawmaking
In the fall of 2007, Senate hearings finally commenced on the United Nations Convention on the Law of the Sea, a treaty that has been languishing in the Senate since 1994, when Bill Clinton was still a fresh face in the White House. (1) Submitted to the Senate under the Treaty Clause of the Constitution, (2) the treaty must gain the consent of two-thirds of the Senate in order to become law for the United States--a hurdle it has been unable to clear for over a decade because of a small but determined opposition. Meanwhile, free trade agreements between the United States and Peru, Colombia, and Panama are also up for approval. But these agreements are proceeding not through the Treaty Clause but as "congressional-executive agreements," subject to approval by a majority of both houses of Congress. Signed in 2006, one has already been approved by Congress and at least one more is likely to be approved later this year. (3)
As these examples show, the process for making binding international agreements in the United States today proceeds along two separate but parallel tracks: one that excludes the House of Representatives and another that includes it, one that requires a supermajority vote in the Senate and another that does not, one that is expressly laid out in the Constitution and one that is not. (4) I refer to both of these methods of making international commitments as "international lawmaking" to emphasize the dependence of international law on individual countries' decisions to commit to it. International law may be negotiated by states in New York or Geneva or Montreal, but it is not made at the negotiating table. It is made by countries when they agree as a matter of law to a binding international commitment. For it is the act of consent by each country that transforms an international agreement from a piece of paper devoid of any legal force into law that binds. (5)
Of the two methods for malting international law in the United States, the Treaty Clause-which requires a two-thirds vote in the Senate and bypasses the House of Representatives- is the better known of the two; it is principally used to conclude agreements on extradition, taxation, and investment and commercial matters. But an increasingly common path is the congressional-executive agreement, now used in virtually every area of international law. Each year, hundreds of congressional-executive agreements on a wide range of international legal topics are enacted by simple majorities in the House and Senate and signed into law by the President, outside the traditional Treaty Clause process. (Executive agreements entered into by the President alone-often called sole executive agreements-are also on the rise and involve no formal congressional involvement at all. (6))
It is puzzling that two distinct methods of lawmaking operate side-by-side within a single nation--all the more so because virtually no other country deals with international law as we do. Most other countries make international law in the same way they make domestic law-a norm followed by one of our two methods (congressional-executive agreements) but not the other (the Treaty Clause). Because the Treaty Clause requires that all but thirty-three members of the Senate assent to a treaty and includes no provision for participation by members of the House, it surely makes a substantial difference which of these two methods is used. For this reason alone, it would be natural to expect that there are compelling, consistent reasons why each method is used in particular areas or instances.
Yet that is not the case. Although there are patterns to the current practice of using one type of agreement or another, those patterns have no identifiable rational basis. For example, most free trade agreements are concluded through congressional-executive agreements. By contrast, agreements on investment and commercial matters-issues no less critical to the smooth operation of the global economy-are concluded through both treaties and congressional- executive agreements. The Law of the Sea Convention mentioned at the outset was brought to the Senate under the Treaty Clause. But most other fisheries and maritime agreements are concluded through congressional-executive agreements. Human rights agreements are concluded as treaties. Meanwhile, the vast majority of education, health, and debt-restructuring agreements with developing countries--issues that can be just as important to human dignity--are concluded as congressional-executive agreements. Compared with agreements authorized as congressional-executive agreements, a higher share of agreements considered under the Treaty Clause are multilateral. Nonetheless, the vast majority of multilateral agreements are concluded through congressional-executive agreements.
There is, I argue, no persuasive explanation for these differences based on the subject matter, form, topic, or any other substantive basis. The explanation for these differences lies not in reason, but in history-a history that it is now time to leave behind. Rooted in now-irrelevant (and discredited) concerns of slaveholding states, overtaken by actual political practice almost from the Constitution's beginning, the Treaty Clause was the product of circumstances that have little continuing relevance.
The current bifurcated system took its shape over the course of the twentieth century. The United States gradually abandoned the mercantilist, protectionist trade policy that it had pursued since the Civil War in favor of a policy built on reciprocal trade agreements with foreign states. The legal innovation that enabled this transformation subsequently expanded to include almost every area of international law-an expansion fueled by the perceived cumbersomeness of the Treaty Clause alongside the desire and need for the country to engage more fully in the international sphere. Meanwhile, opposition to human rights agreements motivated significant opposition to treaties in the second half of the century. In the 1950s, a series of proposed amendments to the Constitution (generally referred to collectively as "the Bricker Amendment" after the chief sponsor in the Senate) aimed to prevent the United States from entering international human rights agreements that some feared would be used to challenge segregation and Jim Crow. The controversy ended in a "compromise" in which the amendment was defeated at the cost of future human rights agreements, which would henceforth be concluded only as treaties that had been rendered almost entirely unenforceable through reservations, understandings, and declarations. (7) All of the rest of international law was haphazardly carved up between these two tracks-with some areas assigned to the Treaty Clause route, others to the congressional-executive agreement, and many uncomfortably straddling the two.
Paying fealty to this history by requiring that treaties continue to be used in certain historically contingent areas of international law comes at a substantial continuing cost: compared to congressional-executive agreements, treaties have weaker democratic legitimacy, are more cumbersome and politically vulnerable, and create less reliable legal commitments. The final failure is particularly worrisome, since the central purpose of international lawmaking is to create reliable commitments between states.
This Article makes the case for a new direction: nearly everything that is done through the Treaty Clause can and should be done through congressional-executive agreements approved by both houses of Congress. The congressional-executive agreement includes the House of Representatives in the lawmaking process, is less subject than is a treaty to stonewalling by an extreme minority, and rarely requires the passage of separate implementing legislation to enter into effect. Moreover, the agreement is often easier to enforce and can be subject to more stringent rules regarding unilateral withdrawal, thus allowing the United States to make stronger and more consistent international commitments. A congressional-executive agreement might seem to lack the "'dignity' of a treaty." (8) But in fact a congressional-executive agreement that is expressly approved by Congress is more legitimate and more reliable than a treaty, and it can and should be used for even the most important international commitments. (9)
In laying out the case for "treaties' end," I examine U.S. international lawmaking through empirical, comparative, historical, and policy lenses. I begin in Part I with a broad empirical assessment of the international lawmaking practice of the United States during the last two decades of the twentieth century. What I find has implications for the longstanding debate over the "interchangeability" of treaties and congressional-executive agreements-calling into question the empirical claims of many of those on both sides of the debate.
I next consider the treaty-making process in cross-national comparative perspective. The United States, it turns out, is extraordinarily unusual. The process for making international law that is outlined in the U.S. Constitution is close to unique. Together with the evidence about recent U.S. practice, these findings pose a puzzle: why does the United States have such an anomalous system for making international law?
In Part II, I develop a historical account that provides some answers. It traces the current odd and unsatisfactory international lawmaking arrangement back to the Founding. The current system of international lawmaking in the United States rests, I show, on rules and patterns of practice developed in response to specific contingent events-events that for the most part have little or no continuing significance.
In Part III, I show that the Treaty Clause, besides having no strong legal or historical claim for priority today, is demonstrably inferior as a matter of U.S. public policy to congressional-executive agreements expressly approved by both houses of Congress on nearly all crucial dimensions: ease of use, democratic legitimacy, and strength of the international legal commitments that are created.
I conclude in Part IV by presenting a vision for the future of international lawmaking in the United States that charts a course toward ending the Treaty Clause for all but a handful of international agreements. By gradually replacing most Article II treaties with congressional-executive agreements, policy makers can make America's domestic engagement with international law more sensible, more effective, and more democratic.
I. U.S. INTERNATIONAL LAWMAKING AND THE DEBATE OVER INTERCHANGEABILITY
Since at least the 1940s, most scholars of U.S. international lawmaking have fallen into two broad camps-opposite sides in what is often called "the interchangeability debate." On one side stand those who argue that congressional-executive agreements and Article II treaties are and should be treated as wholly interchangeable. On the other are those who say that they are not and should not be: treaties and congressional-executive agreements have appropriately separate spheres that can be described and justified with legal and analytical reasons.
A notable feature of this debate is that most of the arguments rest upon a remarkably thin understanding of the current international lawmaking practice in the United States-and yet many on both sides make strong (and conflicting) claims that their normative views are reflected in actual practice. Thus, after outlining the two sides in the interchangeability controversy, I begin to fill this gap by undertaking an examination of current practice. Doing so is an important step toward settling the debate over interchangeability. Examining the empirical record makes it possible to determine which, if any, of the descriptive claims is accurate--and shows that neither side gets the story right. Moreover, examining current practice is essential for assessing that practice, and ultimately reforming it, as I argue we should.
My examination proceeds in two stages. I begin with an analysis of how international law is currently made in the United States. The legal and regulatory framework that applies to international lawmaking is thin and opaque. Aimed primarily at preventing the evasion of congressional oversight over international agreements altogether, the framework provides scant guidance with regard to what should happen within Congress. The result is a practice of international lawmaking that is not consistent with either side of the interchangeability debate. Treaties and congressional-executive agreements are not fully interchangeable, for there are many areas of law in which one instrument or the other is exclusively or almost exclusively used. At the same time, the use of treaties and congressional-executive agreements does not conform to the predictions of those who argue that the two lawmaking instruments operate in separate spheres, for there are many areas of law where the two are used interchangeably. These findings, fully consistent with no existing theory, are in themselves deeply puzzling.
They are all the more so, I show, when viewed in comparative context--the second stage of my examination. The United States, it turns out, is extraordinarily unusual in the way it makes much of its international law. Indeed, by examining the international (and domestic) lawmaking procedures of nearly every nation in the world, I am able to say confidently that the U.S. Treaty Clause creates a process for making international law that has almost no parallel abroad. This distinctive process results, moreover, in predictable yet haphazard divisions that have important consequences for the United States' ability to engage in international cooperation.
A. The Interchangeability Debate
The modern debate over the interchangeability of congressional-executive agreements and Article II treaties dates to the 1940s, when executive agreements began to rise in prominence. On one side of this vigorous exchange are those who argue that treaties and executive agreements are wholly interchangeable. On the other are those who argue that treaties and congressional-executive agreements are instead exclusive instruments, and that the Constitution requires that each be used in limited circumstances.
Since at least the 1940s, the weight of scholarly opinion has rested with the first view. Wallace McClure wrote in 1941 that "executive agreements and treaties have been used interchangeably to accomplish seemingly identical purposes" and hence "there is, prima facie, no reason to deny the existence of constitutional authorization for the use of executive agreements relating to whatever subjects may be dealt with by the treaty-malting power." (10) Shortly thereafter Edward Corwin concluded that executive agreements through incremental "constitutional development" had come to serve many of the same purposes as treaties. (11) And near the close of World War II, Myres McDougal and Asher Lans wrote that "our constitutional law today makes available two parallel and completely inter-changeable procedures, wholly applicable to the same subject matters and of identical domestic and international legal consequences, for the consummation of intergovernmental agreements." (12)
This view has continued to hold sway among much of the scholarly community. Louis Henkin, writing in the mid-1990s, concluded that "it is now widely accepted that the Congressional-Executive agreement is available for wide use, even general use, and is a complete alternative to a treaty." (13) The 1987 Restatement (Third) of Foreign Relations Law endorsed interchangeability, noting that, "[a]t one time it was argued that some agreements can be made only as treaties.... Scholarly opinion has rejected that view." (14) More recently, Bruce Ackerman and David Golove offered a powerful defense of the interchangeability position, albeit a modified one. They argued that the concept of interchangeability dates not to the founding of the nation, but to the New Deal era. Over the course of the late 1930s and early 1940s, Congress, the President, the courts, and legal scholars together developed a new constitutional consensus that permitted the President to submit international agreements to both houses of Congress for approval in lieu of the Article II process, is By the time a majority of both houses of Congress approved the 1945 Bretton-Woods Agreement that would create the foundations of a new world economic order, the congressional-executive agreement had ascended to the core of U.S. international lawmaking, where it remains today. (16)
On the other side of the debate stands a smaller but outspoken group of critics who have repeatedly challenged the eclipse of the Treaty Clause. (17) In the 1940s, these critics-most prominently Edwin Borchard-argued that the Constitution required certain international agreements to be made by treaty alone and that any effort to change this requirement was ill-advised. (18) Though Borchard's pro-status quo view was in the scholarly minority at the time, it appeared to triumph in the halls of Congress, with the failure of a proposed Amendment to the Constitution in 1945 that would have ended the two-thirds provision in favor of a majority vote in both houses of Congress. (19) The reason for the failure, however, likely did not please the Treaty Clause purists: the Amendment failed to gain support at least in part because many in Congress concluded that they could achieve the same result without it, by substituting congressional-executive agreements for Article II treaties. (20)
In recent years, at least two separate anti-interchangeability positions have arisen. The first, put forward by Laurence Tribe, is a modern version of the purist position that international agreements must be made through the Treaty Clause. In the mid-1990s, Tribe launched a broad-scale and hard-hitting attack on Ackerman and Golove's method of analysis and their conclusion that the "congressional-executive agreement [is] an all-purpose substitute for the treaty." (21) Using congressional-executive agreements as if they were fully interchangeable with treaties, he argued, is inconsistent with the text, structure, and history of the Constitution and would allow Congress to exceed the powers expressly granted to it. (22) Instead, Tribe advocated that the Treaty Clause be treated as the exclusive method for treaty approval. Consequently, U.S. participation in the North American Free Trade Agreement (NAFTA), the World Trade Organization, the Bretton Woods Agreements, and many other agreements concluded by congressional-executive agreement should have been submitted to the Senate under the Treaty Clause. (23)
A second anti-interchangeability theory might be called the "separate spheres" approach. One version, advanced by John Yoo, argues that Article I and Article II agreements are both constitutional but have their own mutually exclusive zones of authority. (24) Congressional-executive agreements, in Yoo's words, "must be used to approve international agreements that regulate matters within Congress's Article I powers" to ensure "that the same public lawmaking process will apply to the same subjects, regardless of whether an international agreement is involved or not." (25) The Article II treaty process is required, on the other hand, "if the nation seeks to make agreements outside of Congress's competence or bind itself in areas where both President and Congress exercise competing, overlapping powers." (26) In Yoo's telling, not only is this division of labor constitutionally required, but it almost perfectly comports with the practices of the nation over the past fifty years--"whatever is, is right." (27)
A second version of the separate spheres approach is associated with Peter Spiro. (28) In an article published almost simultaneously with Yoo's, Spiro sought, like Yoo, to make sense of the "persistent patterns of instrument choice." (29) He did so, however, not primarily through textual analysis of the Constitution but by examining the acceptance, contestedness, age, and pedigree of existing practices. (30) In Spiro's account, the "ongoing interplay among the branches gives rise to an accreted refinement of norms, in much the same way as judicial decisions do in other areas of the law." (31) This historical interplay, he argued, had given rise to separate domains for the Treaty Clause and congressional-executive agreements that were logical, embedded, and worthy of respect.
B. How International Law Is Made in the United States
For all the broad claims made in the interchangeability debate, surprisingly little is known about current international lawmaking practice in the United States. My aim in this Section, therefore, is to describe how international lawmaking actually takes place in the United States. The following Section compares these practices with those of other nations.
As will become clear, the evidence is consistent with neither side of the interchangeability argument. Treaties and congressional-executive agreements are not used as perfect substitutes for one another, as interchangeability advocates would have it. Yet neither do the two instruments of international lawmaking have well-defined, exclusive, and defensible areas of authority, as critics of the interchangeability position contend. Although it is possible to detect patterns in the use of one instrument over another, these patterns have little or no identifiable rational basis.
Far from resolving the debate, then, these findings present a puzzle of their own: why does the United States have such a confused system for making international law?
1. Legislative and Regulatory Guidelines
Commentators who favor interchangeability as a normative ideal almost universally argue that interchangeability is reflected in practice, with treaties and congressional-executive agreements used almost entirely interchangeably. Those, on the other hand, who oppose interchangeability argue that Article II and congressional-executive agreements are used differently-usually in ways that coincide with the author's normative framework, whatever that might be. A deeper examination of the empirical evidence over last two decades of the twentieth century proves both claims wrong. Treaties and congressional-executive agreements are not used interchangeably. But neither are the differences between them driven by any reasoned analytical differences. In short, the decision to pursue an agreement through one or the other of the two major international lawmaking processes is driven principally by historical happenstance and political considerations.
The decision to conclude an international agreement as a treaty, a congressional-executive agreement, or a sole executive agreement is made in the first instance by the U.S. Department of State. The Department is guided in its decision by rules and regulations first enacted in the 1950s, now known as the Circular 175 Procedure. (32)
The procedure requires that a request for authorization to negotiate or sign a treaty or other international agreement must take the form of an action memorandum that includes, among other things, a discussion of the basis for the type of agreement recommended. Eight factors are to be taken into consideration:
The extent to which the agreement involves commitments or risks affecting the nation as a whole; Whether the agreement is intended to affect state laws; Whether the agreement can be given effect without the enactment of subsequent legislation by the Congress; Past U.S. practice as to similar agreements; The preference of Congress as to a particular type of agreement; The degree of formality desired for an agreement; The proposed duration of the agreement, the need for prompt conclusion of an agreement, and the desirability of concluding a routine or short-term agreement; and The general international practice as to similar agreements. (33)
The Circular 175 procedure also requires the Office of the Legal Adviser to provide a memorandum prior to negotiating an international agreement that discusses and justifies the use of the Article II treaty process or the use of an executive agreement and, inter alia, an "analysis of the Constitutional powers relied upon." (34)
This set of regulations, which at first appears comprehensive, leaves a great deal of room for the exercise of discretion. This should come as no surprise. The regulations were not crafted to prevent evasion of the Treaty Clause through congressional-executive agreements. They aimed instead at ensuring congressional involvement in the making of international agreements. In particular, members of Congress sought to prevent the President from making excessive use of executive agreements concluded with little or no oversight by Congress (sometimes called sole executive agreements). (35)
While the Circular 175 and accompanying regulations provide some direction about the situations in which sole executive agreements are not appropriate, they give relatively little guidance regarding the choice between a treaty and a congressionally authorized executive agreement. For example, the first factor--how extensive the commitment or risks affecting the nation--was likely intended to discourage the use of sole executive agreements in situations with extensive commitments or risks. But it is far from clear how this factor would affect the choice between an Article II treaty and a congressional-executive agreement. Clearly, an agreement that involves commitment or risks affecting the nation as a whole should have the approval of more than one branch of government. Less clear is whether such an agreement should be subject to a supermajority vote in the Senate, as required by the Article II process, or instead to a majority vote in both houses of Congress. (36) That is true of many of the eight factors, which were almost certainly not intended to guide the choice between the two different forms of congressional approval. (37)
Only two of the listed eight factors have any significant bearing on the choice between the Article II and congressional-executive agreements processes. Past U.S. practice as to similar agreements (the fourth factor above) can offer guidance regarding the use of the Treaty Clause as opposed to a joint resolution. As elaborated below, patterns of practice have evolved around particular subject areas of international lawmaking-practices that are increasingly entrenched. In addition, the preference of Congress as to a particular type of agreement (the fifth factor) can also influence the choice between the two types of agreements. (38) Where the Senate makes clear its preference that an agreement proceed through the Article II process, that would likely influence the executive's choice of instrument--not simply because of deference to the Senate's constitutional authority but also (perhaps primarily) because a majority of the Senate can defeat a joint resolution if it feels its unique constitutional authority is threatened. Nonetheless, this power is a limited one, for a majority of the Senate may approve a congressional-executive agreement even over the opposition of a significant minority. (39)
The regulations have thus done little to illuminate and guide the practice of choosing between international lawmaking processes--much less encourage transparent and principled distinctions. As the next Subsection shows, recent international lawmaking practice in the United States shows many clear signs of this indistinct guidance.
2. Article II Treaties vs. Congressional-Executive Agreements: The Empirical Evidence
To date, relatively little attention has been paid to when treaties and congressional-executive agreements are actually used in U.S. international lawmaking. Early work on the topic simply noted that that there are distinct uses of the two types of instruments. (40) Recent work often stops at the observation that trade is an area in which congressional-executive agreements are prominent, whereas human rights and arms control are areas in which treaties are more common. (41) The most comprehensive examination of the empirical evidence, John Yoo's examination of the treaties in force in 2000, (42) only focuses on a few areas of law and does not say much about the United States' current treaty practice. (43) Peter Spiro's similar examination stops at the observation that of the three major areas of international lawmaking, the treaty form is used in two (arms control agreements and mutual security pacts and human rights conventions), and congressional-executive agreements in the third (trade agreements). Yet this leaves the vast bulk of international agreements unaccounted for.
Part of the reason for the relative absence of empirical work is that there is no single comprehensive database available that delineates sole executive agreements, congressional-executive agreements, and Article II treaties. Nonetheless, it is possible to gain a reasonably complete picture of international lawmaking in the United States by cross-referencing a variety of databases. (44) That is what I do here in an effort to provide insight into U.S. international lawmaking over the course of the two decades from 1980 to 2000. This examination reveals two trends in lawmaking. (45) First, the line between the different types of international lawmaking--particularly the various kinds of executive agreements--is not nearly as distinct as usually assumed. Second, treaties and congressional-executive agreements are used in ways that do not conform to any of the existing academic accounts.
Descriptions of international lawmaking in the United States generally break international agreements into two categories: executive agreements and Article II treaties. But there are, in fact, three very different kinds of executive agreements that differ significantly in the amount of interbranch cooperation they require. First are congressional-executive agreements. These are agreements concluded by the President and either authorized in advance or approved after the fact through the same process used for ordinary federal legislation. The second and third types of executive agreements are both commonly referred to as "sole executive agreements." One is concluded pursuant to a treaty obligation and the other is concluded solely on the President's own constitutional authority. (46) Unlike ex post congressional-executive agreements, sole executive agreements require nothing more than congressional inaction to take effect. (47)
The discussion that follows focuses broadly on the first type of executive agreements just mentioned: congressional-executive agreements. It is worth noting, however, that this broad category actually contains at least two subtypes that differ in the degree of control retained by Congress and hence in the amount of interbranch cooperation that they require. First are congressional-executive agreements authorized in advance by legislation ("ex ante congressional-executive agreements" or "congressionally authorized executive agreements") involve relatively little interbranch cooperation. Indeed, in many cases, they do not involve a more significant congressional role than do sole executive agreements based on a prior treaty arrangement. This is particularly true of the numerous agreements based on broad authority granted well in advance, which make up the largest group of congressional-executive agreements. (48) Second are congressional-executive agreements that require approval by Congress only after the agreement is negotiated ("ex post congressional-executive agreements") perforce involve deeper interbranch cooperation. However, such agreements are much less common than their less restrictive counterparts. Although an accurate count is almost impossible, during the twenty-year period under examination here, it appears there were a small number, including agreements on fisheries, trade, atomic energy, investment, education, and the environment. (49)
In the remainder of this Subsection, I put these distinctions to one side and focus on congressional-executive agreements taken as whole, excluding, when possible, sole executive agreements. When we compare the substantial body of congressional-executive agreements authorized in some form by Congress to Article II treaties, we find that both sides of the interchangeability argument fail. (50) To make this comparison, it is necessary to begin by examining the full set of Article II treaties. During the last two decades of the twentieth century, such treaties were used in a relatively narrow set of areas. Table 1 groups all of the treaties entered by the United States during this period by subject matter. (51) The two most prevalent types of treaties are extradition treaties, which make up fully twenty-seven percent of treaties, and taxation treaties, which make up nineteen percent of treaties. Next are treaties on investment (eleven percent), commercial matters (seven percent), fisheries and wildlife (seven percent), arms control (four percent), maritime matters (four percent), shipping and marine pollution (four percent), and the environment (two percent). A variety of other areas-including aviation, consular relations, maritime matters, telecommunications, international law and organization, human rights, labor, nuclear safety, intellectual property/copyrights, dispute settlement and arbitration, and legal documents--have also seen multiple treaties, though not in large numbers.
Congressional-executive agreements, by contrast, have been used in a wide variety of areas--over one hundred during the 1980s and 1990s, according to the most comprehensive available data. Table 2 lists the twenty most common subject areas. Executive agreements do not, except on rare occasions, identify the source of the authority under which they are concluded--whether under the President's constitutional power or an ex ante or ex post congressional authorization. As a consequence, separating executive agreements that are congressionally authorized from those that are not requires a painstaking search for authorizing legislation. To determine whether an agreement is a congressional-executive agreement, it is necessary to search the Statutes at Large prior to the date the agreement went into effect for terms related to that subject area. (52) Then it is necessary to read each statute to determine whether it actually authorizes the relevant international agreements. Using this method, a research assistant and I found authorizing legislation for agreements in every subject area specifically listed in Table 2 (though not for those in the "Other" category). We were unable to find any relevant authorizing legislation for agreements in the following areas: aviation, finance, taxation, telecommunication, scientific cooperation, and arms limitation. As a result, we excluded them from the table along with those agreements that are most obviously sole executive agreements. Among the subject areas included in the table, agreements regarding defense are the most prevalent, constituting just over thirteen percent of all agreements entered into during the decade. Other areas in which congressional-executive agreements are commonly used include trade (eight percent), debts (eight percent), postal matters (seven percent), agriculture (six percent), atomic energy (four percent), and economic cooperation (four percent).
There are a few areas of law in which the Article II process was used exclusively during the 1980s and 1990s. Foremost was extradition, about which there were 103 treaties (twenty-seven percent of all treaties during this period) and apparently no executive agreements of any kind. (54) The only other areas in which the Article II process was used exclusively were human rights (five Article II treaties) and dispute settlement (two Article II treaties). (55) In addition, there are several areas of law in which all significant international agreements were concluded through the Article II process, and any congressional-executive agreements appear to be entered pursuant to obligations under a treaty obligation or under the sole authority of the President (where there are, in other words, no true congressional-executive agreements). This includes arms control (with fifteen Article II treaties devoted to the topic (56)), aviation (with seven Article II treaties (57)), the environment (eight Article II treaties), (58) labor (five Article II treaties), (59) consular relations (six Article II treaties), (60) taxation (seventy-three Article II treaties), (61) and telecommunications (six Article II treaties). (62)
In many areas, a significant number of agreements were concluded as Article II treaties, but congressional-executive agreements appear to be important as well-these are areas where the interchangeability thesis comes closest to being accurate. These include investment (forty-three Article II treaties and seventy-seven congressional-executive agreements), (63) maritime matters (fifty-three Article II treaties and sixty-eight congressional-executive agreements), (64) education (one Article II treaty and sixty-seven congressional executive agreements), (65) nuclear safety and technology (five Article II treaties and nineteen congressional-executive agreements), (66) and judicial and criminal assistance (more than twenty Article II treaties and congressional-executive agreements). (67) Trade, usually thought of as an area in which congressional-executive agreements dominate, is also an area of shared authority: over two hundred congressional-executive agreements on trade were concluded during the 1980s and 1990s, (68) but there were significant numbers of Article II treaties as well. (69)
Finally, there are numerous areas of international law in which agreements are concluded exclusively or almost exclusively through congressional-executive agreements. Agreements on defense matters are the most numerous, with 358 such agreements during the 1980s and 1990s. (70) These agreements include status of forces agreements, (71) training agreements, (72) and mutual logistical support, (73) among others. Debt agreements were also exclusively concluded through executive agreements, most of which appear to have been congressionally authorized in advance. (74) In the modern era, postal agreements are all concluded through congressional-executive agreements. (75) Agreements on agriculture are almost exclusively concluded through congressional-executive agreements, with 167 such agreements and only one Article II treaty concluded during the 1980s and 1990s (and the one treaty was concluded in 1980). (76) The same is true of agreements on atomic energy, where there were 127 agreements and one Article II treaty. (77) Congressional-executive agreements were exclusively used in the areas of economic cooperation, where there were 115 agreements, (78) and employment. (79) These are just a few of the areas covered by executive agreements. Indeed, in close to one hundred different areas of law, the United States enters agreements exclusively by means of executive agreements. (80)
This summary of the empirical evidence calls into question the claims of both sides of the interchangeability debate. Scholars who argue that congressionally authorized executive agreements and treaties are (and ought to be) fully interchangeable fail to accurately describe both the past and present practices of the United States. Though there are some areas of law in which treaties and congressionally authorized executive agreements are today used interchangeably, there are also significant areas that are dominated almost entirely by one process or the other. By ignoring the distinct uses to which the two different processes for making international law are put, those who favor interchangeability undermine confidence in their accounts. And in seeing the world as they wish it to be, they fail to recognize that the continuing use of the Treaty Clause has had a disproportionately large effect on U.S. participation in some areas of international law-including human rights- while leaving other areas entirely unaffected.
Yet those who argue that the two processes are not interchangeable (and ought not to be) also miss important parts of the story. These scholars correctly note that the Treaty Clause and congressional-executive agreements are not treated as fully interchangeable. They err, however, in providing an incomplete picture of current practices in the United States. They fail to acknowledge, for instance, that the instruments are both used in several areas of law. They also attempt to shoehorn the patterns of practice that they detect into reasoned theories of constitutional law that simply do not fit the facts. And finally, they fail analytically to provide a coherent normative account that justifies the different uses to which the two processes are put.
For example, Yoo's claim that congressionally authorized executive agreements are used exclusively for agreements that fall within Congress's Article I powers whereas treaties are used for agreements that extend beyond Article I is contradicted by the evidence. There is little evidence that the two instruments are used exclusively in certain areas of law, much less in the constitutionally guided manner that Yoo suggests. Quite the contrary: in many areas of international law-including investment, maritime matters, education, nuclear safety and technology, judicial and criminal assistance, and trade--Article II treaties and congressional-executive agreements are used side-by-side. Moreover, areas of law in which Article II treaties are used extensively, including human rights, dispute resolution, arms control, aviation, the environment, labor, consular relations, taxation, and telecommunications, almost never extend beyond Congress's Article I powers. If agreements on human rights, labor, and taxation were beyond Congress's Article I powers, then the Civil Rights Acts, the Labor Department, and the Internal Revenue Service would seem to be unconstitutional exercises of federal power as well.
This examination of the current international lawmaking process of the United States suggests that empirical reality does not fit the expectations of scholars on either side of the interchangeability debate. The two types of international agreements are neither treated as fully interchangeable nor used in ways that reflect relevant legal differences. Far from resolving the debate, then, the findings simply complicate the puzzle posed by U.S. international lawmaking: if neither side is right, as appears to be the case, then what explains the current international lawmaking practices of the United States? This puzzle deepens when we consider the international lawmaking practices of the United States in comparison with those of the rest of the world. (81)
C. U.S. Practice in Comparative Perspective
International law provides strikingly little guidance to states about how they ought to make international law. The 1972 Vienna Convention on the Law of Treaties, which incorporates widely accepted principles of international law, provides states with some guidance. It dictates that in order for a state to bind itself to an international agreement, it must express its consent. But how that consent is expressed or determined is left entirely to domestic law. (82) As a consequence, there is a wide variety of practices among states.
The only way to know how states make international law, then, is to look to the domestic legal rules that govern the process. Working with a team of researchers, I have taken a step in this direction by compiling a comprehensive database of the treaty-making and domestic lawmaking practices of every country in the world that had a constitution in the year 2007. It turns out that the U.S. Treaty Clause stands out as a remarkably unusual method of making international law. Only five other countries in the world-Algeria, Burundi, Iraq, Micronesia, and the Philippines-require a supermajority vote in their legislature in order for the country to ratify a treaty. (83) By contrast, most states require that international law be made through a simple or absolute majority vote in the legislature.
The United States is also one of a small handful of countries that combine two features in their Constitution-an international lawmaking process that provides for less involvement by part of the legislature in international treaty making than in domestic lawmaking and the automatic incorporation of the results of that process into domestic law. The vast majority of states provide in their constitutions for an international lawmaking process that mirrors the domestic lawmaking process. One hundred and twenty-four states currently have voting thresholds in the legislature for treaties that are the same as those for domestic laws. (84) By contrast, fifty-nine (including the United States) provide for different voting thresholds in either house of the legislature for treaties than for domestic legislation. Of these, only ten (again, including the United States) explicitly provide in their constitution for some level of automatic incorporation of international law into domestic law. (85) Those countries are listed in Table 3.
Four of the ten countries (Ecuador, Georgia, Serbia, and the Slovak Republic) have marginally higher voting standards for treaties than for domestic legislation. They require that a treaty be passed by a majority of all of the members of the legislature, rather than a simple majority of those present. In each case, the legislature is unicameral and hence no part of the legislature that is involved in domestic lawmaking is excluded from international lawmaking. (Moreover, the Slovak Republic makes only a subset of treaties explicitly self-executing. (86))
The six remaining countries-Cyprus, Ethiopia, Mexico, Slovenia, Tajikistan, and the United States- provide for less involvement by a part of the legislature in treaty making than in domestic lawmaking. Of these, Cyprus provides that treaties are supreme over ordinary legislation, but does not explicitly grant treaties independent legal force. In addition, Cyprus, Ethiopia, and Slovenia all provide for extremely limited upper house involvement in domestic legislation. (Cyprus's upper house is involved only in legislation affecting subnational communities (and, indeed, there is a special voting procedure involving both houses for treaties that involve the competence of the upper house, hence the domestic and international processes are effectively the same); Ethiopia's upper house is primarily responsible for interpreting the Constitution and for federal-regional issues and is not usually involved in the regular legislative process; and Slovenia's upper house involvement in ordinary legislation is largely limited to a veto over legislation that can be overridden by the lower house through simple re-passage.) This leaves Mexico, Tajikistan, and the United States as the only countries in the world that provide for significantly less involvement by a part of the legislature in treaty-malting than in domestic lawmaking and make the results of this process automatically part of domestic law in more than a few confined areas of law.
The United States is therefore unusual in requiring a supermajority legislative vote to approve treaties, it is in the distinct minority in excluding a part of the legislature that is usually involved in domestic lawmaking from international lawmaking, and it is among a small handful of countries that combine the latter feature with a rule that makes treaties automatically a part of domestic law. That the process for making treaties in the United States is extremely unusual does not mean, of course, that it is necessarily wrong or misguided. But it does raise questions, to which I shall return later in the Article, about the legitimacy of this method of international lawmaking. It also deepens the puzzle of U.S. international lawmaking: why is the system so unusual? Part II is devoted to answering this question. It examines how we arrived at the unusual compromise represented by the Treaty Clause.
II. A BRIEF HISTORY OF INTERNATIONAL LAWMAKING IN THE UNITED STATES
International lawmaking has changed dramatically over the more than two centuries since the country's founding. Examining this transformation helps to explain why the country has two separate methods for making international law whose distinct uses are not well defined. And it helps to explain why the United States adopted a process for making international law that is so unusual in comparative perspective.
The examination of the history of international lawmaking in the United States serves another related purpose as well. Many of those on opposing sides of the interchangeability debate argue that their normative claims are reflected in (and hence find support from) past and present practice. Hence, several interchangeability scholars argue that treaties and congressional-executive agreements not only ought to be interchangeable, but that they in fact are (and long have been) treated this way by policymakers. Similarly, many of those supporting the separate spheres approach argue not only that the two methods ought not be used interchangeably, but that each type of agreement is (and long has been) used in precisely the way that they advocate (for example, as noted earlier, Yoo argues that congressional-executive agreements are used for agreements that fall under Congress's Article I authority, and treaties are used for agreements that exceed this authority).
The blurring of the line between the normative and positive in the debate over the two tracks of international lawmaking likely stems at least in part from the natural reflex of lawyers to look to the weight of history-or precedent-to guide future practice. But there are reasons behind this reflex beyond a simple preference for continuity. Rules developed over time often have developed in response to functional needs-hence the practices of the present are forged in the furnace of history and address needs of which we may be only dimly aware. Moreover, past practices can serve as a guide (albeit an imperfect one) as to what practices are and are not permitted or prohibited by the Treaty Clause. One need not hold an originalist view of constitutional interpretation to believe that past uses and interpretation of the Constitution provide some guide as to what is and is not permitted by the text.
Yet the mere fact that current practice has been shaped through the accretion of historical precedent-as is true of the use of the Treaty Clause and congressional-executive agreements-does not in itself offer a normative justification for that practice. That a set of practices exists is not reason enough to assume that they are either functionally or legally the best practices-nor that how things are is how they must (or ought to) be. Indeed, where the reasons that gave rise to current practices have been discredited and rendered obsolete, as I shall argue is the case here, the fact that practices are as they are tells little about what they ought to be. (87)
Examining the history of international lawmaking practices in the United States and how they have developed over time reveals that they have been shaped directly in response to a set of particular historical circumstances--many of which no longer hold today. The Treaty Clause was a compromise carefully crafted to hold together the coalition of states in a single government. Two circumstances in particular shaped the Clause: first, the assumption that the Senate would serve as a council of advisors for the President, and second that the supermajority requirement would protect regional interests, particularly those of southern slaveholding states. These original goals are now obsolete.
Executive agreements, on the other hand, began as a quite modest tool, used for a relatively limited set of purposes. That all changed near the end of the nineteenth century, as the country turned to these agreements to facilitate reciprocal trade reductions with other nations. The use of executive agreements gradually expanded over the course of the century to the point that they came to far exceed treaties in scope, number, and importance. Meanwhile, the Treaty Clause once again became the center of controversy in the 1950s. In a sign of an emerging backlash against the human rights revolution-and particularly against the fear that human rights treaties would be used to challenge racial segregation-a series of amendments to the Constitution were proposed to restrict the treaty power of the federal government. This is the history that has shaped the system of international lawmaking in the United States today.
A. The Treaty Clause: A Compromise To Save the Union
The word "treaty" appears four times in the Constitution. The most important of these for the purposes of this discussion is the so-called Treaty Clause, which states that the President "shall have Power, by and with the Advice and Consent of the Senate, to make Treaties, provided two-thirds of the Senators present concur." (88)
This Clause was no mere afterthought. The Confederation that existed at the time of the Constitutional Convention had proven fundamentally incapable of observing many of its treaty obligations. (89) Especially troubling was the failure to abide by the Treaty of Peace with Great Britain, which forbade the United States from placing lawful impediments in the way of British citizens who sought to collect their prewar debts. (90) Because the United States was an unreliable treaty partner--as it was unable to guarantee that the states would observe the Confederation's treaty agreements--it had difficulty negotiating treaties with other nations. (91) Moreover, because the country was unable to live up to many of the agreements it had managed to negotiate, its treaty partners felt justified in doing the same. (92)
An important goal of the Constitutional Convention was, therefore, to strengthen the federal government's power to create enforceable treaties. Yet there was significant uncertainty about where to place the strengthened treaty power. Indeed, during the first day of discussion of the Treaty Clause, Randolph adjourned the conversation by noting that "almost every Speaker had made objections to the clause as it stood." (93) Much of the discussion of the clause at the Convention focused on the question of whether to bring the House into the process. After vigorous debate, they decided to place responsibility for concluding treaties in the hands of the President and the Senate alone.
There were two central reasons for this decision. First, it was expected that the Senate would be directly involved in negotiating treaties and would serve as the President's "council of advisors" in treaty making. Second, it was seen as a way to keep the federal government from bargaining away regional interests. Indeed, the Treaty Clause was ineluctably shaped by a particular set of events that made the southern states exceptionally wary of any process that would allow the north to bargain away their shared interests.
The remainder of this section examines more fully these two rationales for the Treaty Clause. These rationales, I argue, are entirely products of a particular time and a set of circumstances that no longer hold.
1. The Senate as a "Council of Advice" to the President
The two historians to have examined the Treaty Clause most closely--Jack Rakove and Arthur Bestor--both conclude that "[a]dvice ... was to be given at every stage of diplomacy, from the framing of policy and instructions to the final bestowal of consent." (94) The process required a manageable number of participants as well as secrecy--a role that most believed to be better entrusted to the twenty-six-member Senate than to the much larger House. (95) Not everyone shared this view. Gouverneur Morris, James Wilson, and James Madison favored the inclusion of the House in the treaty-making process. The "great obstacle they now confronted," historian Jack Rakove explains, "was the objection that the larger and more popular chamber of the legislature would not possess the requisite secrecy and efficiency to be an effective partner in negotiations, particularly on occasions where urgent matters of war and peace were on the tapis." (96) Repeatedly they argued for broader participation of the House in treaty making and repeatedly their proposals were either ignored or voted down. (97)
When the final draft of the Treaty Clause was read on September 7, Wilson submitted an amendment that would have given the House precisely the same rights as the Senate in treaty malting. (98) He argued that since treaties were "to have the operation of laws, they ought to have the sanction of laws also." (99) He continued: "The circumstances of secrecy in the business of treaties formed the only objection"--an objection he argued was "outweighed" by the argument in favor of "obtaining the Legislative sanction." (100) Roger Sherman responded that the "necessity of secrecy in the case of treaties forbade a reference of them to the whole Legislature." (101) Wilson's proposal was decisively rejected. (102)
The central justification for including the Senate and excluding the House--that the Senate, with its smaller size, could more efficiently engage in negotiations and keep those negotiations secret--quickly proved wrong. Early in his presidency, President George Washington regarded the Senate as a "council of advice" in the treaty-making process. (103) And yet eight years later, he had almost entirely ceased seeking the Senate's advice. Several events contributed to this transformation. The story is often recounted of the first and last visit by a President to the Senate chamber to consult about a treaty: President Washington went to the Senate on August 22, 1789, to consult about proposed treaties with the Southern Indians. (104) He was so frustrated with the experience that he declared it "defeats every purpose of my coming here" and never again appeared in person to discuss a treaty with the Senate. (105) By the end of his second term, President Washington had all but abandoned the process of consulting the Senate prior to opening treaty negotiations. (106) He assumed control of treaty negotiations and generally asked for Senate approval only once the agreements were finalized--a practice that largely continues to this day. (107) As Louis Henkin observed almost two centuries later, "'advice and consent' has effectively been reduced to 'consent."' (108)
The Senate also proved incapable of keeping a secret. Indeed, even the small number of appointed Senators were too many to maintain secrecy. The 1795 Jay Treaty with Great Britain addressed many issues left over from the American Revolution and was central to averting renewed war between the two nations. The terms of the treaty were leaked to a local newspaper by a Senator involved in the negotiations. (109) This event reinforced Washington's opinion that the Senate was not "a safe repository for diplomatic secrets." (110)
Even if these events had not so quickly put an end to the expectations of the Founders, it would likely be impossible today to regard the Senate as a council of advisors on treaty making. Indeed, the very qualities that the Founders believed disqualified the House of Representatives from participation in treaty negotiations--large size and popular electoral base--are today both true of the Senate. The first Senate included a comparatively modest twenty-two members, and those members, unlike representatives in the House, were at the time not subject to direct election but were instead appointed by the state legislatures. Today, the Senate has grown to one hundred members--much larger than the first House of Representatives, which had sixty-five members--and Senators are directly elected. (111)
2. Protecting Regional Interests: The Mississippi River and the Origins of the Treaty Clause
There was a second, equally important--and today, equally irrelevant--justification offered for entrusting the treaty-making power to the Senate and requiring that it approve those treaties by a two-thirds vote: the supermajority requirement in the Senate was seen as a method for preventing the federal government from concluding treaties that would disproportionately disadvantage a particular region or significant subset of states. In maintaining a role for the states through the Senate, the Treaty Clause made it possible for those states to place sole responsibility for international lawmaking in the hands of the federal government.
The Constitutional Convention sought to address a thorny dilemma: On the one hand, the Convention was motivated in significant part by a desire to strengthen the role of the federal government in international affairs, where collective action by the states was essential to the success of the nation. On the other hand, there was great fear that the strengthened national government would act in ways that disfavored and discriminated against a minority of states. To satisfy both concerns, the Convention gave the Senate shared responsibility with the President for treaty making. Madison observed in the discussion of the Treaty Clause that "the Senate represented the States alone." (112) The states would …