From Treaties to International Commitments: The Changing Landscape of Foreign Relations Law

Introduction

In his farewell address, George Washington urged that “[t]he great rule of conduct for us in regard to foreign nations is . . . to have with them as little political connection as possible.”1 This advice illustrates just how wide a gap exists between the world of the Founders and the present day. No longer a small nation struggling for respect, the United States now does not and could not manage its affairs in the absence of strong international cooperation. It has countless ongoing international commitments and continues to pursue new ones.2 During the Obama administration, the United States joined the New START treaty on arms control, the Basel III accords on international financial regulation, the Iran deal regarding nuclear nonproliferation, the Paris Agreement addressing climate change, and numerous lower-profile commitments. The Trump administration is more skeptical of international cooperation and has already announced its intent to withdraw from the Paris Agreement. Yet it too may well end up making or revising at least some important commitments.3

A striking feature of these international commitments is the diversity of legal pathways by which the United States joins them. The Treaty Clause of the Constitution empowers the president to make treaties with the advice and consent of two-thirds of the Senate.4 This is the only way to enter into international commitments that is specified in the Constitution, and yet today international commitments are routinely reached in other ways. Of the four commitments named above, only one—the New START treaty—has gone through the process set out in the Treaty Clause.5 The others have all followed different paths. Basel III is nonbinding as a matter of international law and is being implemented by administrative agencies through powers delegated to them under the Dodd-Frank Wall Street Reform and Consumer Protection Act6 and preexisting statutes.7 The Iran deal is also nonbinding as a matter of international law, and the executive branch can meet the US commitments under it by deploying previously delegated statutory authority.8 The Paris Agreement is binding under international law and took effect without any specific congressional approval, although the Obama administration intended to tie its implementation to previously delegated administrative authority.9 A fifth major agreement negotiated by the Obama administration—the Trans-Pacific Partnership (TPP)—would have required approval and implementation by congressional legislation but has since been abandoned by the Trump administration.10 Collectively, these examples illustrate that the US process for making international commitments has become multifaceted rather than unitary.

Scholars of foreign relations law typically break down US participation in international agreements into three main categories: treaties entered into pursuant to the Treaty Clause, congressional-executive agreements, and sole executive agreements.11 Congressional-executive agreements “are concluded by the president with either the advance authorization or subsequent approval of a majority of both houses of Congress.”12 Sole executive agreements “are concluded by the president alone.”13 This three-part categorization is long-standing—dating back at least to the 1920s—and has become “Lesson I of Foreign Relations Law 101.”14 Yet its usefulness is increasingly questionable. In a speech given during his tenure as State Department Legal Adviser, Professor Harold Koh criticized this framework as a “procrustean construct,” observing that international agreements often “do not fall neatly into any of these boxes.”15 This three-part categorization also takes account only of agreements that are binding under international law and thus does not cover purely nonbinding commitments, even those as important as Basel III or the Iran deal.

The emphasis on these three categories is problematic for a deeper reason, as well. It frames the process of making international commitments using the lens of constitutional law. The very names “congressional-executive agreements” and “sole executive agreements” evoke Articles I and II, and most scholarship engaging with these categories has focused on the extent to which the Constitution permits their use.16 But as important and foundational as this constitutional question indisputably is, there are other questions that one should ask, including: How does the United States decide which form of international commitment it will use? And what structural checks and balances operate in the current system? Answering these questions from a constitutional-law perspective will at best give rise to only partial answers, and at worst may give rise to misleading ones.

This Article explores the multiple pathways available for making international commitments. To understand the structural landscape in which they exist, we must take into account three strands of law—not just constitutional law, but also international law and administrative law. Each strand plays a crucial role in shaping how the United States makes international commitments. The structure of the international legal system both encourages the use of multiple pathways and affects what pathways are available in particular contexts. Constitutional law places meaningful doctrinal limits on the available pathways, although these limits now have more force with regard to how international commitments are implemented than with regard to how they are made in the first place. Perhaps most significantly, administrative law influences the choice of pathways by affecting how international commitments can be implemented, by underlying the State Department’s internal process for determining which pathway to pursue in a given context, and by shaping who is at the negotiating table for the United States.

This approach has important implications for separation-of-powers concerns. From the vantage point of constitutional law, the rise of myriad paths for making international commitments amounts to an unvarnished win for presidential power. The president has the power to choose which pathway to domestic approval to pursue for an international commitment, conditional on the use of this pathway being deemed constitutional. The more constitutionally permissible options there are, the more the president can evade the democratic and deliberative check of legislative review. As the Obama administration increasingly favored bypassing the subsequent approval of the Senate or Congress for international commitments, claims of presidential unilateralism followed quickly. “That’s outrageous, and it’s unlawful. And it’s a clear example of the executive overreach in the area of foreign affairs,” said the convener of a congressional hearing on the administration’s decision to join the Paris Agreement without going to the Senate.17

When all three strands of law are taken into account, the structural landscape looks quite different. For international law and administrative law have also given rise to constraints on presidential power. These constraints are ones that the Framers did not foresee, and yet they further James Madison’s goal of “contriving the interior structure of the government, as that its several constituent parts may, by their mutual relations, be the means of keeping each other in their proper places.”18 These constraints arise at all stages of an international commitment—negotiation, domestic approval, and implementation. Some of these constraints are independent of the constitutional constraints, but others have their strongest bite at times when the constitutional constraints are the weakest. In other words, the more the president seeks to bypass the Senate and Congress, the more he or she is likely to run up not only against constitutional concerns, but also against alternative constraints arising from international and administrative law and from institutions empowered by these bodies of law—including international organizations, administrative agencies, and occasionally even US states. The president’s power with respect to international commitments is thus not the power to avoid constraints entirely. Rather, it is the power to choose between different types of constraints.

The Paris Agreement exemplifies the structural claims made in this Article. From the perspective of constitutional limits on the approval process, it was a textbook example of unchecked presidential power. It was also a signature foreign policy achievement of President Barack Obama: in his words, a “historic” and “ambitious” agreement that will “establish[ ] the enduring framework the world needs to solve the climate crisis.”19 Yet while his predecessors accepted the need to take the two prior major multilateral agreements on climate to the Senate, Obama joined the United States to the Paris Agreement without seeking specific legislative approval—and did this precisely because such approval would not have been forthcoming.

From a broader perspective, however, the Paris Agreement reveals constraint upon constraint. Partly to avoid constitutional issues related to approval, the Obama administration had to accept strong checks in relation to the Agreement’s negotiation and implementation. In the negotiations, the executive branch had to operate within the limits arising from the international legal process, including a requirement of consensus, and yet persuade other nations to craft an agreement that satisfied its constitutional concerns. This process was so fraught that even in the final moments the negotiations almost broke down over a single word.20 During these negotiations, the executive branch also had to ensure that the resulting agreement could be implemented domestically through authority previously delegated by the Clean Air Act21 to the Environmental Protection Agency (EPA) and the states. This in turn required US negotiators to tie their negotiating position to the scope of the Clean Air Act, to involve the EPA in the negotiating process, and to pay close attention to underlying principles of administrative law and federalism. And as challenging as the negotiation of the Paris Agreement was for US negotiators, its future presents even more difficulties. President Donald Trump has already announced that he intends to withdraw the United States from the Paris Agreement, although this may not be his final word on the subject.22 And even if the United States were to remain in the Paris Agreement, there would still be much to be worked out both internationally through the process that governs further negotiations and domestically in terms of practical implementation. Overall, the story of the Paris Agreement illustrates both the reach and the limits of the president’s power to make international commitments.

This Article’s descriptive account of the current system for making international commitments stands apart from the desirability of this system from a normative perspective. On the normative question, this Article goes on to offer a qualified defense of the current system. This system strikes a reasonable balance between two related problems of our contemporary governmental landscape: presidential overreaching and legislative inaction. As to presidential overreaching, it incorporates a set of constraints that reduce the risks of abuses of power. As to legislative inaction, it provides the executive branch with alternatives to obtaining specific approval from the Senate or Congress, while folding in alternative forms of democratic accountability. In general, the rise of the current system is broadly faithful to other developments within public law, including the way in which administrative-law values have come to complement and sometimes substitute for constitutional principles.

The normative claims of this Article, if accepted, in turn have implications for several ongoing structural and doctrinal debates within the field of foreign relations law. One implication is that we should resist calls for sweeping changes to the process of how the executive branch makes international commitments, although some refinements would be beneficial. A second is that the Senate and Congress would do well to reduce the barriers to specific legislative approval that currently exist in order to give the executive branch more incentives to pursue traditional paths to approval. A third implication is that courts should be cautious in crediting certain strong claims of presidential power, including claims that the president and his or her agents have exclusive power to communicate with foreign governmental actors and claims that international commitments made solely by the executive branch have the power to preempt state law.

The rest of the Article develops the arguments summarized above. Part I describes how constitutional law, international law, and administrative law each contribute to the reconfigured landscape of international commitments. Part II explores the role that presidential power plays under this framework and discusses the rise of a diffuse yet robust set of checks on this power. Part III offers a qualified defense of the existing system and proposes several doctrinal improvements. Finally, Part IV illustrates the claims made in this Article through a case study of the Paris Agreement.

Two caveats to this Article require particular mention. First, in describing international law and administrative law as sources of growing checks on presidential power, this Article does not seek to imply that they operate in the same way or to the same degree. As a general matter, international law tends to operate more as an independent check on presidential power, and administrative law tends to operate more as a substitute for constitutional checks—but both the strength of these checks and their degree of interconnection to constitutional checks are highly dependent on context. Second, this Article focuses on describing and evaluating the current landscape rather than on dating its various features. The origins of alternatives to the Treaty Clause lie deep in our constitutional history, and modern international and administrative law began to influence the process by which the United States joins international commitments by at least the end of World War II. Regardless of whether the choices made by the Obama administration are characterized as incremental developments or seismic shifts, they illustrate the importance of all three strands of law for the process by which the United States joins international commitments. That administration’s commitment to global engagement, the partisan gridlock in Congress, and the increased importance of international regulatory cooperation together put international commitments made without the explicit approval of the Senate or Congress at the center of US foreign policy.

Looking ahead, the template created by the Obama administration may not get much use during the Trump administration. To date, Trump appears more focused on undoing international commitments than on making new ones—and his actions make clear to the international community just how fragile international commitments made without the Senate or Congress can be. But it is early in his term, and the creation or revision of other international commitments may lie ahead of him. Still further ahead, future presidents interested in pursuing international cooperation will likely find the precedents from the Obama administration to be compelling, especially as legislative approval for international commitments will remain difficult to get in the absence of structural reforms. The international commitments of the Obama administration thus merit close scrutiny not only on their own account but also for the future.

I. International Commitments and Three Strands of Law

How should we understand the different ways in which the United States enters international commitments? This question is usually approached from the perspective of constitutional law, which requires reconciling today’s multifaceted practice with the much more unitary approach set out in the Constitution. But as interesting as the constitutional perspective is, it explains only a fraction of what is actually going on. A far more complete picture emerges when we look not just to constitutional law but also to international law and administrative law. This Part describes how each of these three strands of law shapes the processes by which the United States joins international commitments.

A. Constitutional Law

The Treaty Clause sets a high bar for treaty approval. Even in the nineteenth century, proponents of international cooperation viewed the veto it gave to “a malcontent third” of the Senate as “the original mistake in the Constitution.”23 Especially with the development of the party system, the challenges of getting even slightly controversial treaties through the Senate have been and remain formidable. This in turn has led to deep interest in developing and justifying alternative paths to making international commitments.

By the early twentieth century, scholars and practitioners were beginning to categorize these paths and assess their constitutional reach.24 In 1922, not long after the Treaty of Versailles failed in the Senate, the solicitor for the State Department prepared a memorandum for a leading member of Congress that identified two alternative paths to international agreements.25 One path consisted of “agreements made pursuant to authority contained in acts of Congress” and the other of “agreements entered into purely as executive acts without legislative authorization.”26 As examples of the former, the memorandum named “postal arrangements made with foreign postal authorities; reciprocal tariff arrangements; arrangements respecting discriminatory duties, copyrights and trademarks; and agreements made with Indians.”27 As examples of the latter, the memorandum emphasized “agreements relating to the settlement of pecuniary claims of American citizens against foreign countries,” but also mentioned several examples pertaining to different issues.28

The constitutional scope of these alternatives became an important issue during World War II. Could the United States join the future UN Charter through a process other than the Treaty Clause, thus preventing a minority of the Senate from dealing it the same fate as the Treaty of Versailles? In a 250-page article published in two parts in the Yale Law Journal, Professor Myres McDougal and Asher Lans argued in the affirmative.29 Their analysis used categories similar to those set out in the 1922 State Department memorandum: they focused on (1) treaties, (2) congressional-executive agreements, and (3) presidential agreements.30

Unlike the memorandum, McDougal and Lans staked out bold constitutional claims. Drawing on historical practice, they argued that congressional-executive agreements were constitutionally permissible substitutes for treaties in all contexts.31 In making this argument, they equated congressional-executive agreements that Congress authorized before their negotiation and ones that Congress approved after their negotiation. (Today, we refer to these as ex ante congressional-executive agreements and ex post congressional-executive agreements, respectively.) Aggregating these types of agreements strengthened McDougal and Lans’s constitutional argument, as it enabled them to cite to more cumulative historical practice.32 McDougal and Lans also defended expansive presidential authority to make presidential agreements, although they did not claim that this authority reached as far as the other two categories.33 Throughout, they considered only the constitutionality of these various pathways and indeed firmly disclaimed any connection between these alternatives and international law.34

Reading McDougal and Lans in some ways shows just how little foreign relations law scholarship has changed over the years. The three categories used by McDougal and Lans have become “Lesson I of Foreign Relations Law 101,” with the slight update that presidential agreements are now more commonly called “sole executive agreements.”35 Scholarship considering these pathways continues to focus on their constitutional dimensions. As to congressional-executive agreements, scholars are still debating the extent to which they are constitutionally interchangeable with treaties. Professors Bruce Ackerman, David Golove, Oona Hathaway, Peter Spiro, Laurence Tribe, and others have all written on this topic, reaching a range of conclusions about interchangeability.36

As to sole executive agreements, their meaning and their reach has assumed increased importance in this era of high congressional gridlock and presidential unilateralism. Should we classify international agreements for which there are some signals of congressional support for the aims of these agreements as sole executive agreements or instead as something more nuanced? Can the executive branch commit the United States to any kind of international agreement? Or is the reach of sole executive agreements constitutionally limited in one or more of the magnitude, duration, or subject matter of the agreement? President Obama’s willingness to make important international agreements without explicit authorization from the Senate or Congress fueled an intense constitutional conversation in both Congress and the academy.37

A particularly important issue for sole executive agreements is their implementation. Supreme Court precedent establishes that sole executive agreements relating to claims settlement can be enforceable domestic law for purposes of the Supremacy Clause.38 For the most part, however, the president needs Congress in order to create domestic law. This constitutional limit is important—functionally the most important domestic legal limit on the scope of sole executive agreements. But as discussed later in this Part, preexisting congressional delegations to the president or to administrative agencies place considerable power in the hands of the executive branch, which can be used to implement carefully crafted international agreements that lack specific congressional authorization.

The constitutional reach of congressional-executive agreements and sole executive agreements are important issues in foreign relations law—yet these categories and the constitutional questions they evoke only partly explain how the United States participates in international commitments. These categories do not, for example, encompass nonbinding commitments, which are a major way in which the United States conducts diplomacy. Nor do these categories explain how decision-making occurs within the executive branch between the president, executive branch agencies, and independent agencies with respect to international commitments. The US government functions very differently in 2016 from how it did in 1922 or in 1945, yet foreign relations law scholarship still tends to treat the executive branch as equivalent to the president. As Professor Koh put it recently, “[w]e need a better way to describe the texture of the tapestry of modern international lawmaking and related activities that stays truer to reality than this procrustean construct” of treaties, congressional-executive agreements, and sole executive agreements.39 Such a description must look beyond constitutional law.

B. International Law

“Fragmented,” “multi-hub,” “transnational,” and “pluralist”—these are words that scholars use to describe today’s international legal order.40 It has changed radically since World War II. The UN Charter has come into being, with its focus on peace, and it remains at the core of the international legal system. But it is only a piece of the web of international legal regimes that have emerged in the last sixty years and especially since the end of the Cold War. Separate multilateral agreements and institutional structures exist with regard to trade, finance, disarmament, humanitarian law, the environment, and human-rights law, in addition to countless regional and bilateral arrangements. In addition to these various fora of cooperation among nations, the international legal order increasingly encompasses participation by nongovernmental organizations, corporations, and subnational governmental entities.

Foreign relations law scholarship rarely considers the modern structure of the international legal system in thinking about the pathways by which the United States joins international commitments. When international law comes up at all in the context of the pathways, it is usually with regard to whether a now-eroded distinction between “treaties” and “agreements” found in the work of the eighteenth-century international legal scholar Emmerich de Vattel sheds light on the constitutional scope of treaties as opposed to congressional-executive agreements and sole executive agreements.41 This minimal use of international law is consistent with the constitutional-law perspective that dominates the field.

The architecture of the international legal order nonetheless plays a vital role in shaping how the United States joins international commitments. As a formal matter, international law has little to say regarding the constitutional pathways by which the United States enters into international agreements.42 In practice, however, the superstructure of international law affects the form of international commitments, the identity of governmental actors who participate in them, and the fora within which they are made. These factors in turn influence the internal pathways that are available to and used by the United States.

One important feature of the international legal order is the place that it accords to soft-law commitments: formal political undertakings that are not binding under international law. These commitments can be onetime affairs, such as the Iran nuclear deal reached during the Obama administration. They can also be made under the auspices of long-standing institutional structures, such as the food safety standards developed through the Codex Alimentarius.43 Such soft-law commitments have largely been overlooked by foreign relations law scholars, even though they have now been around for quite some time.44 Because they do not impose any international legal obligations on nations, they are not treaties, congressional-executive agreements, or sole executive agreements and are thus largely invisible from a constitutional-law perspective.45 Yet because they are a permissible way within the international legal order of doing business—including very important business—they have become an important way in which the United States engages in international cooperation.

The rising pluralism of the international legal system also opens the door to participation by a variety of domestic legal actors. Although the text of the Constitution bars US states from making binding international agreements without the consent of Congress,46 in practice states have come to engage in considerable cooperation with foreign nations and subnational entities.47 A recent example is California’s highly formalized arrangement with Quebec to integrate their respective cap-and-trade programs for greenhouse gases.48 Even looking within the US national government, international cooperation is increasingly carried out by actors outside the State Department and the White House, including at times by leaders of independent agencies. The structure of international institutions—especially soft-law ones—facilitates this actuality. By way of example, because the Basel Committee is in essence a meeting of international bankers, US participation there is led by the Federal Reserve Board, with the Office of the Comptroller of the Currency, the Federal Deposit Insurance Corporation, and even the Federal Reserve Bank of New York also participating.49

International institutions also shape and stabilize the international processes for making international agreements in ways that can influence the domestic pathways for joining these agreements. Sometimes an international agreement itself spells out the future process, like the way that the UN Charter gives the Security Council the “primary responsibility for the maintenance of international peace and security.”50 Other times, international practice establishes patterns that serve as strong defaults. Major environmental commitments tend to take the form of multilateral, internationally binding agreements tailored to specific environmental issues;51 international investment commitments rest either in bilateral agreements or as part of regional trade pacts;52 and so on. As discussed in the next Section, the administrative process by which the State Department determines the pathway by which the United States will join an agreement in turn takes account of these patterns.

C. Administrative Law

A third strand of law that affects how the United States enters into international commitments is administrative law.53 Executive branch agencies and independent agencies play crucial roles in the negotiation and implementation of international commitments, often but not solely due to congressional delegations of authority. In the process, principles of regularity, accountability, and transparency have come to play an increased role both in the making of international commitments and in decisions about the pathways by which these commitments shall be made.

The constitutional focus of foreign relations law tends to lead to equation of the president and the executive branch.54 In her important article on presidential power in the making of international agreements, for example, Hathaway concludes that there were almost four thousand “international agreements entered by the President acting alone” between 1980 and 2000.55 This description of the “President acting alone” is understandable from a constitutional perspective. Practically speaking, however, the president could not have personally negotiated (or likely even known about) most of these agreements—instead, they would have been done by agencies.56

Executive branch agencies owe strong allegiance to the president, both legally and functionally. Their leaders are appointed by the president with the advice and consent of the Senate, can be removed by him or her, and are subject to various forms of White House supervision, especially for important matters. Moreover, they are more likely in the foreign affairs context than in the domestic context to be exercising delegated presidential powers, including the power to conduct international negotiations and the commander-in-chief power. This is especially true of the Departments of State and Defense, but other executive branch agencies can channel these powers, as well.

But the loyalties of executive branch agencies are nonetheless divided. They are of course answerable to Congress as well as the president. Sometimes Congress has passed laws that explicitly relate to how agencies will participate in international negotiations—for example, Congress requires the Food and Drug Administration (FDA) to publish an annual notice describing intended US participation in standard setting done by the Codex Alimentarius and to provide an opportunity for public comment on this issue.57 In addition, the implementation of many international commitments—especially ones with a regulatory component—depends on the domestic legal powers that Congress has delegated to these agencies. To continue with this example, the FDA implements decisions of the Codex Alimentarius through its preexisting powers under the Food, Drug, and Cosmetic Act.58 Finally, at a general level, the agencies are answerable to Congress because their budgets come from Congress. To the extent that they focus on presidential interests at the expense of congressional ones, they may face not only congressional complaints, inquiries, and hearings, but also threats to their bottom line. In a 2016 hearing on the Paris Agreement, for example, one witness encouraged Congress to use its appropriations power to block the EPA from expending any funds that might implement the Agreement.59 In addition to Congress, agencies are also strongly responsive to the interests of their own civil servants and to various outside constituencies.60

The distinction between the president and executive branch agencies with respect to international commitments is demonstrated by efforts by the president to supervise agency actions in this context. For international agreements, a federal regulation provides that “[t]he Secretary of State is responsible, on behalf of the President, for ensuring that all proposed international agreements of the United States are fully consistent with United States foreign policy objectives.”61 Agencies must receive the State Department’s permission to negotiate them and must clear the final texts with the State Department prior to signature.62 There used to be no parallel for nonbinding commitments, but in 2012 Obama issued Executive Order 13609 on Promoting International Regulatory Cooperation, which sets up procedures for furthering interagency cooperation with respect to “international regulatory cooperation activities that are reasonably anticipated to lead to significant regulatory actions.”63 This coordination is to run through a Regulatory Working Group chaired by the head of the Office of Information and Regulatory Affairs within the White House.64

As all this suggests, the making of international commitments is a ripe venue for the same kind of institutional dynamics that play out in domestic administrative law. The president holds the reins, and yet the agencies have considerable power to shape their own agendas. They also have tools to resist presidential oversight. Professor Jennifer Nou’s work describes how, in the domestic rulemaking context, agencies can deploy techniques that “functionally serve to bypass [presidential] review, calibrate its scrutiny, or truncate the amount of time available” for it.65 Such ability of federal agencies to evade or resist presidential power serves as a practical check on the reach of that power. Similar tools will presumably be available for international regulatory coordination, especially when it will ultimately result in domestic rulemaking.66 Finally, also as in the domestic legal context, independent regulatory agencies have even more room to maneuver with respect to international commitments than do executive branch agencies67 —as illustrated by the Federal Reserve’s powerful role in the negotiation and implementation of Basel III. Indeed, Professors Peter Conti-Brown and David Zaring observe that often “the Fed sets its own foreign policy,” sometimes amidst disagreement from executive branch agencies.68

The influence of administrative law on the making of international commitments can be found not only with respect to institutional dynamics but also with respect to underlying values of procedural regularity and reasoned decision-making. The very process by which the executive branch decides how to join an international commitment is designed to promote these values. To decide whether an international agreement should be made as a treaty, a congressional-executive agreement, or a sole executive agreement, the State Department engages in what is known as the “Circular 175 Procedure”—a process set out in the State Department’s Foreign Affairs Manual69 and complemented by provisions in the Code of Federal Regulations.70 The C-175 Procedure sets forth eight factors that State Department lawyers should consider in determining the appropriate pathway for an international agreement.71 These factors relate to constitutional concerns, to the international context in which the agreement was reached, and to various practical considerations. For example, the factor of “[p]ast U.S. practice as to similar agreements” is tied to both international law and constitutional law, as the international legal context helps determine what constitutes a “similar agreement” and past practice is relevant for assessing constitutional concerns. The C-175 Procedure also provides for consultation with congressional leaders and committees “as may be appropriate” when there is debate about whether or not an agreement should be done as a treaty.72

Administrative-law principles also infuse other aspects of the commitment-making process. The C-175 Procedure calls on the negotiators of international agreements to create an opportunity for public comment whenever, in the view of the State Department, “circumstances permit.”73 With regard to nonbinding international commitments, Congress sometimes similarly insists that agencies provide an opportunity in the course of negotiations for notice and comment.74 In addition, in the Case-Zablocki Act,75 Congress required that all international agreements other than treaties be submitted to Congress within sixty days of their entry into force.76 Finally, when administrative agencies rely on congressionally delegated authority in implementing international commitments into domestic law, their actions will be subject to the Administrative Procedure Act77 (APA) and whatever administrative procedures are set forth in the legislation from which their delegated authority stems.

Administrative-law principles thus pervade the making of international commitments. But it is important to acknowledge that they often do so in weaker ways than in many domestic legal contexts. For one thing, administrative agencies engaged in the making of international commitments can invoke various limits or exceptions that reduce transparency and public participation. The APA exempts various foreign affairs–related issues from notice-and-comment rulemaking,78 and the Supreme Court has interpreted the Government in the Sunshine Act79 narrowly when it comes to the participation of US agency leaders in international negotiations.80 For another thing, the executive branch maintains a veil over how it decides whether to push for binding versus nonbinding commitments and over how the C-175 Procedure is applied to particular agreements. Thus, the State Department does not typically publish or otherwise disclose its legal reasoning with respect to the C-175 factors (and therefore provides no opportunity for public comment on this reasoning). Overall, while the constraints of administrative law with respect to international commitments can be less extensive than the constraints created by administrative law in traditional domestic contexts, they are still considerable.

II. International Commitments and the Separation of Powers

The rise of multiple pathways to making international commitments has fundamentally reshaped the separation of powers with respect to them. In the years since the Founding, the executive branch has gained a new and crucial power—the power of choice. The executive branch selects the pathway by which the United States will join an international commitment. In constitutional terms, this gives the president the opportunity to bypass the Senate and often Congress, and thus to evade the traditional constraints on presidential overreaching. Yet with this opportunity come new checks rooted in both international law and administrative law. The more the president chooses to bypass the Senate and Congress, the stronger these checks are likely to be. Overall, the rise of multiple pathways has led to a structural shift away from a single, concentrated check on presidential power to a set of checks that are individually diffuse but collectively strong. These checks are spread across all three stages of the commitment-making process: negotiation, domestic approval, and implementation. Not all the checks described here are present for every international commitment—there is considerable variation across particular commitments—but most commitments are subject to a robust set of checks. This Part describes this structural shift and connects it to broader trends with respect to the separation of powers.

A. The President’s Power to Choose

From a constitutional perspective, the rise of alternatives to the Treaty Clause has greatly enhanced the president’s powers. With these alternatives, the president has gained not only the substantive power to make at least certain types of agreements under his or her own authority, but also the procedural power to choose the pathway by which an international commitment will be approved. This power to choose is immensely important. Given all the constitutionally defensible alternative pathways available today, the president can almost always pursue an approach that bypasses the subsequent approval of the Senate and Congress.

Historical practice firmly establishes the president’s authority to choose the pathway by which an international commitment will be approved.81 Of course, the president may pick only a pathway that is constitutionally justifiable. But in choosing a pathway, the president can simultaneously widen it. As such choices have accumulated through historical practice over time, they have made these alternative pathways increasingly defensible as a matter of constitutional law.82 Today, the president always has the following three options: (1) for agreements that are binding as a matter of international law, to go to the Senate for approval pursuant to the Treaty Clause; (2) for any commitments dealing with issues within the scope of Congress’s Article I powers, to go to Congress for its approval following their negotiation; or (3) for commitments that are nonbinding as a matter of international law, to approve them under his or her own authority.83 In addition, the president has a fourth option for at least many international agreements, which is to approve them without any postnegotiation action by the Senate or Congress. The constitutional basis for this last option is one or more of a prior congressional authorization, a prior authorization set out in a treaty, prior legislation that implicitly supports the president’s action, or the president’s independent constitutional powers.84

This power to choose gives the president the ability to sidestep the check of legislative approval. Especially for important agreements, this raises structural constitutional concerns. As Professor Louis Henkin put it, the “highly uncertain” “reaches of the President’s power to make executive agreements . . . might tempt activist Presidents into far-reaching undertakings.”85 If going to the Senate or to Congress for specific approval is only optional for the president, then there are few if any meaningful constitutional checks on the process of approving domestic commitments.

Following the Obama administration’s strategic use of the president’s power to choose, this concern has received significant attention. “Expanded use of sole executive agreements . . . reduces democratic control over international lawmaking . . . [and] raises serious questions about the potential of these agreements to undermine democratic lawmaking writ large,” write Professors Hathaway and Amy Kapcyzynski with regard to the Anti-Counterfeiting Trade Agreement86 (ACTA).87 In a similar vein, Professor Michael Ramsey considers that the “aggressive approach” taken by the Obama administration with regard to the Iran deal and the Paris Agreement “threatens to evade the limitations on the President imposed by the treaty making power.”88 In Congress, the Committee on the Judiciary Task Force on Executive Overreach held a hearing on this very issue.89

The president’s power to choose does indeed weaken the core constitutional constraint on international commitments. Yet this does not translate into a straightforward increase in presidential power. When we take into account other bodies of law, other institutional actors, and other stages of the commitment-making process, the picture is far more complex and constrained. As the analysis below shows, the president’s power of choice is not the power to avoid constraints entirely. Rather, it is the power to choose between different types of constraints.

B. Diffuse Checks, Collective Balance

Checks on presidential power do exist under the myriad pathways for making international commitments, but they look very little like the check built into the original constitutional design. The Treaty Clause provides a single, concentrated constitutional check on presidential overreach. At its core, this check is grounded in one institution (the Senate), tied to one stage of the commitment-making process (domestic approval), and manifested in one type of activity (a supermajority vote).90 By contrast, the various pathways for making international commitments have given rise to a set of checks which are spread out across institutions, across phases of the commitment-making process, and across types of activities. These checks stem from international law and administrative law as well as from constitutional law. Collectively they serve as robust structural safeguards on presidential power.

While some of these checks are independent, others are closely interconnected. These relationships can be seen by looking separately at the three phases of the commitment-making process: negotiation, domestic approval, and implementation. As a general rule of thumb, the more the president wishes to duck the need for approval from the Senate or Congress, the more limits he or she must accept with respect to the negotiation and implementation of the commitments.

1. Negotiation.

International negotiations have always come with the practical limit that it takes at least two to tango. In today’s world, these negotiations also take place against the backdrop of a well-developed international legal order. The president and his or her agents are bound structurally and substantively in negotiations by this international legal order, even as they also face domestic legal constraints imposed by the future need for domestic approval and implementation of any negotiated agreement.

International law is no longer in an era of creation like that following the end of World War II or, later, the end of the Cold War. Substantively, major multilateral agreements already address most areas of international concern, and their foundational legal principles cannot realistically be revisited. Structurally, these agreements also have created institutions through which future negotiations are channeled—for example, worldwide trade negotiations are held under the auspices of the World Trade Organization, and worldwide security or human-rights negotiations go through the United Nations. This thick existing framework limits what the president and his or her agents can pursue in worldwide negotiations. It does not make new worldwide agreements impossible, but it does notably “condition[ ] the traditional use of state power,” including by bringing formalized procedures and often some level of transparency of process into the negotiations.91

The existing international legal structure also limits what can be done through nonbinding commitments and through bilateral or regional agreements. In negotiating these commitments, executive branch officials likely confront fewer procedural limits on the negotiating process but more substantive limits set by preexisting international law. Even modest commitments that the president probably knows little or nothing about need to be negotiated with an eye to the broader international superstructure. Bilateral fisheries agreements, for example, are structured to ensure “consisten[cy] with international law.”92 For major commitments, this need is even more acute. The terms of the Iran deal, for example, had to be carefully negotiated in light of the existing international legal backdrop. Among other things, because these terms dealt with sanctions previously imposed by the UN Security Council, the negotiators had to make sure that the deal would be one that the Security Council would accept and embrace through the passage of a subsequent resolution.93

Turning to the domestic legal backdrop, the president’s power to negotiate seems all-encompassing at first. As a matter of constitutional law, the classic statement is that “the President alone has the power to speak or listen as a representative of the nation. . . . Into the field of negotiation the Senate cannot intrude; and Congress itself is powerless to invade it.”94 But even accepting these lines as formally correct—and the Supreme Court has recently sent mixed signals on that front95 —they do not reflect the practical realities that stem from institutional design and from shadow-of-the-law effects related to approval and implementation. In practice, these leave the president limited in terms of both his or her control over the negotiations and the substance of negotiating terms.

If the president will need the Senate to approve a treaty or Congress to approve an ex post congressional-executive agreement, then as a practical matter the president needs to solicit congressional input during the negotiating process. This happens informally and also in certain formal ways. In relation to treaties, the Senate sometimes passes resolutions requesting that the president undertake certain negotiations or seek certain terms.96 In relation to trade agreements, Congress has over the years passed “fast-track” statutes that give the president an expedited route to an ex post floor vote if the president pursues certain negotiating objectives and involves members of Congress in the negotiations in certain formalized ways.97 And even when they are not asked for their views, members of Congress can find ways to weigh in—as with the letter sent by Senator Tom Cotton and numerous other senators to Iranian leaders during the negotiations of the Iran deal.98

A more overlooked but comparably important way in which the president must share negotiating power stems from the role of administrative agencies. The president depends on agencies to carry out most international negotiations. The routine small-scale negotiations that make up the vast majority of international commitments are less exercises of presidential power than of bureaucratic power over which the president keeps a supervisory eye through his or her agents in the State Department and the White House. And even major negotiations can occur with relatively little oversight from the president. The increased use of nonbinding political commitments has increased the president’s power to bypass the Senate and Congress, but it has also increased the power of agencies and US states to negotiate in the foreign affairs space with relatively little presidential oversight.99 In addition to the example of Basel III, international regulatory coordination with respect to insurance policy is currently carried out through cooperative efforts between US states and various federal agencies, including the Federal Insurance Office, an office within the Treasury Department created by Congress for this purpose in the Dodd-Frank Act.100

Even when the negotiation of international commitments lies with the president’s traditional diplomatic agents in the State Department, these diplomats can have strong incentives to listen to agencies. Agencies can bring valuable contributions to the table on commitments that relate to their expertise. Even more importantly, an agency will be essential to implementation as an institutional matter if the president intends to bypass subsequent approval by the Senate or Congress and instead implement the commitment by drawing on an agency’s preexisting domestic authority. The negotiation of the Minamata Convention on Mercury is a good example. This multilateral international agreement requires state parties to take specific steps to reduce the amount of mercury in the environment.101 The United States joined this agreement in 2013—the same year that its negotiation was finalized—without receiving the advice and consent of the Senate or the ex post approval of Congress.102 The State Department explained that “[t]he United States has already taken significant steps to reduce the amount of mercury we generate and release to the environment, and can implement Convention obligations under existing legislative and regulatory authority.”103 To ensure that the United States could in fact implement the Convention in light of preexisting regulatory power, however, US diplomats had to make sure during the Convention’s negotiations that its obligations would not go beyond what the United States already had the power to implement. These diplomats thus had to engage the EPA in these negotiations and rely on its understanding of its regulatory power under congressional statutes.104 In turn, US negotiators could credibly signal that they were constrained negotiators to foreign counterparts.105

2. Approval.

The choices available to the president for approval depend on the form and content of the negotiated agreement and on whether the executive branch has the preexisting power to implement it. This choice thus implicates issues of international law, constitutional law, and administrative law.

For commitments that are nonbinding as a matter of international law, the president can approve them without further action by the Senate or Congress. In theory, this gives the president enormous power—so much so that scholars are beginning to urge that constitutional limits be read into the president’s power to approve nonbinding commitments.106 In practice, the power looks less overwhelming. Internationally, the use of a nonbinding commitment must not only be consented to by negotiating partners (as with all international commitments), but will also need to be structured in a way that is consistent with existing international law because a nonbinding agreement cannot change international law.107 Domestically, the president may face political pressure to pursue a binding rather than a nonbinding agreement. Moreover, as discussed shortly, if the executive branch does not have the preexisting power to implement the commitment, it will need to obtain either implementing legislation from Congress or the support of US states.

For international agreements, the president faces legal as well as political constraints on the power to choose. Internationally, there may be political constraints imposed by negotiating partners, who may want the president to obtain the stronger signal of commitment embodied by a domestic pathway that includes the explicit approval of the Senate or Congress.108 Domestically, as a matter of process, decision-making with respect to the president’s power of choice is carried out by officials in the State Department who are applying the C-175 Procedure. For almost all agreements, their decisions will presumably track existing practice. It is in only a few—but important—situations when presidential strategy will presumably come into play. Indeed, the practice of the Obama administration suggests the limited practical reach of the president’s power to choose the path of international agreements. The administration has sent treaties to the Senate when past practice almost uniformly supports the use of the treaty route for a particular type of major agreement, including the New START Treaty (which received advice and consent by a 71–26 vote)109 and the UN Convention on the Rights of Persons with Disabilities (which failed by a 61–38 vote).110 Consistent with both past practice and the practical need for implementing legislation, the Obama administration could not join the TPP without congressional approval, and this agreement has since been abandoned by the Trump administration.111 By contrast, when the Obama administration signaled its intent to bypass the subsequent approval of the Senate or Congress despite substantial practice in favor of pursuing such approval, it did so only when (1) some plausible past practice supported its approach and (2) it considered that it has the preexisting authority to implement these agreements. This was the case for ACTA, the Minamata Convention, and the Paris Agreement.112 Notably, the Obama administration faced considerable pushback with regard to the legality of its decision to bypass the Senate and Congress for both ACTA and the Paris Agreement.113 (The less prominent Minamata Convention flew mostly under the radar.) Perhaps in part because of this pushback, the Obama administration never acted on its asserted power to join the United States to ACTA, although it did indeed join the Paris Agreement and the Minamata Convention.

3. Implementation.

As with the negotiation and approval of international commitments, there is no one-size-fits-all account of implementation. Instead, how a commitment is implemented depends on what the commitment is, how it has been approved, and what existing law relates to its implementation. These factors determine the scope of what the president can do.

At the international level, major multilateral commitments often require transparency with respect to implementation. The Minamata Convention, for example, requires each state party to file reports on how it is implementing the Convention and provides certain soft oversight tools to the Conference of the Parties and an Implementation and Compliance Committee.114 These requirements are designed to enable countries to keep an eye on each other, but they also provide sunshine for domestic actors who wish to scrutinize executive branch action with respect to the implementation of commitments.

Turning to domestic law, to ensure that the United States can implement an international commitment, the executive branch will generally need one or more of the following: (1) an independent power of the president that provides for implementation, (2) the international commitment to receive advice and consent from the Senate as a treaty and/or congressional legislation implementing it, or (3) the terms of the commitment to be ones that can be or already are implemented pursuant to preexisting statutes.115 Each of these categories comes with its own set of constraints.

The main constraint on the president’s independent powers is their limited scope. The president can fulfill commitments to communicate with other nations, to recognize foreign nations, to take steps pursuant to the commander-in-chief power, and to settle certain claims between US citizens and foreign states and other foreign entities.116 Most of these powers relate to outward-facing actions by the United States rather than to the regulation of the conduct of private citizens. The Supreme Court has indicated that, outside the context of claims settlement, the president has starkly limited independent power to implement an international commitment in a way that affects the domestic legal rights of US states or private parties.117

When the president seeks the advice and consent of the Senate and/or specific legislation from Congress, the check is the legislative process, with all the democratic principles that it embodies. For international agreements, the president will typically get any needed legislative authorization to implement the agreement simultaneously with the process of domestic approval.118 For nonbinding commitments, implementing legislation is more likely to postdate the approval process. For example, President George W. Bush made a soft-law commitment on behalf of the United States to stem the import of conflict diamonds and then sought and obtained a specific statute to implement it.119 If there is divergence between the implementing legislation and the international commitment, then it is the terms of the implementing legislation that will control within the United States.120

As to the third category, the constraints that exist when a preexisting statute authorizes implementation will vary with the content of the statute. The broader the statute’s scope, the more the executive branch can use it as implementing authority. The more discretion the statute gives to executive branch actors, the more these actors can use this discretion in the service of implementing an international commitment. And the fewer procedural constraints the statute (and other relevant statutes) has, the easier it will be to use this statute to implement the international commitment.

In practice, however, Congress does not typically pass statutes with sweeping substantive reach, enormous executive branch discretion, and minimal procedural safeguards. Congress comes closest when enacting traditional foreign policy statutes, such as those relating to sanctions, and even these will come with nontrivial limits. For example, in the years prior to the Iran deal, Congress passed numerous statutes imposing sanctions on Iran.121 For the most part—but not entirely—these statutes explicitly gave the president the authority to waive these sanctions under certain conditions.122 Because President Obama planned to rely on his discretion to waive these sanctions in order to implement the core US commitments under the Iran deal, his negotiators had to ensure that the deal stayed within the limits of his discretionary authority. To give only one example, one of the congressional statutes at issue expressly authorized state and local governments to divest from actors owning a certain stake in Iran’s energy sector.123 Obama thus had to make sure that the Iran deal did not commit the United States to ending these divestment measures, as he could not have implemented such a provision.124 Here, as in other ways with respect to the Iran deal, the limits attached to implementation shaped the content of the negotiations. In addition, the very fact that the Iran deal relied so heavily on Obama’s discretionary authority in implementation leaves it deeply vulnerable to abandonment by his successor.125

When the president seeks to use domestic policy statutes for the implementation of international commitments, he or she will likely face even more notable constraints. These statutes may have broad substantive scope, but the discretion they provide is likely to be bestowed on an agency rather than on the president and to include procedural safeguards like judicial review. Consider, for example, a statute that gives broad discretion to an administrative agency to regulate on health, food safety, or the environment. As a practical matter, if the president wishes to implement an international commitment by using the discretion accorded to the agency, then he or she must negotiate it within these limits, must obtain buy-in from the agency, and must also run the risk that this buy-in will be withdrawn—and the commitment’s implementation put at risk—during a future administration.126 In addition, the agency’s actions in implementing the statute will almost certainly be subject to judicial review (under the APA or the statute’s own terms) with respect to whether these actions are consistent with the statute and neither arbitrary nor capricious. As Professor Jack Goldsmith put it, in considering Obama’s energetic use of these kinds of international commitments, their underlying “domestic authority . . . , unlike many assertions of presidential foreign relations power, can be reviewed by domestic federal courts,” making for “significantly more accountability than the vast majority of presidential actions in foreign relations.”127

C. International Commitments in Perspective

The restructured separation of powers described above occurs in the context of international commitments. At a higher level of generality, however, it is consistent with a broader account of how, across many areas of governance, the various strands of applicable law are not hermetically sealed from each other. Instead, they interact in ongoing ways that affect not only the functional separation of powers but also the development of legal doctrine.

Scholars of constitutional law and administrative law have explored their relationship in domestic contexts. As one notable example, Professor Gillian Metzger argues that “ordinary administrative law” serves as “a species of constitutional common law,” bringing constitutional values into the practice of governance.128 Agencies are sensitive to constitutional concerns in their decision-making, and courts applying administrative law in reviewing agency decisions draw on constitutional separation-of-powers principles.129 In turn, “ordinary administrative law and administrative practice [have shaped] the scope of constitutional requirements [in a way that] is of a piece with the numerous ways in which constitutional law has bent and transformed in response to the institutional and regulatory needs of the modern administrative state.”130 Administrative law comes to substitute in part for constitutional law, but it does so largely because it also provides for checks and balances.

Just as functional and institutional needs led constitutional law to accept the administrative state, so have similar needs led to the rise of alternatives to the Treaty Clause. These alternatives in turn rely heavily on administrative law for implementation and thus on all the underlying constitutional principles that this law incorporates. This includes any use of federalism. The more the president seeks to make international commitments whose implementation will require the cooperation of US states, the more influence these states will have on international negotiations despite the president’s formal constitutional control over such negotiations.

The insights in this Article also relate to a body of scholarship that considers the interplay between international law and constitutional law. While in theory international law and constitutional law could serve as structural substitutes,131 in practice their relationship is more complicated. Judicial review is not always available for foreign affairs matters,132 and, especially in the security context, the executive branch sometimes plays international law and constitutional law off each other in ways that erode them both.133 This literature urges caution before assuming that international law can serve successfully as a long-term structural substitute for constitutional law.

Unlike in the security context, however, the president faces important process checks with regard to international commitments at the international level that increase its reliability as a constraint. The executive branch can undertake uses of force based on broad interpretations of international law on self-defense without getting other nations to agree with this position, but it cannot negotiate a multilateral commitment without getting other nations on board.134 The agreement of these nations would be needed not only to make a commitment, but also to make changes in the international legal superstructure of existing major international agreements that exist in most areas of international cooperation. The constraints that apply to the making of international agreements are thus more self-enforcing than the constraints that tend to exist with regard to the interpretation of substantive international law.

III. International Commitments and the Future

Scholars and the political branches have long accepted the practical necessity of putting the executive branch in control of minor commitments, which are the bread and butter of US international engagement. But what about major commitments? The Obama administration deliberately avoided seeking the explicit approval of the Senate or Congress for some important commitments, thus extending past practices with regard to alternative pathways and achieving core policy interests. This Part defends these developments from a normative perspective, taking into account the importance of international cooperation, the existence of the checks and balances described earlier in this Article that are distinct from those embodied in congressional approval, and the further check that arises from the existing uncertainties surrounding where the constitutional boundaries lie. In light of this normative defense, this Part then intervenes in several ongoing structural and doctrinal debates in foreign relations law.

A. In Praise of Multiple Pathways

What are we to make of the current system and in particular of the executive branch’s ability to make major international commitments without getting specific legislative approval? The clearest line drawn by this system with respect to the separation of powers between the executive branch and the legislature is one whereby the executive branch can enter into international commitments on its own but needs some kind of preexisting or subsequent action from the Senate or Congress in order for the terms of these commitments to be implemented through domestic law. This line may not be perfect, but it is better than one that requires the executive branch to get specific legislative approval for all major commitments, or for the narrower subset of all major agreements, or even for the still narrower subset of all major agreements whose implementation requires the use of domestic law. This line promotes international cooperation; it satisfies core structural principles related not only to checks and balances but also more specifically to democratic accountability; and it relies constructively on the very uncertainty surrounding it. Together, these virtues provide the current system with a solid normative grounding.

1. Promoting international cooperation.

The current system favors international engagement by giving the executive branch more ways to make international commitments. Without the rise of ex post congressional-executive agreements as an alternative to the Treaty Clause, we might not have had the annexation of Texas, the creation of the Saint Lawrence Seaway, or free trade agreements like NAFTA.135 If the president could not make any major international commitments on his own, we might not have had the Paris Peace Accords ending the Vietnam War, the Shanghai Communique leading to the normalization of relations with mainland China, or the Helsinki Accords bettering relations between the Soviet Union and the West.136 If the executive branch needed specific legislative approval in order to make international commitments whose implementation depends on US domestic law, then we might not have the Convention on Long-Range Transboundary Air Pollution and its various protocols,137 Basel III,138 the Iran deal, or the Paris Agreement.

One may disagree with the merits of one or more of these commitments, but it is hard to argue with their collective demonstration that the Treaty Clause presents too high a bar to action to be the only route available for major international commitments. The message of the second and third sets of examples is that sometimes specific congressional approval similarly presents too high a bar. In the domestic context, some form of congressional action is usually a starting point for executive branch action. This creates enormous challenges for governance, particularly with the rise of partisan polarization and its attendant congressional gridlock in recent years.139 In the international context, always requiring specific congressional action would be even less palatable. International cooperation is often highly desirable, and sometimes, especially in the context of peace and security, it is essential. Because the United States has less control over international affairs than domestic ones, the executive branch can have an even higher functional need to act in this context—and to do so with flexibility and sometimes with speed. Indeed, if specific congressional approval were required for all major international commitments, then these commitments would face greater process hurdles than exist for domestic legislation, as they would require not only the agreement of Congress and the president but also of the international negotiating partners.

The appropriateness of allowing the executive branch to make major international commitments without specific legislative approval is made all the more clear when considered in tandem with the president’s war powers. The president today can initiate the use of force with relatively few checks, either internationally or domestically.140 If specific legislative approval were required for international commitments, then this would effectively incentivize the president toward force over diplomacy. By allowing the president to unilaterally end the Vietnam War, normalize relationships with China, and ease relations with the Soviet Union, the current system gives greater space for diplomacy. This point holds as well for the Iran deal. If President Obama had needed specific congressional approval for the deal, then he would have had greater incentives to pursue military alternatives to stopping Iran from getting a nuclear weapon.141 The war–diplomacy trade-off is less evident for climate-change agreements—though still present in the long run142 —and absent from some other forms of international commitments, but when it exists it is an important factor to consider.

In defending the use of multiple pathways as a means to promoting international cooperation, this discussion assumes neither that multiple pathways are always necessary to such cooperation nor that such cooperation is always good. The arguments are instead qualified ones. The lower the barriers to international commitments that there are as a matter of legislative process, the less functional need there is for alternative pathways. If our system of governance were parliamentary or were structured in ways that made it much less prone to polarization, then the functional need for alternative pathways would be lessened. Without such changes (and at present they seem highly unlikely), the existence of multiple pathways serves as a workable and valuable substitute. Similarly, the claim that international cooperation is desirable should not be mistaken for the claim that it is always desirable (let alone that it will please everyone). International cooperation is itself only a means to the ends of peace, prosperity, the promotion of human rights, and the protection of shared commons; it can also be used in ways that undermine these values. We should hope and expect that the executive branch will ordinarily seek to use international cooperation in favor of good ends rather than bad ones. The structural checks and balances that come with the alternative pathways increase the likelihood that this will prove true.

2. Satisfying core structural principles.

Turning to structure, the current system builds in important checks on presidential overreaching, including checks that are tied to legislative involvement. Part II showed how this system constrains presidential power through multiple strands of law. To borrow from a broader conversation on presidential power, the final result is one of “power and constraint” rather than the “executive unbound.”143 Importantly, some of these constraints reflect forms of democratic accountability. These forms of oversight are less obvious, less demanding, and less supermajoritarian than the need to get the affirmative approval of either the Senate or Congress. They nonetheless are meaningful and make it highly likely that international commitments that bypass specific legislative approval will do more than simply reflect presidential preferences.

This issue of democratic accountability is particularly salient for important international commitments that will be implemented through US domestic law. The Iran deal and the Paris Agreement are examples of such commitments made by Obama. US commitments under the Iran deal are being implemented by the president using authority delegated to him by preexisting congressional laws that allow him to waive sanctions against Iran if he deems it in the national interest.144 At the Paris negotiations, US negotiators anticipated that US commitments under the Paris Agreement would be implemented through the EPA’s preexisting authority under the Clean Air Act.145 These commitments thus deal with matters that fall under the purview of Congress, as opposed to matters of recognition, security, and diplomacy with regard to which the president is understood to have considerable independent constitutional powers. Yet the Congresses that passed these preexisting laws did not expressly authorize them to be used as implementing authority for international commitments. Given the absence of express authorization, should the executive branch be able to infer that it can harness these laws to its own international objectives?

Two reasons suggest that the answer should generally be yes. First, in relying on these laws to implement international commitments, the Obama administration was furthering the congressional purposes underlying these laws. In the case of the Iran sanctions, Congress deliberately entrusted the president with broad discretion to lift them in order to advance US national interests. Allowing the president to use his delegated powers as negotiating leverage with Iran seems like a patently obvious way to use this discretion.146 In the case of the Clean Air Act, the purposive link is at a higher level of generality. The Congress that passed the Clean Air Act in the 1970s did not have climate change clearly in mind, let alone international coordination in relation to it. But its broader purpose of protecting public health and welfare through the regulation of air pollution is advanced if the United States can persuade other countries to reduce their own air pollution, because air pollution is a transborder problem (especially for climate change).147 To the extent that the Clean Air Act also sought to balance economic concerns, its use as negotiating leverage to get other countries to act with respect to climate change likewise promotes this interest—the more other countries are addressing climate change, the more the playing field is leveled in terms of effects on economic competition. There are doubtless some issues in which international cooperation would not enhance the underlying congressional purposes,148 but in general such cooperation seems likely to do so.

A second reason why the executive branch should be able to harness preexisting laws to implement international commitments that it makes without specific legislative approval is that this process naturally brings the interests of Congress into the negotiation and implementation of an international commitment. If it intends to rely on preexisting legislation for implementing authority, the executive branch can agree only to terms with respect to implementation that are within the bounds of what the Congress that passed this legislation already authorized as a matter of domestic law. Similarly, in implementing the international commitment domestically, the executive branch must accept whatever tools of accountability are built into the law it is using for implementation authority, including judicial review. The executive branch also has some degree of accountability to the current Congress, although far less than if it needed to get that Congress to vote affirmatively in favor of the commitment. At a minimum, the executive branch must have enough supporters in Congress to prevent the passage of a veto-proof law that strips it of its ability to rely on the preexisting legislation, whether by modifying that legislation or providing that it cannot be the basis for implementation.149 The executive branch will also be subject to whatever soft-law tools the current Congress deploys.150 The more the preexisting law gives implementing authority to administrative agencies rather than to the president, the greater Congress’s abilities will be to deploy these tools.

A distinct but related ground for concern about the democratic accountability of international commitments made without specific legislative approval goes to whether they would unduly thwart changes in democratic preferences over time. If the president makes an international commitment whose terms extend beyond his or her time in office, then the next president may feel pressured to honor this commitment despite not approving of it.151 Professor Ramsey, for example, argues that nonbinding political commitment cannot “constrain future Presidents (even informally)” without raising constitutional problems,152 and Professor Goldsmith notes that “the President can effectively change reliance interests through his delegated authorities in ways that are credible and sticky because they are hard for a future president to unwind.”153 For international agreements that do not expressly allow for withdrawal or that allow for withdrawal only after the passage of multiple years, the issue is particularly significant because the future president would have to choose between abiding by the agreement for an extended period of time or failing to comply with an international legal obligation.154

This concern is legitimate but adequately met by the current system. To begin with, the administration that makes an international commitment should foresee the risk that a new administration will have different preferences. While it has incentives to try to nudge this administration toward continuing its approach, it also must recognize that the more partisan a commitment it makes, the greater the odds are that this commitment will be repudiated. It must also recognize that failing to leave the next administration with an internationally lawful option of exit could incentivize that administration to pursue violations of international law in ways that could damage the reputation of the United States. This is one reason that using nonbinding international commitments will be appealing to an administration, as it provides a lawful if diplomatically awkward justification for noncompliance and means of exit.

For international agreements, the legal availability of withdrawal is more complicated. Most international agreements contain withdrawal clauses that the future president could choose to invoke.155 As a matter of administrative procedure, the C-175 Procedure makes the proposed duration of an agreement a factor that is to be considered in determining whether or not to seek specific approval from the Senate or Congress.156 The rare international agreements that lack withdrawal procedures—such as major human-rights instruments—go to the Senate for approval as treaties as a matter of constitutional custom and perhaps constitutional obligation, thus ensuring strong supramajoritarian review.157 While many international agreements provide for withdrawal upon a year or less of notice, some make withdrawal permissible beginning only after several years have elapsed since the agreement’s entry into force. The Minamata Convention, for example, provides that a nation can give one year’s notice of withdrawal at any point after the initial three years following its entry into force.158 The Paris Agreement has a similar provision, except that it also provides a backdoor route for withdrawal after only a single year.159 Such agreements thus left a lawful exit route for the next president, but one that could not be invoked for a while.

Because the administration that makes an international commitment has reasons not to lock its successor in too thoroughly, the future president will typically have a lawful exit option available at some point during his or her term. In addition, regardless of international legal considerations, the next president’s own democratically elected mandate may empower him or her to pursue exit or effective noncompliance. As a candidate, President Trump promised with respect to the Iran deal that he would “be so tough on [Iran] and ultimately that deal will be broken unless they behave better than they’ve ever behaved in their lives, which is probably unlikely.”160 Similarly, on climate, Trump promised during his campaign that “[w]e’re going to cancel the Paris Climate Agreement and stop all payments of U.S. tax dollars to U.N. global warming programs,” as well as to “rescind all the job-destroying Obama executive actions, including the Climate Action Plan.”161 It remains to be seen just how much Trump will now follow through on these assertions, but he is currently pursuing them with vigor. There is no legal impediment to exiting the Iran deal, and Trump is reported to be interested in exit.162 While Trump cannot “cancel” the Paris Agreement immediately as a matter of international law, he has already started rolling back the Clean Action Plan and has announced his intention to withdraw the United States from the Paris Agreement (and not to comply with it in the meantime).163 These examples illustrate how future presidents can roll back the international commitments of their predecessors.

3. Using uncertainty constructively.

There is a great deal of uncertainty built into the current system with respect to constitutional law. The line that the executive branch always needs some kind of preexisting or subsequent action from the Senate or Congress in order for the terms of international commitments to be implemented through domestic law is only a rough one—in the claims-settlement context, for example, the Supreme Court has held that a sole executive agreement does in fact constitute domestic law.164 Similarly, while the president has very broad constitutional authority to make nonbinding commitments, the extent to which he or she can make international agreements without the specific approval of the Senate or Congress remains deeply contested.165

This uncertainty gives the executive branch considerable flexibility and probably promotes the long-term expansion of the president’s constitutional authority through historical practice. Yet at the same time it has a self-policing effect for whatever international commitment is presently at issue because it raises the likelihood of resistance outside the executive branch (and possibly within it, too). The more the executive branch seeks to make a controversial international commitment without the specific approval of the legislature, the more it should anticipate pushback not just with respect to the merits of the commitment but also with respect to the process by which it was made. This in turn should encourage the executive branch to be cautious. Perhaps the Iran deal could constitutionally have been made as an agreement that was binding under international law.166 Perhaps the Paris Agreement could have included binding emissions reduction targets, notwithstanding the positions taken by the executive branch and the Senate Foreign Relations Committee years earlier during the approval of the UN Framework Convention on Climate Change (UNFCCC).167 In opting for the less dramatic position, the executive branch headed off stronger grounds for constitutional contestation from within the executive branch, from Congress, from the public, and potentially from the courts. As long as the courts do not give their stamp of approval to international commitments that bypass the Senate or Congress, this self-policing instinct should remain. This in turn leaves the executive branch likely to continue to go to the Senate or Congress not just for international commitments that require changes in domestic law for their implementation or for ones for which legislative buy-in is especially important on the merits, but also for international agreements that have exceptionally long time horizons or for which there is a particularly strong tradition of obtaining specific approval.

Of course, the executive branch will sometimes choose to provoke constitutional controversy, either out of a deliberate desire to set a precedent or because its international objectives lead it to take constitutionally controversial positions. When this happens, however, the executive branch may be most vulnerable to resistance when it is also least constrained at the international legal level or the administrative-law level. ACTA provides a good example. Opponents to this intellectual property agreement argued that the executive branch could not constitutionally join the United States to it, even assuming all the requirements for its domestic implementation were already provided for under preexisting US law.168 Interwoven with these arguments, however, were complaints about the secrecy of the process by which ACTA was negotiated internationally,169 about its failure to go through the C-175 Procedure in the usual way, and about the failure of the executive branch to provide a consistent and detailed defense of its intent to bypass the Senate and Congress.170 Public outcry against the secrecy of ACTA negotiations also occurred in Europe, and the European Parliament declined to ratify it.171 Whether because of this decision in Europe or because of domestic pressure, the Obama administration did not ratify ACTA, despite having asserted its power to do so.

As this example suggests, the very existence of legal uncertainty itself acts as a further check to the current system. This uncertainty both can act as a direct check on executive branch overreaching and can empower the opponents of international commitments in ways that further robust political debate and the use of soft tools of democratic accountability. The more vulnerable an international commitment is from the perspective of international legal checks and administrative-law checks, the more opponents of the international commitment may be able to draw persuasively on constitutional concerns in the public arena and perhaps also in the courts.

B. Structural and Doctrinal Implications

Understanding the current system to be the product of three strands of law has implications for ongoing structural and doctrinal debates within the field of foreign relations law. This Article has already shown descriptively how this system incorporates checks from all three strands and has offered a qualified normative defense of this system. These descriptive and normative claims in turn have implications for three related debates: first, whether international agreements that do not receive specific legislative approval need to be subject to more accountability mechanisms; second, whether the Senate and Congress need to put in place easier processes for providing legislative approval; and third, how strong the president’s constitutional powers with respect to negotiation and domestic implementation should be understood to be.

1. Balancing accountability and flexibility.

As this Article has shown, international law and administrative law provide accountability mechanisms for international commitments that executive branch actors make without getting specific legislative approval. For major international commitments, international negotiations and subsequent compliance procedures often provide for considerable transparency and opportunities for broader participation. On the domestic front, the Case-Zablocki Act, the C-175 Procedure, any laws specific to agencies involved in international negotiations, and now Executive Order 13609 all further accountability, as do the limits that come with the implementation process and the tools of soft power that members of Congress can invoke.

Professor Hathaway has argued that existing measures are not enough and that there needs to be stronger oversight of all international agreements that do not receive the specific approval of the Senate or Congress.172 She suggests multiple reforms for all such agreements, including that they be provided to Congress before they enter into force, that the State Department be public and specific about the constitutional justification for the pathway that it uses for each agreement, and that Congress pass an APA-like statute providing for public opportunity for notice and comment during the negotiation of international agreements.173

Hathaway’s proposals amount to effortful remedies to a mostly nonexistent problem. In her view, the entire system needs reform not because it is producing bad agreements, but because the president has too much power and too little oversight. This position in turn rests on her perspective that the president acts “alone” and “almost entirely unfettered.”174 Hathaway focuses on the checks that exist from a constitutional perspective and thus does not take into account the existing constraints that come with the international legal process or with the involvement of agencies, who are in turn accountable not only to the president but also to Congress, the courts, and their own constituencies. The steps she recommends would come at a real cost, in both money and time, to the executive branch for the hundreds of international agreements (mostly minor ones) that it makes every year. This in turn would probably have the effect of incentivizing executive branch actors more toward nonbinding commitments, which are not subject to Hathaway’s proposals. Of course, this could be addressed by trying to tighten oversight with regard to nonbinding commitments—but that would require more time, cost, and hassle, and thus have a deterrent effect on international cooperation.

Retail fixes offer a better way of dealing with accountability concerns than do wholesale ones. The type and strengths of constraints vary across specific contexts at both the international and domestic levels. Negotiations of new international agreements that are occurring under UN auspices will have greater transparency and opportunities for participation than will regional trade agreements. International commitments that are to be implemented by authority previously delegated to the president will have fewer administrative-law safeguards than ones to be implemented by authority delegated to agencies. Some administrative agencies involved in international cooperation operate with less congressional scrutiny than do others.175 Context will determine how good the balance is in any particular situation, and calls for reform should therefore be contextually grounded.

One retail fix that the executive branch could undertake on its own or with the encouragement of Congress would be to make public its legal reasoning on the constitutionality of especially significant international agreements for which it does not obtain specific legislative approval. The Obama administration was public with regard to its legal analysis on many sensitive national security issues. Through speeches, testimony, and the release of legal memoranda, it explained its legal positions with respect to issues as sensitive as the use of force in Libya176 and the targeted killing abroad of American citizens.177 Yet when asked for the legal reasoning behind its claim that the United States could join ACTA without the specific approval of the Senate or Congress, the administration gave brief and somewhat conflicting responses.178 By making its legal reasoning clearer for major international agreements, the administration would be providing another venue for administrative-law principles of transparency and accountability, even as over time such transparency would probably strengthen its constitutional arguments as a matter of historical practice.

2. Making legislative approval easier.

The current system will be with us for the foreseeable future. This is true not only with respect to the many international commitments that do not receive specific approval from the Senate or Congress, but also for the pathways that do require such specific approval, including treaties and ex post congressional-executive agreements. The executive branch will continue to make treaties and ex post congressional-executive agreements when changes to domestic law are needed for implementation, when there is a particular desire for domestic support, when there is a need to signal to other nations the seriousness of the US commitment, and when, for certain types of agreements, long-standing practice perhaps amounting to constitutional law effectively requires legislative approval.

Right now, obtaining legislative approval for international agreements is a deeply unpredictable process. This is true not only with respect to whether the votes will be there to support a particular agreement, but also with respect to whether and when the agreement can get through all the procedural veto gates in order to receive a floor vote. Right now, most major international agreements that require specific legislative approval go to the Senate as treaties and can further require the passage of implementing legislation from Congress.179 It can take years or decades for these treaties to receive floor votes or for the passage of their implementing legislation.180 The only major category of international agreements that has gone to Congress for approval as ex post congressional-executive agreements in recent years is trade agreements.181 In practice, the ability of these agreements to receive a floor vote depends on whether or not there is preexisting “fast-track” legislation that guarantees a speedy vote for these agreements.182 Such legislation has existed on and off over the years; the present version applies for the next handful of years.183

In recent years, scholars have made numerous proposals for making the treaty process and the ex post congressional-executive agreement process more efficient. For treaties, the Senate’s advice-and-consent process could provide conditional prenegotiation approval184 or set up a fast-track process.185 The Senate could also take steps to reduce or eliminate the need for implementing legislation,186 or alternatively Congress could streamline the process for passing it.187 For ex post congressional-executive agreements, Congress could provide a long-term fast-track option for all agreements.188

The alternatives to treaties described in this Article offer reasons why the Senate and Congress should be willing to consider these moves. The harder it is for the executive branch to get specific approval from the Senate and Congress—and the more uncertainty there is about this process—the more incentives the executive branch has to try to structure an international commitment in ways that allow it to bypass specific approval entirely. As Goldsmith wrote toward the end of the Obama administration, “Senate, Congress, wake up and pay attention!”189 If the Senate or Congress wants more of a role, then they need to bring either carrots or sticks to bear—and, because of the president’s veto power, carrots are more likely to succeed as legislation than sticks.

Easier legislative approval would have advantages not only for the institutional power of the Senate and Congress, but also for US interests more generally. It would offer the executive branch more flexibility in international negotiations, as it would have more reason to think itself able to obtain congressional approval and thus more reason to pursue terms that would require changes to US domestic law. It would also make it easier to negotiate agreements aimed at changing the superstructure of international law. The major international agreements that define the international legal system may need updating over time, and legislative approval would be important to such endeavors.

3. Retaining checks in negotiation and implementation.

This Article has defended the diminishment of specific legislative approval for international commitments largely on the grounds that alternative constraints exist with respect to the negotiation and implementation of these commitments. One further set of implications from this Article, therefore, is that especially strong constitutional claims of presidential power in these domains should be resisted.

With regard to negotiations, the executive branch has long favored the sweeping language from United States v Curtiss-Wright Export Corp190 that “the President alone has the power to speak or listen as a representative of the nation . . . he alone negotiates.”191 In a soft form, this is undeniably correct: the president and his or her agents act internationally for the United States. But this language should not be taken to mean that members of Congress, independent regulatory agencies, and US states cannot constitutionally share their own positions with respect to international negotiations and even directly with negotiators from other countries. The letter authored by Senator Cotton and many other Republican senators to the “Leaders of the Islamic Republic of Iran”192 may have been unprecedented and an exercise of poor judgment, but, contra to the view of then–Secretary of State John Kerry, it should not be considered “unconstitutional.”193 Instead, it should be viewed as an example of how the institutional powers exercised by members of Congress may change over time as counterweights to developments in the president’s power to choose. The Supreme Court has recently signaled an interest in rolling back the strong language from Curtiss-Wright.194 Such an approach would promote the structural checks and balances discussed in this Article.

Turning to implementation, as discussed earlier, the president typically needs the Senate or Congress in order to implement international commitments that require the alteration of domestic law. But this line is a rough one. In the claims-settlement context, for example, the Supreme Court has held that sole executive agreements—and even presidential policy—can preempt state law.195 In addition, the Supreme Court has recently held that the president has the exclusive power of recognizing foreign nations,196 and it therefore seems logical that decisions and international commitments made by the president with regard to recognition should have effect as domestic law.197 Outside of these contexts, though, the analysis in this Article suggests that courts should energetically resist attempts to treat international commitments that do not have the clear approval of Congress as domestic law or as preempting state law. In addition, given the useful structural role that state and local governments can play as a counterweight to presidential power, courts should be wary of strong attempts to box these actors out of engagement with foreign affairs.198

With respect to international and administrative law, however, courts need not be as concerned about giving some leeway to the executive branch in implementation. Courts typically defer to the executive branch with respect to the interpretation both of international law and administrative statutes. In practice, however, courts have considerable flexibility in terms of how much deference they actually give, and this flexibility gives them another tool of control over executive branch overreaching.199 With the move toward international commitments that do not receive specific legislative approval but will be implemented through preexisting administrative law, one interesting question is whether courts should pay any attention to the international commitment in reviewing agency actions taken under the authority of the preexisting statute. While courts should be wary of applying a strong version of the Charming Betsy canon200 in these contexts, a touch of added deference seems appropriate.201

IV. The Paris Agreement as a Case Study

In his last year in office, President Obama joined the United States to the Paris Agreement on climate without the explicit approval of either the Senate or Congress.202 This decision showcases the dynamics described in this Article. On the one hand, it demonstrates how constitutional constraints on the approval of international agreements have eroded in modern times. Indeed, it was called “outrageous,” “unlawful,” and “a clear example of the executive overreach in the area of foreign affairs.”203 On the other hand, when the negotiation and implementation of the Paris Agreement are taken into account, it is apparent that Obama acted under powerful constraints arising from the conjunction of international law, constitutional law, and administrative law. As subsequent events are already revealing, his actions remain deeply vulnerable to resistance and reversal.

Overall, the Paris Agreement demonstrates just how important the availability of multiple pathways to making international commitments has become to the functioning of foreign relations law. For those convinced of the urgent need for action with respect to climate change, it also illustrates the value of this diversity of pathways. This Part first briefly sets the Paris Agreement in the broader context of US participation in climate negotiations. It then describes the web of checks on Obama’s uses of power that have arisen or will arise with respect to its negotiation, approval, and implementation.

A. From Rio to Paris

In 1992, President George H.W. Bush signed the UNFCCC at the Earth Summit in Rio de Janeiro.204 Finalized earlier that year, the UNFCCC had both substantive and procedural elements. Substantively, it committed state parties to taking steps to combat climate change but left vague many aspects relating to the content of these steps.205 Procedurally, it set up an institutional and legal framework for the conduct of future negotiations about climate change, including through the establishment of a Conference of Parties that would hold annual meetings.206

Bush then submitted the UNFCCC to the Senate for advice and consent.207 This was the easy and obvious choice. As a matter of politics, obtaining the Senate’s approval was straightforward. Both Democrats and Republicans wanted it done before the 1992 elections, and the treaty received the Senate’s advice and consent with astonishing speed—within a mere month of its submission.208 As a matter of law, treating the UNFCCC as a treaty put it on unquestionable constitutional footing and was consistent with how the recent Vienna Convention for the Protection of the Ozone Layer and related protocols had been treated.209

During the advice-and-consent process, actors in both the executive branch and the Senate considered what domestic pathway would be appropriate for future protocols to the UNFCCC. The administration stated to the Senate Foreign Relations Committee that “we would expect” a future “protocol containing targets and timetables [to] be submitted to the Senate” for advice and consent.210 The Senate Foreign Relations Committee stated in its committee report that a “decision . . . to adopt targets and timetables would have to be submitted to the Senate for its advice and consent.”211

Between 1992 and 2009, however, two important developments dramatically changed the incentives for a president concerned about climate change to return to the Senate with a substantive treaty. First, what had been merely predictable became painstakingly obvious: getting domestic approval for a treaty that provided for strong substantive action on climate change would be about as easy as getting a camel through the eye of the proverbial needle. Two-thirds of the Senate would never back such a treaty under a Democratic president, and no Republican president would put such a treaty forward. During the Clinton administration, this was manifested by the failure of the 1997 Kyoto Protocol, which had no chance in the Senate.212 In the George W. Bush era, this was made clear by the administration’s underwhelming interest in addressing climate change.213

The second important change was the Supreme Court’s 5–4 decision in Massachusetts v Environmental Protection Agency.214 The Court held that, contrary to the views taken by the Bush II administration, the Clean Air Act permitted and indeed effectively obligated the EPA to regulate greenhouse gas emissions.215 Although this decision was purely about the interpretation of the Clean Air Act, it had major implications for international commitments. Most importantly, it meant that the EPA could now implement an international commitment that required domestic action, as long as this domestic action followed proper procedures and lay within the substantive scope of its now-clarified authority under the Clean Air Act.

Obama thus came into office with a pathway for implementing a negotiated climate commitment domestically in a way that would not require congressional legislation. Consistent with the long time horizons associated with the negotiation of major international agreements, only when the Paris Agreement was reached in December 2015 was this pathway fully realized.216

B. Paris and Presidential Power

The Paris Agreement reveals both the opportunities and limits of the president’s power to choose. Obama described it as “historic,” as “ambitious,” as “establish[ing] the enduring framework the world needs to solve the climate crisis,” and overall as something that “can be a turning point for the world.”217 So much was acknowledged to be at stake—yet Obama chose not to go to the Senate or to Congress for approval. More than that, he bypassed the Senate and Congress precisely because the issue was so important and the odds of legislative approval were so low. Focusing just on the approval process, this seems like a quintessential example of unchecked presidential power. But a close look at the international, constitutional, and administrative legal issues underlying the negotiation, approval, and implementation of the Paris Agreement reveals a very different story. In actuality, the Obama administration was so hemmed in on every front that the Paris Agreement amounted to an improbable and brilliant success. The acts of the Trump administration to date suggest that this success may prove to be a fleeting one.

1. Negotiation.

In negotiating the Paris Agreement, the Obama administration had to contend not only with the constraints accompanying the negotiating process but also with those that it foresaw arising with respect to approval and implementation. This is true for all international commitments (as Part II described), but what is unusual about the Paris Agreement is just how formidable all of these constraints were. As a matter of international law, the negotiating process was designed to be incredibly cumbersome. The negotiations took place as part of the UNFCCC Conference of the Parties, and, under the applicable rules of procedure, the nearly two hundred negotiating states need to reach consensus in order to make an agreement.218 This challenging requirement was further complicated by all the differences between nations that climate change brings out—differences that go to who bears the greatest historic responsibility, who has the most capacity to reduce greenhouse gas emissions and/or support carbon sinks, and who is most steeply suffering the effects of climate change. All these factors help explain the painfully slow progression of negotiations under the UNFCCC process.

By the time of the Paris Agreement, the accretion of prior negotiating decisions made through the UNFCCC process limited what could be done at Paris. One of these limiting prior decisions involved the legal form which could be used for a commitment. In 2011, the Conference of the Parties had agreed in their annual negotiations that the future commitment would be “a protocol, another legal instrument or an agreed outcome with legal force under the Convention applicable to all Parties.”219 The Obama administration agreed to this formulation in return for a major substantive concession,220 but in return it had to forgo the domestic constitutional advantages that would have come with having the future instrument be nonbinding under international law. In other words, by the time the Conference at Paris came around, the president’s negotiators were committed internationally to a process that gave up his easiest domestic constitutional path to approval.221

Domestic-law considerations also tightly channeled the scope of negotiating possibilities for the Obama administration. Thinking ahead to approval, the administration knew it needed an agreement that it could join without the Senate or Congress—first, as a matter of constitutional law; second, as a matter of the C-175 process; and third, despite the legislative history from the Senate’s advice and consent to the UNFCCC. Thinking ahead to implementation, the administration knew it needed an agreement whose terms could at least theoretically be implemented through a combination of the president’s independent constitutional powers and preexisting delegations by Congress to administrative agencies and US states, most notably in the Clean Air Act.

As a matter of process, this meant that the Obama administration did not even try to leave the Paris negotiations solely in the hands of traditional diplomats. The EPA played a particularly prominent role in the negotiations—Administrator Gina McCarthy spent the entire week in Paris222 —but many other agencies were represented as well.223 Nor were actors outside of the executive branch shy about making their voices heard. Congressional committees and subcommittees held hearings in the weeks leading up to Paris224 —including one hearing convened by Senator Ted Cruz—and members of both the House and the Senate made appearances at Paris. Representatives from nongovernmental organizations in both industry and environment similarly turned up in force at Paris.225 So did a large contingent from the government of California; Governor Jerry Brown made multiple international commitments just before and during the conference.226 These external forces did not alter the executive branch’s formal control over US negotiations, but they did make the overall process more participatory and more scrutinized.

As a matter of substance, the domestic constraints mentioned above tied the Obama administration’s negotiating hands with respect to two central issues under negotiation: first, the reduction of greenhouse gas emissions and, second, money. On the issue of emissions reductions, the Obama administration needed to avoid making an internationally legally binding commitment to reduce emissions and, especially, to avoid such a commitment with respect to specified targets. A legally binding commitment to reduce emissions would have put the Paris Agreement on thinner ice as a matter of constitutional law, though still defensible and also consistent with the approach taken in some earlier environmental agreements.227 Moreover, it would have been in tension with the legislative history associated with the Senate’s advice and consent to the UNFCCC, when the executive branch had signaled that it would submit any future protocol with targets and timetables as a treaty.228 Combined, these two factors would have made it difficult and perhaps impossible for the State Department to conclude through the C-175 process that no subsequent approval from the Senate or Congress was needed.

The Obama administration resolved this challenge by insisting that the Paris Agreement not contain any legally binding commitments to reduce emissions. Instead, the Agreement requires each nation to set a nationally determined contribution (NDC) that it “intends to achieve.”229 Nations are legally obligated to “prepare, communicate and maintain” these NDCs, but they are not legally obligated to meet them.230 Structuring the Paris Agreement this way required expert legal work, with careful distinctions drawn between language meant to signify internationally legally binding commitments (most notably “shall”) and language meant to signify nonbinding aspects of the agreement (for example, “should,” “aim,” and “are encouraged”). The importance of these distinctions to presidential negotiators was highlighted by a dramatic last-minute change. The initial final text of Article 4(4) stated that: “Developed country Parties shall continue taking the lead by undertaking economy-wide absolute emission reduction targets.”231 Horrified and believing that this language was introduced in error, US negotiators insisted that the word “should” be substituted for “shall.” Secretary Kerry said in essence that, “Either it changes, or President Obama and the United States will not be able to support this agreement.”232 In the final moments, the change was made, nominally as a correction to a drafting error.233

Another key domestic constraint related to money. Part of the core climate deal is for developing countries to receive very large amounts of financial support for their efforts to combat climate change. In the 2009 Copenhagen negotiations, the commitment that was reached set specified levels of support that developed countries would mobilize for developing countries—levels which were to rise to the overall sum of $100 billion a year by 2020.234 At Paris, the developing countries wanted this commitment made in a legally binding way. While the Obama administration could be sure of its ability to provide some seed money to developing countries,235 it could not be sure a successor administration would do so, and it certainly could not commit to large, continuing contributions at a level that would require congressional appropriations. The Obama administration thus had strong reasons for wanting to avoid any international legal obligation to commit meaningful sums of money to developing countries. Looking to approval, such an obligation would strengthen the constitutional concerns about bypassing the Senate and Congress;236 and looking to implementation, it would be close to impossible to meet such an obligation without affirmative support from Congress. In the end, negotiators once again struck a delicate balance. The Paris Agreement itself states that “[d]eveloped country Parties shall provide financial resources to assist developing country Parties,”237 but it does not specify the amounts for which developed countries (let alone individual countries among them) are responsible. Instead, the $100-billion-a-year figure from Copenhagen was reiterated only in a separate conference decision that accompanied the Paris Agreement.238

As all this suggests, the Obama administration had very limited space for maneuvering in the Paris negotiations. Its hands were tied not only by political constraints, but also by legal ones. From international law, it had to contend with layers and layers of international process and with the substantive limits locked in by previous negotiations. On the domestic side, its constraints did not come from asserted claims about what the Senate or Congress would approve, as it was clear that they were not going to end up approving the Paris Agreement. Instead, these domestic constraints derived from the limited scope of what could be done in the absence of the Senate or Congress. That these constraints were strong and credible is shown by how much the Paris Agreement was tailored to accommodate them.

2. Approval.

As the discussion above shows, concerns about domestic approval folded into the negotiating process. Although the international legal backdrop prevented the climate commitment from being entirely nonbinding, the Obama administration succeeded in negotiating an agreement that it considered it could join without subsequent approval from the Senate as a treaty or Congress as an ex post congressional-executive agreement. It did so by insisting that the aspects of the agreement most vulnerable to constitutional concern be made nonbinding or nonspecific. The end result is an agreement that illustrates Professor Koh’s observation that “we are now moving to a whole host of less crystalline, more nuanced forms of international legal engagement and cooperation that do not fall neatly within any of the[ ] three pigeonholes” of treaty, congressional-executive agreement, and sole executive agreement.239 Obama made the Paris Agreement mainly on his own constitutional authority, but he was buttressed in doing so by the fact that this Agreement furthers both the preexisting UNFCCC and the goals underlying the Clean Air Act. The expert lawyering that enabled the Obama administration to walk this line sparked admiration from former Bush administration lawyers John Bellinger and Goldsmith.240

The Paris Agreement illustrates the broader overall shift described in this Article from constitutional checks to administrative ones. Despite its careful crafting to minimize constitutional concerns, it will probably be viewed broadly by future administrations as a historical practice justifying sweeping constitutional powers of the president to make sole executive agreements when the commitments made in these agreements can be implemented through preexisting domestic law.241 Yet it may also lead in the long run to changes to the C-175 Procedure in ways that make it less opaque and perhaps more regularized. Intriguingly, at a 2016 congressional hearing involving the Paris Agreement, one witness who was asserting that the Paris Agreement should go to the Senate based this claim on the C-175 Procedure factors far more than on the Constitution.242 To the interest of several present representatives, he argued for more transparency to the C-175 Procedure.243 The more the president is unconstrained constitutionally with regard to approval, the more such administrative-law principles may come to the fore.

3. Implementation.

The decision to forgo the unachievable approval of the Senate or Congress left Obama in a fragile position with regard to implementation. As a matter of international law, he locked the next administration into the Paris Agreement for four years from its entry into force, unless the Trump administration wishes to invoke a one-year route that is available if it also withdraws from the UNFCCC.244 As a matter of domestic law, however, Obama’s ability to have the Paris Agreement implemented depended in the immediate future on the EPA and the federal courts—and in the longer term on the next administration, on future Congresses, and on the international legal process. In short, as in the negotiations, the President’s powers were curtailed on all sides.

Most immediately, in relying on the Clean Air Act for implementation of the NDC submitted by the United States, Obama necessarily accepted the process and limits that come with that Act. He thus became dependent on the EPA, which in turn must act within the constraints of the Clean Air Act and the APA.245 In June 2014, the EPA provided for notice and comment a proposed rule known as the Clean Power Plan. This proposed rule sought to reduce emissions from existing power plants and gave considerable flexibility to state agencies with regard to implementation.246 After the receipt of over four million comments, the EPA issued its final rule in the fall of 2015.247 This rule was arguably crucial to the ability of the United States to meet its NDC.248 But over twenty states and countless other parties challenged many aspects of the rule, and the Supreme Court stayed its implementation during this litigation by a 5–4 vote just a few days before the death of Justice Antonin Scalia in February 2016.249

This challenge to implementation would have awaited even a successor to Obama who was friendly to the Paris Agreement. So too would other challenges. At the international level, the Paris Agreement left many issues to the complex and ongoing negotiating process, including aspects of such crucial matters as increased emissions reduction, compliance, and funding. At the domestic level, congressional support would have been needed in the long run for funding purposes, so as to support developing countries in pursuing emission reductions.

The first year of the Trump administration has illustrated yet another check on Obama’s power with respect to the Paris Agreement: the ability of a future president to roll back the commitments of his or her predecessor. Shortly after taking office, President Trump instructed his EPA administrator to review the Clean Power Plan and “if appropriate . . . as soon as practicable suspend, revise, or rescind” it.250 He then announced that the United States would withdraw from the Paris Agreement, presumably pursuant to its withdrawal provisions, and would in the meantime fail to implement it.251 By contrast, while Trump has expressed skepticism of some major international commitments that received the explicit approval of the Senate or Congress (such as NAFTA and NATO), to date he has not triggered withdrawal from these commitments.252

Trump’s decision to withdraw from the Paris Agreement illustrates to other nations just how fragile international commitments made by the executive branch without explicit congressional buy-in can be. Assuming that Trump’s decision to withdraw is indeed carried out, one might wonder whether it will cause other nations to be unwilling to make such commitments with the United States in the future. But while it should increase their skepticism, it is unlikely to turn them off such commitments entirely. For one thing, a commitment that receives the explicit approval of Congress or the Senate will often be impossible to obtain, and other nations frequently will prefer an executive branch commitment, with all its limits, to no commitment at all. Furthermore, even withdrawn executive branch commitments can leave footprints that are followed by other institutional actors. The courts will have power to review the Trump administration’s efforts to repeal the Clean Power Plan. And already many US cities and some states have signaled their intention to reduce emissions in compliance with the Paris Agreement despite Trump’s proposed withdrawal.253 For just as other institutional actors can serve as checks on executive power in the making of international commitments, so too can they check the effects of withdrawal.

* * *

Overall, the Paris Agreement demonstrates both the strength and the fragility of the president’s power to choose. In order to bypass the Senate or Congress at the approval stage, Obama needed the nearly two hundred other countries at Paris to agree to core US negotiating demands and for the EPA to conclude it could and would issue various emissions-reducing regulations under its preexisting statutory authority. Yet the success of the Paris Agreement has proved fragile at best. For the United States to meet its commitments, Obama knew or should have known that all of the following would be needed: the courts to uphold the EPA’s regulations aimed at climate mitigation, the next administration to retain existing efforts aimed at the implementation of the Paris Agreement, and Congress to eventually provide appropriations that further mitigation efforts outside the United States.254 That Obama chose to accept the full force of these constraints rather than aim for specific legislative approval of the Paris Agreement is a signal of just how impossible it would have been to get this approval. For those who believe that there is no substitute for specific approval from the Senate or Congress, the existence of an alternative set of constraints will be no consolation. But those who are concerned about both unconstrained presidential power and undue legislative gridlock should have reason to value the current system.

Conclusion

US foreign policy depends heavily on international commitments made without the specific approval of the Senate or Congress. This has long been true, but high-profile commitments like the Iran deal and the Paris Agreement made this salient and the subject of renewed scrutiny. The Obama administration’s bold decisions to bypass the Senate and Congress appropriately invite inquiry as to whether the president now has too much power in this domain. This concern is an especially important one at the start of a new administration whose actions on other fronts have already triggered concerns about executive overreaching.

As this Article has shown, important checks on presidential power remain with respect to the making of international commitments. But we will not see most of them if we look only at constitutional law and only at the process of domestic approval for commitments. Instead, the checks are spread across strands of law—constitutional, international, and administrative—and across the negotiation, approval, and implementation of commitments. Institutionally, they exist not only through the classic constitutional actors on Capitol Hill, but also via international organizations, administrative agencies, and even sometimes US states. The resulting web of checks is not perfect. In some places it is too weak and in other places it is too strong. But overall it does a good job of balancing the imperatives of US international engagement with the need for constraints on presidential power.

Although this Article has focused on the making of international commitments, the approach taken here has implications for foreign relations law more generally. Thinking about checks and balances only from the perspective of constitutional law is like looking for the keys under the lamppost. It is a natural choice but not always the right one. The international landscape is increasingly shaped by legal order rather than anarchy, and administrative law more and more affects how the executive branch engages internationally. These changes matter for how power is allocated and constrained in the practice of US foreign relations law.

• 1George Washington, Farewell Address of Sept 17, 1796, in James D. Richardson, ed, 1 A Compilation of the Messages and Papers of the Presidents 205, 214 (Bureau of National Literature 1897).
• 2Because this Article focuses on how the United States participates in international commitments, all references to “international commitments,” “international agreements,” and “treaties” refer to those involving the United States (unless otherwise specified). I use “international commitments” broadly to cover formalized exchanges of promises by the United States and one or more other nations, regardless of whether these promises are binding under international law. I use “international agreements” more narrowly to refer to instruments that contain commitments that are binding as a matter of international law. I use “treaties” still more narrowly to refer to agreements that go through the process set out in the Treaty Clause.
• 3See Statement by President Donald J. Trump on Signing the “Countering America’s Adversaries through Sanctions Act” (White House Office of the Press Secretary, Aug 2, 2017), archived at http://perma.cc/Q9TG-P5HT (stating that “[a]s President, I can make far better deals with foreign countries than Congress”).
• 4US Const Art II, § 2, cl 2.
• 5See Senate Resolution of Advice and Consent to the New START Treaty with the Russian Federation, 111th Cong, 2d Sess, in 156 Cong Rec 23472 (Dec 22, 2010).
• 6Pub L No 111-203, 124 Stat 1376 (2010).
• 7See Jean Galbraith and David Zaring, Soft Law as Foreign Relations Law, 99 Cornell L Rev 735, 784–87 (2014).
• 8See Jack Goldsmith, Why Congress Is Effectively Powerless to Stop the Iran Deal (and Why the Answer Is Not the Iran Review Act) (Lawfare, July 20, 2015), archived at http://‌perma.cc/‌GB4U‌-SC5A.
• 9See Senior State Department Official on the Paris Agreement Signing Ceremony (US Department of State, Apr 20, 2016), archived at http://‌perma.cc/‌VK2Y‌-JWMD; Juliet Eilperin, Obama Hails ‘Historic’ Ratification of Paris Climate Agreement (Wash Post, Oct 5, 2016), archived at http://‌perma.cc/‌JKY8‌-FE2R.
• 10See Statement by the President on the Signing of the Trans-Pacific Partnership (White House Office of the Press Secretary, Feb 3, 2016), archived at http://‌perma.cc/‌J5BY‌-34N5; Presidential Memorandum regarding Withdrawal of the United States from the Trans-Pacific Partnership Negotiations and Agreement (White House Office of the Press Secretary, Jan 23, 2017), archived at http://‌perma.cc/‌6FCM‌-A36J.
• 11For coverage in leading treatises, see, for example, Curtis A. Bradley, International Law in the US Legal System 31–95 (Oxford 2013); Louis Henkin, Foreign Affairs and the United States Constitution 175–224 (Oxford 2d ed 1996); Restatement (Third) of the Foreign Relations Law of the United States § 303 (1987). In these and other sources, there are several variations on the overall terminology—for example, sole executive agreements are sometimes called presidential agreements, and congressional-executive agreements and sole executive agreements are sometimes collectively referred to as executive agreements. An additional category, which scholars frequently note but tend to treat more briefly, is that of agreements entered into by the executive branch that are authorized by a preexisting treaty.
• 12Bradley, International Law at 75 (cited in note 11).
• 13Id.
• 14Harold Hongju Koh, Remarks: Twenty-First-Century International Lawmaking, 101 Georgetown L J 725, 726 (2013). For the origins of the categorization, see Part I.A.
• 15Id at 727, 732.
• 16See, for example, Oona A. Hathaway, Treaties’ End: The Past, Present, and Future of International Lawmaking in the United States, 117 Yale L J 1236, 1338–49 (2008); Bradford R. Clark, Domesticating Sole Executive Agreements, 93 Va L Rev 1573, 1578–1617, 1654–60 (2007); Peter J. Spiro, Treaties, Executive Agreements, and Constitutional Method, 79 Tex L Rev 961, 964–1009 (2001); Michael D. Ramsey, Executive Agreements and the (Non)Treaty Power, 77 NC L Rev 133, 160–83, 218–35 (1998); David M. Golove, Against Free-Form Formalism, 73 NYU L Rev 1791, 1798–1805 (1998); Laurence H. Tribe, Taking Text and Structure Seriously: Reflections on Free-Form Method in Constitutional Interpretation, 108 Harv L Rev 1221, 1228–35 (1995); Bruce Ackerman and David Golove, Is NAFTA Constitutional?, 108 Harv L Rev 799, 808–13 (1995).
• 17Executive Overreach in Foreign Affairs, Hearing before the Executive Overreach Task Force of the House Committee on the Judiciary, 114th Cong, 2d Sess 2 (2016) (“House Executive Overreach Task Force Hearing”) (statement of Rep King).
• 18Federalist 51 (Madison), in The Federalist 347, 347–48 (Wesleyan 1961) (Jacob E. Cooke, ed).
• 19Statement by the President on the Paris Climate Agreement (White House Office of the Press Secretary, Dec 12, 2015), archived at http://perma.cc/‌A75H‌-DL3M.
• 20Joby Warrick, How One Word Nearly Killed the Climate Deal (Wash Post, Dec 13, 2015), archived at http://‌perma.cc/‌2JBX‌-98DU.
• 21See Clean Air Act, Pub L No 88-206, 77 Stat 392 (1963), codified as amended at 42 USC § 7401 et seq; Clear Air Amendments of 1970, Pub L No 91-604, 84 Stat 1676; Clear Air Act Amendments of 1977, Pub L No 95-95, 91 Stat 685.
• 22Statement by President Trump on the Paris Climate Accord (White House Office of the Press Secretary, June 1, 2017), archived at http://‌perma.cc/‌6GZ7‌-GJXP. But see Madeline Conway, Trump: ‘Something Could Happen’ on Paris Agreement (Politico, July 13, 2017), archived at http://‌perma.cc/‌5STQ‌-6TYX (describing Trump’s statement in a press conference that “something could happen with respect to the Paris accord. We’ll see what happens. . . . And if it happens, that will be wonderful, and if it doesn’t, that will be OK, too”).
• 23Letter from John Hay to Henry Adams (Aug 5, 1899), quoted in W. Stull Holt, Treaties Defeated by the Senate: A Study of the Struggle between the President and the Senate over the Conduct of Foreign Relations 177 (Johns Hopkins 1933).
• 24For early articles analyzing historical practice, see generally John Bassett Moore, Treaties and Executive Agreements, 20 Polit Sci Q 385 (1905); James F. Barnett, International Agreements without the Advice and Consent of the Senate, 15 Yale L J 63 (1905).
• 25See generally Memorandum from the Solicitor for the Department of State to Senator Henry Cabot Lodge on International Executive Agreements Not Submitted to the Senate (Aug 23, 1922) (on file with author). Senator Henry Cabot Lodge was the chair of the Foreign Relations Committee and bore significant responsibility for the failure of the Treaty of Versailles in the Senate.
• 26Id at 1.
• 27Id.
• 28Id. The first thirty-one pages of the memorandum give examples of sole executive agreements in the claims-settlement context. See id at 2–31. Only the last two pages mention other examples, some of them modi vivendi, related to military affairs, peace, fisheries, and boundaries. Id at 32–33.
• 29See generally Myres S. McDougal and Asher Lans, Treaties and Congressional-Executive or Presidential Agreements: Interchangeable Instruments of National Policy: I, 54 Yale L J 181 (1945); Myres S. McDougal and Asher Lans, Treaties and Congressional-Executive or Presidential Agreements: Interchangeable Instruments of National Policy: II, 54 Yale L J 534 (1945). Other important contemporary scholars engaged with these constitutional questions as well, reaching a range of conclusions. See Ackerman and Golove, 108 Harv L Rev at 853–56, 866–73 (cited in note 16) (summarizing the debates). The UN Charter was ultimately approved as a treaty, but some other major agreements in the years after World War II were done as congressional-executive agreements. Id at 889–96.
• 30McDougal and Lans, 54 Yale L J at 187, 203–06 (cited in note 29) (also briefly noting the additional category of agreements based on authorizations in existing treaties).
• 31Id at 187.
• 32See Ackerman and Golove, 108 Harv L Rev at 868–69 (cited in note 16) (noting the use of this technique by defenders of interchangeability).
• 33McDougal and Lans, 54 Yale L J at 187 (cited in note 29) (arguing that the president has authority to make international agreements that are “within the scope of his own constitutional powers”).
• 34Id at 197 (“Whatever distinction there is between treaties and executive agreements must be found, not in the practices and doctrines of international law, but in our own unique constitutional law.”).
• 35Koh, 101 Georgetown L J at 726 (cited in note 14).
• 36See generally, for example, Hathaway, 117 Yale L J 1236 (cited in note 16); Spiro, 79 Tex L Rev 961 (cited in note 16); Golove, 73 NYU L Rev 1791 (cited in note 16); Tribe, 108 Harv L Rev 1221 (cited in note 16); Ackerman and Golove, 108 Harv L Rev 799 (cited in note 16).
• 37For congressional hearings addressing this question, see note 224. For recent articles, see generally, for example, Daniel Bodansky and Peter Spiro, Executive Agreements+, 49 Vand J Transnatl L 885 (2016); Jack Goldsmith, The Contributions of the Obama Administration to the Practice and Theory of International Law, 57 Harv Intl L J 455 (2016); David A. Wirth, The International and Domestic Law of Climate Change: A Binding International Agreement without the Senate or Congress?, 39 Harv Envir L Rev 515 (2015); Oona A. Hathaway and Amy Kapczynski, Going It Alone: The Anti-Counterfeiting Trade Agreement as a Sole Executive Agreement (ASIL Insights, Aug 24, 2011), archived at http://‌perma.cc/‌7TWW‌-6X4G.
• 38See United States v Belmont, 301 US 324, 331 (1937).
• 39Koh, 101 Georgetown L J at 726–27 (cited in note 14). See also Harold Hongju Koh, Triptych’s End: A Better Framework to Evaluate 21st Century International Lawmaking, 126 Yale L J F 338, 345–49 (2017) (proposing a three-factor test for evaluating whether international commitments have been joined in a constitutionally appropriate manner).
• 40See, for example, Martti Koskenniemi and Päivi Leino, Fragmentation of International Law? Postmodern Anxieties, 15 Leiden J Intl L 553, 556 (2002) (fragmented); Paul Schiff Berman, Global Legal Pluralism: A Jurisprudence of Law beyond Borders 10–14 (Cambridge 2012) (pluralist); William W. Burke-White, Power Shifts in International Law: Structural Realignment and Substantive Pluralism, 56 Harv Intl L J 1, 5 (2015) (multi-hub). “Transnational” has the oldest pedigree. See Philip C. Jessup, Transnational Law 2 (Yale 1956).
• 41See, for example, Bradley, International Law at 91 (cited in note 11); Golove, 73 NYU L Rev at 1900–13 & n 36 (cited in note 16); Clark, 93 Va L Rev at 1592–93 (cited in note 16).
• 42Pursuant to the Vienna Convention on the Law of Treaties, which is now widely regarded as embodying customary international law, heads of state and other properly authorized national representatives have the capacity to represent their nations “for the purpose of expressing the consent of the State to be bound by a treaty.” Vienna Convention on the Law of Treaties Art 7, 1155 UNTS 331, 334, TIAS No 18232 (May 23, 1969, entered into force Jan 27, 1980) (“Vienna Convention”). Moreover, a nation “may not invoke the fact that its consent to be bound by [an international agreement] has been expressed in violation of a provision of its internal law regarding competence to conclude [international agreements] as invalidating its consent unless that violation was manifest and concerned a rule of its internal law of fundamental importance.” Vienna Convention Art 46, 1155 UNTS at 343.
• 43See Galbraith and Zaring, 99 Cornell L Rev at 769 (cited in note 7).
• 44For exceptions, see generally Duncan B. Hollis and Joshua J. Newcomer, “Political” Commitments and the Constitution, 49 Va J Intl L 507 (2009) (considering soft-law commitments from the perspective of constitutional law); Galbraith and Zaring, 99 Cornell L Rev 735 (cited in note 7) (analyzing how soft-law agreements engage with principles of both foreign relations law and administrative law). Following the Iran deal, this is beginning to change. See generally, for example, Michael D. Ramsey, Evading the Treaty Power?: The Constitutionality of Nonbinding Agreements, 11 FIU L Rev 371 (2016).
• 45Hollis and Newcomer, 49 Va J Intl L at 512 (cited in note 44).
• 46See US Const Art I, § 10, cl 1 (“No State shall enter into any Treaty, Alliance, or Confederation.”). See also US Const Art I, § 10, cl 3 (“No State shall, without the Consent of Congress . . . enter into any Agreement or Compact with another State, or with a foreign Power.”).
• 47See Duncan B. Hollis, Unpacking the Compact Clause, 88 Tex L Rev 741, 747–59 (2010) (describing this growing phenomenon and explaining how changes in the international legal structure have facilitated it).
• 48Agreement between the California Air Resources Board and the Gouvernement du Québec concerning the Harmonization and Integration of Cap-and-Trade Programs for Reducing Greenhouse Gas Emissions (Sept 25, 2013), archived at http://‌perma.cc/‌SJ3A‌-8ZAU. The structure of this document very much resembles a classic international-law agreement, complete with twenty articles, an entry-into-force provision, and a termination provision.
• 49See Michael S. Barr and Geoffrey P. Miller, Global Administrative Law: The View from Basel, 17 Eur J Intl L 15, 32–33 (2006).
• 50UN Charter Art 24(1).
• 51See Dan Bodansky, Jutta Brunnée, and Ellen Hey, International Environmental Law: Mapping the Field, in Dan Bodansky, Jutta Brunnée, and Ellen Hey, eds, The Oxford Handbook of International Environmental Law 1, 20 (Oxford 2007).
• 52Timothy Meyer, Codifying Custom, 160 U Pa L Rev 995, 998, 1057–68 (2012).
• 53As the discussion of “administrative law” in this Section indicates, this Article uses the term capaciously, including within its broad ambit transsubstantive statutes like the Administrative Procedure Act (APA), 60 Stat 237 (1946), codified as amended in various sections of Title 5, specific statutes applicable to agency action, and internal executive branch processes that are designed to further regularity, accountability, and transparency—all of which have implications for institutional dynamics within the executive branch.
• 54The literature unpacking the executive branch is more robust in relation to national security. See, for example, Neal Kumar Katyal, Internal Separation of Powers: Checking Today’s Most Dangerous Branch from Within, 115 Yale L J 2314, 2319–42 (2006) (considering intra–executive branch checks with a particular focus on the post-9/11 landscape).
• 55Oona A. Hathaway, Presidential Power over International Law: Restoring the Balance, 119 Yale L J 140, 152–53 (2009) (including both ex ante congressional-executive agreements and sole executive agreements in this tally).
• 56Indeed, some of the statutes that Hathaway identifies as delegating power to make these agreements do not delegate power to the president but rather to cabinet officials or other agency heads. See id at 159–65 (providing a list of statutes, some of which delegate authority to the secretary of defense, the secretary of state, the administrator of the US Agency for International Development, and “[t]he Postal Service, with the consent of the President”).
• 57Uruguay Round Agreements Act § 491, Pub L No 103-465, 108 Stat 4809, 4970–71 (1994).
• 5852 Stat 1040 (1938), codified as amended at 21 USC § 301 et seq.
• 59House Executive Overreach Task Force Hearing, 114th Cong, 2d Sess at 54 (cited in note 17) (testimony of Steven Groves of The Heritage Foundation’s Freedom Project). For an example of how Congress can use its power to influence agency participation in ongoing negotiations within an international institution, see Kristina Daugirdas, Congress Underestimated: The Case of the World Bank, 107 Am J Intl L 517, 537–39 (2013) (describing how Congress used its appropriations power to pressure the Department of the Treasury to advocate for certain policies within the World Bank).
• 60See, for example, Jon D. Michaels, An Enduring, Evolving Separation of Powers, 115 Colum L Rev 515, 530–51 (2015).
• 6122 CFR § 181.4(a). Presidential control here thus runs through the State Department, which may use its power in ways that further a particular vision of the president’s agenda. Nonetheless, as a matter of constitutional law, the State Department is acting here as the agent of the president. Functionally, its decisions are especially likely to reflect presidential preferences for international agreements that are important enough to attract the personal involvement of the president.
• 62This is part of the C-175 Procedure, see notes 69–73 and accompanying text, which goes back to the 1950s. Interestingly, it is also promoted by Congress under the Case-Zablocki Act, Pub L No 92-403, 86 Stat 619 (1972), codified as amended at 1 USC § 112b (providing that no international agreement may be signed “without prior consultation with the Secretary of State”).
• 63Executive Order 13609 § 2(a) (2013), 3 CFR 255, 255–56. Agencies must report their intent to engage in such cooperative activities in their Regulatory Plan, which is made public annually. Executive Order 13609 § 3(a), 3 CFR at 256 (cross-referencing Executive Order 12866). See also Executive Order 12866 § 4(c)(7) (1994), 3 CFR 638, 643 (providing that the Regulatory Plans shall be published each October).
• 64Executive Order 13609 § 2, 3 CFR at 255–56.
• 65Jennifer Nou, Agency Self-Insulation under Presidential Review, 126 Harv L Rev 1755, 1764 (2013).
• 66See Galbraith and Zaring, 99 Cornell L Rev at 769 n 168 (cited in note 7) (considering how the techniques described by Nou could play out with respect to nonbinding international commitments).
• 67For example, independent regulatory agencies are simply “encouraged to comply” with the provisions of Executive Order 13609, unlike other agencies that are required to do so. See Executive Order 13609 §§ 4(a), 5, 3 CFR at 257.
• 68Peter Conti-Brown and David Zaring, Foreign Affairs and the Federal Reserve *4, 6–7 (unpublished draft on file with author).
• 69US Department of State, 11 Foreign Affairs Manual §§ 720–27, archived at http://‌perma.cc/‌J6UY‌-DBPP (“11 FAM”).
• 7022 CFR § 181.
• 7111 FAM § 723.3 (cited in note 69). The eight factors are:

  (1) [t]he extent to which the agreement involves commitments or risks affecting the nation as a whole; (2) [w]hether the agreement is intended to affect state laws; (3) [w]hether the agreement can be given effect without the enactment of subsequent legislation by the Congress; (4) [p]ast U.S. practice as to similar agreements; (5) [t]he preference of the Congress as to a particular type of agreement; (6) [t]he degree of formality desired for an agreement; (7) [t]he 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 (8) [t]he general international practice as to similar agreements.

• 7211 FAM § 723.4 (cited in note 69).
• 7311 FAM §§ 722(5), 725.1(6) (cited in note 69).
• 74See note 57 and accompanying text.
• 75Pub L No 92-403, 86 Stat 619 (1972), codified as amended at 1 USC § 112b.
• 76See Case-Zablocki Act § 1, 86 Stat at 619, 1 USC § 112b(a) (further limiting the reporting of agreements “the immediate public disclosure of which would, in the opinion of the President, be prejudicial to the national security of the United States”). The C-175 Procedure also provides that “unless classified, [international agreements] generally are published by the Office of the Assistant Legal Adviser for Treaty Affairs” some time following their entry into force. 11 FAM § 725.3(a) (cited in note 69).
• 7760 Stat 237 (1946), codified as amended in various sections of Title 5.
• 78See Galbraith and Zaring, 99 Cornell L Rev at 775–77 (cited in note 7); 5 USC § 553(a)(1) (exempting a “military or foreign affairs function of the United States” from notice-and-comment rulemaking).
• 79Pub L No 94-409, 90 Stat 1241 (1976), codified in various sections of Title 5.
• 80Federal Communications Commission v ITT World Communications, Inc, 466 US 463, 473 (1984) (holding that the international negotiation at issue was not a meeting “of an agency” for purposes of the Government in the Sunshine Act because the FCC did not convene this negotiation and did not have unilateral control over its procedures).
• 81This approach is currently embodied in the C-175 Procedure, discussed in notes 69–73 and accompanying text, in which, as a constitutional matter, the State Department acts as the agent of the president. Congress has not to date sought to dictate the choice of pathway (although the Case-Zablocki Act sets out procedural requirements related to this choice), and the extent to which it could do so is an interesting constitutional question. See note 236 (discussing a “sense of the Senate” resolution in 1969 on this issue). Sometimes in the course of obtaining advice and consent for treaties, the executive branch makes explicit commitments with regard to the path that will be pursued with respect to future, related agreements.
• 82See Jean Galbraith, International Law and the Domestic Separation of Powers, 99 Va L Rev 987, 1027–33 (2013) (discussing sole executive agreements); Curtis A. Bradley and Trevor W. Morrison, Historical Gloss and the Separation of Powers, 126 Harv L Rev 411, 468–76 (2012) (discussing congressional-executive agreements).
• 83See Part I.A. Constitutional law provides enough support for these options for the president to proceed with them in the first instance. It is theoretically possible but, in my view, highly unlikely that a court might one day find an exercise of one of these options unconstitutional. See, for example, Made in the USA Foundation v United States, 242 F3d 1300, 1319–20 (11th Cir 2001) (finding the question whether NAFTA was appropriately done as a congressional-executive agreement to be a “nonjusticiable political question”).
• 84See Part I.A.
• 85Henkin, Foreign Affairs at 224 (cited in note 11).
• 8650 ILM 239 (2011).
• 87Hathaway and Kapczynski, Going It Alone (cited in note 37). ACTA “establishes new norms across a range of intellectual property rights, with an emphasis on heightened penalties, more summary proceedings, more extensive border enforcement, and the introduction of obligations for third parties.” Id.
• 88Ramsey, 11 FIU L Rev at 387 (cited in note 44).
• 89See generally House Executive Overreach Task Force Hearing, 114th Cong, 2d Sess (cited in note 17).
• 90It isn’t quite this simple; hence the qualifier “at its core.” See Jean Galbraith, Congress’s Treaty-Implementing Power in Historical Practice, 56 Wm & Mary L Rev 59, 83–93 (2014) (discussing how the implementation of some treaties has come through practice to depend on the passage of further legislation by Congress); Jean Galbraith, Prospective Advice and Consent, 37 Yale J Intl L 247, 251–74, 285 (2012) (discussing how the Senate has on rare occasions used its advice-and-consent power at the negotiating stage and also noting how the Senate’s internal procedures enhance the already-high bar of the supermajority voting requirement).
• 91Jose E. Alvarez, The New Treaty Makers, 25 BC Intl & Comp L Rev 213, 227 (2002) (describing how existing treaty regimes limit the traditional diplomatic powers of powerful states).
• 92See, for example, Agreement between the Government of the United States of America and the Government of Canada on Fisheries Enforcement, TIAS No 11753 (Sept 26, 1990, entered into force Dec 16, 1991) (further referencing various aspects of the law of the sea).
• 93See Jean Galbraith, Comment, Ending Security Council Resolutions, 109 Am J Intl L 806, 807–10 (2015) (describing the interconnections between the Iran deal and a subsequent Security Council resolution approving it).
• 94United States v Curtiss-Wright Export Corp, 299 US 304, 319 (1936).
• 95See Zivotofsky v Kerry, 135 S Ct 2076, 2086, 2089–90 (2015) (signaling some disapproval of this language yet simultaneously affirming that the president “has the sole power to negotiate treaties”).
• 96See Galbraith, 37 Yale J Intl L at 251–52, 303–04 (cited in note 90) (describing the Senate’s resolution during the Kyoto Protocol negotiations that the United States should not sign any treaties that committed itself to limiting greenhouse gas emissions unless developing counties similarly bound themselves and the treaty would not seriously harm the US economy).
• 97For an overview, see generally Ian F. Fergusson, Trade Promotion Authority (TPA) and the Role of Congress in Trade Policy (Congressional Research Service, June 15, 2015), archived at http://‌perma.cc/‌PQH4‌-9CW3.
• 98Tom Cotton, et al, Open Letter to the Leaders of the Islamic Republic of Iran (Mar 9, 2015), archived at http://‌perma.cc/‌KZ73‌-5UX7. For a detailed picture drawn from Wikileaks data of the direct diplomatic activities of members of Congress, see Ryan M. Scoville, Legislative Diplomacy, 112 Mich L Rev 331, 336–56 (2013).
• 99For elaboration of this point with respect to the Federal Reserve, see generally Conti-Brown and Zaring, Foreign Affairs and the Federal Reserve (cited in note 68). For discussion of how states make nonbinding commitments with foreign states, see Hollis, 88 Tex L Rev at 743–44 (cited in note 47) (concluding that such commitments have become “remarkably common”).
• 100See The Impact of International Regulatory Standards on the Competitiveness of U.S. Insurers, Part II, Hearing before the Subcommittee on Housing and Insurance of the House Committee on Financial Services, 114th Cong, 2d Sess 6–8 (2016) (testimony of Professor David Zaring) (describing ways in which a proposed bill in Congress would incorporate more administrative-law principles into this negotiating process).
• 101Minamata Convention on Mercury, 55 ILM 586 (2016) (not in force).
• 102See Duncan Hollis, Doesn’t the U.S. Senate Care about Mercury? (Opinio Juris, Nov 12, 2013), archived at http://‌perma.cc/‌5GHZ‌-3XMC.
• 103US Department of State, United States Joins Minamata Convention on Mercury (Nov 6, 2013), archived at http://‌perma.cc/‌66VU‌-SNSJ.
• 104See Minamata Convention on Mercury (EPA), archived at http://‌perma.cc/‌U5LC‌
  -HVSY (noting that “EPA worked closely with the State Department and other federal agencies in the negotiation of this agreement”).
• 105The fact that US negotiators can be domestically constrained even without the subsequent need for legislative approval is recognized, if less discussed, in classic work on international bargaining. See, for example, Robert D. Putnam, Diplomacy and Domestic Politics: The Logic of Two-Level Games, 42 Intl Org 427, 429, 448 (1988) (giving an example of how the president was a constrained negotiator even though approval from the Senate and Congress was not required, although also offering other examples when such approval would be required).
• 106See Hollis and Newcomer, 49 Va J Intl L at 538–75 (cited in note 44). See also generally Ramsey, 11 FIU L Rev 371 (cited in note 44).
• 107See, for example, notes 219–21 and accompanying text (detailing how the president’s negotiators could not achieve having the overall structure of the Paris Agreement take a nonbinding form).
• 108See generally Lisa L. Martin, The President and International Commitments: Treaties as Signaling Devices, 35 Pres Stud Q 440 (2005) (arguing that the form of domestic ratification serves as a signal of the strength of the US intentions in ways that other nations may view as significant).
• 109Senate Resolution of Advice and Consent to the New START Treaty at 23472 (cited in note 5).
• 110Resolution of Advice and Consent to Ratification of the Convention on the Rights of Persons with Disabilities, 112th Cong, 2d Sess, in 158 Cong Rec 16184–85 (Dec 4, 2012).
• 111See Statement by the President on the Signing of the Trans-Pacific Partnership (cited in note 10); Presidential Memorandum regarding Withdrawal of the United States (cited in note 10).
• 112For past practice relevant to ACTA, see Galbraith, 99 Va L Rev at 1040–41 (cited in note 82). For past practice relevant to the Minamata Convention and the Paris Agreement, see Wirth, 39 Harv Envir L Rev at 552–61 (cited in note 37). This Article takes no position on whether the administration’s decision to bypass the Senate and Congress for the Minamata Convention stemmed from concerns about resistance to the Convention or instead from an interest in establishing a precedent that in turn would be useful for the Paris Agreement.
• 113See Part IV (describing resistance to the administration’s choice with respect to the Paris Agreement). See also Galbraith, 99 Va L Rev at 1040–41 (cited in note 82) (describing resistance to the administration’s choice with respect to the ACTA).
• 114Minamata Convention on Mercury Arts 15, 21–23, 55 ILM at 599–600, 602–04.
• 115A fourth, rarer category is a commitment whose terms will be implemented by US states. In a very early example of a significant nonbinding commitment, for example, President Theodore Roosevelt reached an arrangement with Japan whose terms tacitly depended on Roosevelt’s ability to persuade California to improve its treatment of Japanese schoolchildren. See Devon W. Carbado, Yellow by Law, 97 Cal L Rev 633, 641–43 (2009) (describing the “Gentlemen’s Agreement”).
• 116This list is exemplary rather than exclusive. For a longer discussion of the president’s powers, see Henkin, Foreign Affairs at 31–62 (cited in note 11).
• 117Medellin v Texas, 552 US 491, 527–32 (2008) (emphasizing that the president lacks the power to make domestic law and describing the claims-settlement context as “involv[ing] a narrow set of circumstances”).
• 118The authority to implement treaties sometimes flows from their status as the law of the land (which in turn requires that they have received the Senate’s advice and consent and been ratified) and sometimes from implementing legislation passed by Congress, which in practice will typically be passed before the president ratifies the treaty. See Galbraith, 56 Wm & Mary L Rev at 76 & n 62 (cited in note 90). For ex post congressional-executive agreements, in essence Congress votes simultaneously on the approval and implementation of the agreement. See Fergusson, Trade Promotion Authority at *1, 10 (cited in note 97).
• 119See Galbraith and Zaring, 99 Cornell L Rev at 778–81 (cited in note 7).
• 120See id at 786–87 (giving an example of the controlling effect of the terms of the implementing legislation with respect to Basel III and the Dodd-Frank Act).
• 121For a fuller and more complex picture, see Dianne E. Rennack, Iran: U.S. Economic Sanctions and the Authority to Lift Restrictions *5–32 (Congressional Research Service, Jan 22, 2016), archived at http://‌perma.cc/‌XLP4‌-2ZE6.
• 122See id (listing the various statutes and noting the availability of waivers).
• 123Comprehensive Iran Sanctions, Accountability, and Divestment Act of 2010 § 202, Pub L No 111-195, 124 Stat 1312, 1342–43, codified at 22 USC § 8532.
• 124See Jack Goldsmith and Amira Mikhail, Does the Iran Deal Require the USG to Seek Preemption of (Some) State Sanctions? (Lawfare, Apr 27, 2016), archived at http://‌perma.cc/‌4SK7‌-ARL2. For a list of many other sanctions that the Iran deal left in place, see Rennack, Iran at *33–36 (cited in note 121).
• 125See David E. Sanger, Trump Seeks Way to Declare Iran in Violation of Nuclear Deal (NY Times, July 27, 2017), online at http://‌www.nytimes.com/‌2017/‌07/‌27/‌world/‌middleeast/‌trump‌-iran‌-nuclear‌-agreement‌.html (visited Aug 19, 2017) (Perma archive unavailable) (describing how President Trump is interested in abandoning the Iran deal despite having continued to abide by it so far).
• 126See Part IV.B (discussing this possibility with respect to the Paris Agreement).
• 127Goldsmith, 57 Harv Intl L J at 471 (cited in note 37).
• 128Gillian E. Metzger, Ordinary Administrative Law as Constitutional Common Law, 110 Colum L Rev 479, 485 (2010).
• 129See id at 486–512. For an extended historical account of how agencies can engage in constitutional decision-making, see generally Sophia Z. Lee, The Workplace Constitution from the New Deal to the New Right (Cambridge 2014).
• 130Metzger, 110 Colum L Rev at 508 (cited in note 128). See also Michaels, 115 Colum L Rev at 520 (cited in note 60). The role played by federalism in administrative law can similarly have structural implications for constitutional principles of the separation of powers, as the work of Professor Jessica Bulman-Pozen shows. Jessica Bulman-Pozen, Federalism as a Safeguard of the Separation of Powers, 112 Colum L Rev 459, 461–63 (2012).
• 131See Daniel Abebe, The Global Determinants of U.S. Foreign Affairs Law, 49 Stan J Intl L 1, 36–49 (2013) (arguing that the president’s foreign affairs powers as a matter of domestic law should vary with the level of constraint arising from the international political context).
• 132By contrast, administrative law’s ability to channel constitutional values and to substitute for other forms of constitutional checks is due in no small part to the availability of judicial review. See Metzger, 110 Colum L Rev at 485 (cited in note 128) (noting the significance of “judicial development of administrative law doctrines that respond to constitutional concerns associated with administrative government”).
• 133See Curtis A. Bradley and Jean Galbraith, Presidential War Powers as an Interactive Dynamic: International Law, Domestic Law, and Practice-Based Legal Change, 91 NYU L Rev 689, 708–12 (2016); Rebecca Ingber, International Law Constraints as Executive Power, 57 Harv Intl L J 49, 57–90 (2016); Galbraith, 99 Va L Rev at 1008–33 (cited in note 82).
• 134See Ashley Deeks, Checks and Balances from Abroad, 83 U Chi L Rev 65, 68 (2016) (noting that if “one recognizes that US national security increasingly relies on relationships with foreign partners, then the idea that the executive responds to foreign critiques and concerns to enable ongoing partnerships has bite”).
• 135See Ackerman and Golove, 108 Harv L Rev at 802–06, 832–36, 893–95 (cited in note 16).
• 136See Galbraith, 37 Yale J Intl L at 287–88 (cited in note 90); Hollis and Newcomer, 49 Va J Intl L at 510–11 (cited in note 44).
• 137Wirth, 39 Harv Envir L Rev at 552 (cited in note 37).
• 138See Galbraith and Zaring, 99 Cornell L Rev at 786 (cited in note 7).
• 139See Jessica Bulman-Pozen, Executive Federalism Comes to America, 102 Va L Rev 953, 959–63 (2016) (describing the general rise of partisan polarization). See also generally Glen S. Krutz and Jeffrey S. Peake, President Obama, Congress and International Agreements: An Initial Assessment (APSA 2011 Annual Meeting Paper, Aug 28, 2011), archived at http://‌perma.cc/‌S45R‌-5523 (concluding that Senate polarization with respect to treaties had been even higher for President Obama than for prior presidents).
• 140See generally Bradley and Galbraith, 91 NYU L Rev 689 (cited in note 133) (describing the growth of the president’s unilateral war powers over time, including the power to act without congressional authorization).
• 141Even with diplomatic options, forcible intervention remains a possibility, as demonstrated by the Stuxnet cyberattack on Iran’s nuclear enrichment program, which was carried out in part by the United States. See David E. Sanger, Obama Order Sped Up Wave of Cyberattacks against Iran (NY Times, June 1, 2012), archived at http://‌perma.cc/‌2F2T‌-LPYS.
• 142See US Department of Defense, National Security Implications of Climate-Related Risks and a Changing Climate *3 (July 23, 2015), archived at http://‌perma.cc/‌C6BG‌-Y7SQ (noting that “climate change is an urgent and growing threat to our national security, contributing to increased natural disasters, refugee flows, and conflicts over basic resources such as food and water”).
• 143Compare generally Jack Goldsmith, Power and Constraint: The Accountable Presidency after 9/11 (Norton 2012), with Eric A. Posner and Adrian Vermeule, The Executive Unbound: After the Madisonian Republic (Oxford 2010).
• 144See Rennack, Iran at *5–31 (cited in note 121) (outlining the fuller and more complex picture).
• 145See Part IV.B.
• 146See Japan Whaling Association v American Cetacean Society, 478 US 221, 241 (1986) (concluding that the secretary of commerce “furthered [the] objective” of a statute authorizing the impositions of sanctions related to whaling when he chose not to impose sanctions but instead to “enter[ ] into [an] agreement with Japan, calling for that nation’s acceptance of [a] worldwide moratorium on commercial whaling”).
• 147See 42 USC § 7415 (noting the prospect of “[i]nternational air pollution”).
• 148See Goldsmith, 57 Harv Intl L J at 467–68 (cited in note 37) (discussing a congressional statute that sought to limit certain forms of collaboration with China in the technology space).
• 149For a discussion of the complicated way in which these issues played out in relation to the Iran deal, see David M. Herszenhorn, The Iran Nuclear Deal: Congress Has Its Say (NY Times, Sept 7, 2015), archived at http://‌perma.cc/‌U7QP‌-VV2T.
• 150See Part I.B.
• 151Similarly, Congress might feel pressured to honor an international commitment and therefore not repeal or alter preexisting legislation that is being used to implement this commitment. Professors Hathaway and Kapczynski make this argument in relation to ACTA, see Hathaway and Kapczynski, Going It Alone (cited in note 37), even though it has a 180-day withdrawal provision that could be invoked. See ACTA Art 41, 50 ILM at 256. I focus here only on the future president’s incentives because any efforts by Congress to get the United States to violate or withdraw from an international commitment would almost certainly need his or her support.
• 152Ramsey, 11 FIU L Rev at 375–76 (cited in note 44).
• 153Goldsmith, 57 Harv Intl L J at 471 (cited in note 37). See also Koh, 126 Yale L J F at 359–61 (cited in note 39) (discussing how the “default path of least resistance becomes compliance”).
• 154See Hathaway, 119 Yale L J at 258–59 (cited in note 55) (arguing that as a prudential matter, international agreements whose terms require more than a year’s notice for withdrawal should go to the Senate or Congress for specific approval).
• 155See Barbara Koremenos, The Continent of International Law: Explaining Agreement Design 143 (Cambridge 2016).
• 15611 FAM § 723.3(7) (cited in note 69).
• 157See Spiro, 79 Tex L Rev at 966, 1000–02 (cited in note 16).
• 158Minamata Convention on Mercury Art 33, 55 ILM at 607.
• 159See note 244 and accompanying text.
• 160Transcript of Republican Presidential Candidates’ Debate in Miami (CNN, Mar 15, 2016), archived at http://‌perma.cc/‌Z6VS‌-UKQG.
• 161An America First Energy Plan (Donald J. Trump for President, May 26, 2016), archived at http://‌perma.cc/‌2MMZ‌-4K45 (transcript of speech delivered by Trump in Bismarck, North Dakota).
• 162See Sanger, Trump Seeks Way (cited in note 125).
• 163See Statement by President Trump on the Paris Climate Accord (cited in note 22) (announcing that “as of today, the United States will cease all implementation” and that it “is time to exit the Paris Accord”); Executive Order 13783 § 4, 82 Fed Reg 16093, 16095 (2017) (instructing the EPA administrator to seek to “suspend, revise, or rescind” the Clean Power Plan).
• 164See United States v Belmont, 301 US 324, 331 (1937).
• 165See Part I.A.
• 166See David Golove, Congress Just Gave the President Power to Adopt a Binding Legal Agreement with Iran (Just Security, May 14, 2015), archived at http://‌perma.cc/‌7SH7‌-VZBB (arguing that the Iran Nuclear Agreement Review Act of 2015 unintentionally authorized the president as a matter of constitutional law to make a binding international legal agreement with Iran).
• 167See notes 210–11 and accompanying text.
• 168See, for example, Letter of Law Professors to the Senate Committee on Finance (May 16, 2012), archived at http://‌perma.cc/‌RMY8‌-NAG2; Hathaway and Kapczynski, Going It Alone (cited in note 37); Jack Goldsmith and Lawrence Lessig, Anti-Counterfeiting Agreement Raises Constitutional Concerns (Wash Post, Mar 26, 2010), archived at http://‌perma.cc/‌HBM2‌-F58R.
• 169See, for example, Hathaway and Kapczynski, Going It Alone (cited in note 37):

  [N]egotiating partners complained to the United States about the unusual degree of secrecy, arguing that the level of confidentiality in these ACTA negotiations has been set at a higher level than is customary for non-security agreements and that the secrecy had inhibited consultation with those who would be affected by the agreement.

  See also, for example, Goldsmith and Lessig, Anti-Counterfeiting Agreement (cited in note 168) (“These mostly secret negotiations have already violated the Obama administration’s pledge for greater transparency. . . . Congress should resist this attempt to evade the checks established by our Framers.”).

• 170Gwen Hinze, U.S. Law Professors Cast Further Doubt on ACTA’s Constitutionality—State Department Confirms No ACTA Pre-review (Electronic Frontier Foundation, May 16, 2012), archived at http://‌perma.cc/‌YX5S‌-45MA (noting that a FOIA request produced the response that no C-175 Memorandum was ever issued for ACTA and describing the executive branch’s approach of first defending ACTA as a sole executive agreement and then as an ex ante congressional-executive agreement as a “surprising about-face”).
• 171Charles Arthur, ACTA Down but Not Out, as Europe Votes against Controversial Treaty (The Guardian, July 4, 2012), archived at http://‌perma.cc/‌GE5P‌-VEWC.
• 172Hathaway, 119 Yale L J at 239–53 (cited in note 55).
• 173Id. These proposed remedies are effectively administrative-law solutions to Hathaway’s concern about excessive presidential power.
• 174Id at 144–45.
• 175Helen V. Milner and Dustin Tingley, Sailing the Water’s Edge: The Domestic Politics of American Foreign Policy 165–68 (Princeton 2015) (showing that there is considerable variation with how answerable agencies engaged in foreign affairs are to Congress and finding in particular that agencies involved in military affairs are typically less answerable).
• 176Memorandum from Caroline D. Krass, Principal Deputy Assistant Attorney General, Office of Legal Counsel, to Eric Holder, Attorney General, Authority to Use Military Force in Libya 12 (Apr 1, 2011), archived at http://‌perma.cc/‌C2W7‌-ZJXN; Libya and War Powers, Hearing before the Senate Committee on Foreign Relations, 112th Cong, 1st Sess 11 (2011) (statement of Harold H. Koh, Legal Adviser, US Department of State).
• 177See generally Department of Justice White Paper, Lawfulness of a Lethal Operation Directed against a U.S. Citizen Who Is a Senior Operational Leader of Al-Qa’ida or an Associated Force (Nov 8, 2011), archived at http://‌perma.cc/‌X2XD‌-W2CK.
• 178Hinze, U.S. Law Professors Cast Further Doubt (cited in note 170).
• 179Jean Galbraith, Making Treaty Implementation More Like Statutory Interpretation, 115 Mich L Rev 1309, 1319–20 (2017).
• 180Id at 1322–23.
• 181See Hathaway, 119 Yale L J at 150 n 16 (cited in note 55) (finding only nine ex post congressional-executive agreements made between 1980 and 2000, most of which related to trade).
• 182This was understood, for example, to be “a vital prerequisite” to the TPP. Jonathan Weisman, Trade Authority Bill Wins Final Approval in Senate (NY Times, June 24, 2015), archived at http://‌perma.cc/‌4MDW‌-934J.
• 183Id (noting that this fast-track process will apply to agreements negotiated in the six years following the most recent enactment).
• 184Galbraith, 37 Yale J Intl L at 249–51 (cited in note 90).
• 185Ronald A. Lehmann, Note, Reinterpreting Advice and Consent: A Congressional Fast Track for Arms Control Treaties, 98 Yale L J 885, 896–97 (1989).
• 186Galbraith, 115 Mich L Rev at 1358–59 (cited in note 179).
• 187John F. Coyle, The Case for Writing International Law into the U.S. Code, 56 BC L Rev 433, 482–84 (2015).
• 188Hathaway, 117 Yale L J at 1352–54 (cited in note 16).
• 189Jack Goldsmith, The Obama Administration’s Contributions to International Law (Lawfare, June 28, 2016), archived at http://‌perma.cc/‌SH3P‌-2J9Vhttp://‌perma.cc/‌SH3P‌
  -2J9V‌?type‌=‌image.
• 190299 US 304 (1936).
• 191Id at 319.
• 192Cotton, et al, Open Letter (cited in note 98). For purposes of this discussion, I assume that the Logan Act, 1 Stat 613 (1799), codified at 18 USC § 953, is unlikely to serve as a meaningful bar to international engagement by governmental actors outside the executive branch.
• 193Transcript of Face the Nation (CBS News, Mar 15, 2015), archived at http://‌perma.cc/‌K2GU‌-JXQS (quoting Kerry for the proposition that the letter is “contrary to the Constitution with respect to the executive’s right to negotiate” and “unconstitutional”).
• 194See note 95 (discussing the mixed signals sent in Zivotofsky).
• 195Belmont, 301 US at 331; American Insurance Association v Garamendi, 539 US 396, 413–20 (2003).
• 196See Zivotofsky v Kerry, 135 S Ct 2076, 2088 (2015).
• 197This is reinforced by the fact that some of the claims-settlement cases, including Belmont, were closely tied to recognition decisions.
• 198For scholarship urging resistance to expansive executive power with respect to implementation and preemption, see, for example, Sarah H. Cleveland, Crosby and the “One-Voice” Myth in U.S. Foreign Relations, 46 Vill L Rev 975, 995 (2001); David H. Moore, Beyond One Voice, 98 Minn L Rev 953, 1017–32 (2014); Peter J. Spiro, Foreign Relations Federalism, 70 U Colo L Rev 1223, 1259–70 (1999). For discussion of how state and local governments interact with the federal government with respect to foreign affairs, see generally Jean Galbraith, Book Review, Cooperative and Uncooperative Foreign Affairs Federalism, 130 Harv L Rev 2131 (2017).
• 199A draft of the Restatement (Fourth) of the Foreign Relations Law of the United States, for example, signals this flexibility by stating that US courts “will ordinarily give great weight to an interpretation [of a treaty] by the executive branch.” Restatement (Fourth) of the Foreign Relations Law of the United States § 106 (April 2015 Discussion Draft) (emphasis added). This language is more flexible than that of the Restatement (Third). See Restatement (Third) of the Foreign Relations Law of the United States § 326(2) (1987) (stating that courts “will give great weight to an interpretation made by the Executive Branch”). With respect to administrative law, “agencies seeking to defend statutory interpretations in court can anticipate with confidence neither what standard [of deference] will be applied nor how the court will apply it.” Jud Mathews, Deference Lotteries, 91 Tex L Rev 1349, 1351 (2013) (discussing the scholarship on this issue). See also William N. Eskridge Jr and Lauren E. Baer, The Continuum of Deference: Supreme Court Treatment of Agency Statutory Interpretations from Chevron to Hamdan, 96 Georgetown L J 1083, 1091 (2008) (concluding that “there is no clear guide as to when the Court will invoke particular deference regimes, and why”).
• 200Murray v Schooner Charming Betsy, 6 US (2 Cranch) 64, 118 (1804) (“[A]n act of Congress ought never to be construed to violate the law of nations if any other possible construction remains.”).
• 201See Galbraith and Zaring, 99 Cornell L Rev at 773–75, 790–92 (cited in note 7).
• 202See Eilperin, Obama Hails ‘Historic’ Ratification (cited in note 9).
• 203House Executive Overreach Task Force Hearing, 114th Cong, 2d Sess at 2 (cited in note 17) (statement of Rep King).
• 204William K. Stevens, With Climate Treaty Signed, All Say They’ll Do Even More (NY Times, June 13, 1992), archived at http://‌perma.cc/‌6JXV‌-HUTP.
• 205See UN Framework Convention on Climate Change Art 4, 1771 UNTS 107, 170 (June 4, 1992, entered into force Mar 21, 1994) (“UNFCCC”).
• 206UNFCCC Arts 7–18, 1771 UNTS at 176–85.
• 207Transmission of UN Framework Convention on Climate Change for the Advice and Consent of the Senate, 102d Cong, 2d Sess, in 138 Cong Rec 23902 (Sept 8, 1992).
• 208Resolution of Ratification of UN Framework Convention on Climate Change, 102d Cong, 2d Sess, in 138 Cong Rec 33527 (Oct 7, 1992).
• 209See David A. Wirth, A Matchmaker’s Challenge: Marrying International Law and American Environmental Law, 32 Va J Intl L 377, 400 (1992).
• 210U.N. Framework Convention on Climate Change, Hearing before the Senate Committee on Foreign Relations, 102d Cong, 2d Sess 106 (1992) (“Responses of the Administration to Questions Asked by the Foreign Relations Committee”) (stating when asked about protocols more generally that “[w]e would expect” them to be submitted to the Senate but that “given that a protocol could be adopted on any number of subjects, treatment of any given protocol would depend on its subject matter”).
• 211United Nations Framework Convention on Climate Change, S Rep No 102-55, 102d Cong, 2d Sess 14 (1992).
• 212See Galbraith, 37 Yale J Intl L at 303–04 (cited in note 90).
• 213See John R. Crook, ed, Note, U.S. Positions in International Climate Change Negotiations, 102 Am J Intl L 155, 166 (2008).
• 214549 US 497 (2007).
• 215Id at 528–35.
• 216An earlier nonbinding commitment done in 2009 at the UNFCCC Conference of the Parties at Copenhagen was in part a precursor to the Paris Agreement. See Copenhagen Accord (Dec 18, 2009), UN Doc FCCC/CP/2009/11/Add.1 4–7.
• 217Statement by the President on the Paris Climate Agreement (cited in note 19).
• 218For an overview of the complexities of the consensus requirement, see Alan Boyle and Navraj Singh Ghaleigh, Climate Change and International Law beyond the UNFCCC, in Cinnamon P. Carlarne, Kevin R. Gray, and Richard G. Tarasofsky, eds, The Oxford Handbook of International Climate Change Law 26, 37–39 (Oxford 2016).
• 219Report of the Conference of the Parties on Its Seventeenth Session, Held in Durban from 28 November to 11 December 2011: Part Two, UNFCCC Conference of the Parties, 17th sess (Dec 11, 2011), UN Doc FCCC/CP/2011/9/Add.1 2.
• 220Daniel Bodansky, The Durban Platform Negotiations: Goal and Options *2 (Harvard Project on Climate Agreements, July 2012), archived at http://‌perma.cc/‌3GZY‌-QASR (explaining that “[t]he United States insisted that it would accept a mandate to negotiate a new outcome of a legal nature only if the mandate was ‘symmetrical’ in its application to developing as well as developed countries”).
• 221The Obama administration did use nonbinding commitments as supplements to the Paris negotiations, such as a bilateral commitment made with China a year before the Paris Conference. See U.S.-China Joint Announcement on Climate Change (White House Office of the Press Secretary, Nov 12, 2014), archived at http://‌perma.cc/‌7B39‌-MZ6S.
• 222For details of her role there, see U.S. Environmental Regulation after the Paris Climate Talks: A Conversation with Gina McCarthy (Council on Foreign Relations, Jan 7, 2016), archived at http://‌perma.cc/‌FL42‌-ZFKD.
• 223See Provisional List of Participants, UNFCCC Conference of the Parties, 21st sess (Dec 1, 2015), UN Doc FCCC/CP/2015/MISC.2 40–43 (listing representatives from not only the White House and the State Department, but also the Departments of Agriculture, Commerce, Energy, the Interior, and the Treasury, the EPA, Office of the US Trade Representative, US Agency for International Development, National Oceanic and Atmospheric Administration, and the Forest Service, as well as participants from Congress and its staff).
• 224See generally, for example, Examining the International Climate Negotiations, Hearing before the Senate Committee on Environment and Public Works, 114th Cong, 1st Sess (2015) (“Senate Hearing on Examining Climate Negotiations”); Pitfalls of Unilateral Negotiations at the Paris Climate Change Conference, Hearing before the House Committee on Science, Space, and Technology, 114th Cong, 1st Sess (2015); Valerie Richardson, Republicans Move to Undermine Obama on Paris Climate Deal (Wash Times, Dec 7, 2015), archived at http://‌perma.cc/‌CA8G‌-MB3R (describing the hearing convened by Senator Ted Cruz).
• 225See Provisional List of Participants at 2 (cited in note 223) (noting over seven thousand registered participants from NGOs).
• 226See David Siders, Jerry Brown Wants All New Cars in California Zero-Emission by 2050 (Sacramento Bee, Dec 3, 2015), archived at http://‌perma.cc/‌R3W3‌-SJD7 (discussing a political commitment for future zero-emission cars made between California, Germany, the United Kingdom, New York, and several other sovereigns); UN Climate Conference: Governor Brown, German Government Announce 43 New Signatories to Under 2 MOU Climate Pact (California Office of the Governor, Dec 9, 2015), archived at http://‌perma.cc/‌2PPZ‌-8QMW (mentioning, among other things, a joint declaration with France); Chris Megerian, California Isn’t a Country, so Why Are So Many in the State Headed to Climate Talks in Paris? (LA Times, Dec 2, 2015), archived at http://‌perma.cc/‌5QM7‌-XGK2 (describing the sizeable California contingent, which included Brown, eight state legislators, and “a number of top Brown administration officials”).
• 227Compare Senate Hearing on Examining Climate Negotiations, 114th Cong 1st Sess at 8 (cited in note 224) (testimony of Professor Julian Ku) (“I don’t believe the Constitution allows the President to use a sole executive agreement . . . to legally bind the United States to particular greenhouse gas emissions targets.”), with Wirth, 39 Harv Envir L Rev at 552–61 (cited in note 37) (noting various precedents for making sole executive agreements with substantive requirements that could be fulfilled under existing domestic environmental law, including the 2013 Minamata Convention, a 1991 bilateral agreement with Canada, the 1979 Convention on Long-Range Transboundary Air Pollution, and three of that Convention’s subsequent protocols).
• 228See Responses of the Administration to Questions Asked by the Foreign Relations Committee, 102d Cong, 2d Sess at 106 (cited in note 210).
• 229Paris Agreement Art 4(2) (opened for signature Apr 22, 2016), archived at http://‌perma.cc/‌2BXZ‌-DK7W. The Paris Agreement does use the legally binding language “shall” for some substantive commitments—for example, it states that the “Parties shall pursue domestic mitigation measures,” but, instead of requiring any particular type of commitment, it then simply specifies that these measures will have “the aim of achieving the objectives of” the NDCs. Paris Agreement Art 4(2) (emphasis added).
• 230See Paris Agreement Art 4(2).
• 231See Adoption of the Paris Agreement, Proposal by the President, Draft Decision, UNFCCC Conference of the Parties, 21st sess (Dec 12, 2015), UN Doc FCCC/CP/2015/L.9 21 (emphasis added).
• 232Warrick, How One Word Nearly Killed the Climate Deal (cited in note 20) (quoting Kerry’s description of his conversation with the Conference President).
• 233See John Vidal, How a ‘Typo’ Nearly Derailed the Paris Climate Deal (The Guardian, Dec 16, 2015), archived at http://perma.cc/‌S8LZ‌-H8GC (describing this sharply contentious issue).
• 234See Copenhagen Accord at 7 (cited in note 216).
• 235In 2016, for example, the Obama administration provided $500 million for this purpose using discretionary funding. Timothy Cama, Obama Pays $500M to UN Climate Change Fund (The Hill, Mar 3, 2016), archived at http://‌perma.cc/‌32T2‌-KFUF.
• 236See Henkin, Foreign Affairs at 222–23 (cited in note 11) (discussing a 1969 Senate resolution expressing the view that a sole executive agreement should not “promise to assist a foreign country . . . by the use of the . . . financial resources of the United States”). Although this resolution does not formally bind the president, it does signal the need to tread especially cautiously as a matter of constitutional law when financial commitments are concerned. See id.
• 237Paris Agreement Art 9(1).
• 238Adoption of the Paris Agreement, UNFCCC Conference of the Parties, 21st sess (Jan 29, 2016), UN Doc FCCC/CP/2015/10/Add.1 8, 17 at ¶¶ 53, 114.
• 239Koh, 101 Georgetown L J at 726 (cited in note 14). See also Bodansky and Spiro, 49 Vand J Transnatl L at 887 (cited in note 37) (describing international agreements that are supported but not specifically authorized by Congress as “executive agreements+”).
• 240See Chris Mooney and Juliet Eilperin, Obama’s Rapid Move to Join the Paris Climate Agreement Could Tie Up the Next President (Wash Post, Apr 11, 2016), archived at http://‌perma.cc/‌9G4G‌-VG8H (quoting Bellinger); Goldsmith, 57 Harv Intl L J at 466–67, 469–72 (cited in note 37).
• 241See Galbraith, 99 Va L Rev at 1042–45 (cited in note 82) (describing how the executive branch tends to read its past practices broadly). While President Trump has announced his intent to withdraw the United States from the Paris Agreement, so far he has not repudiated Obama’s legal authority to have made the commitment in the first place. See Nick Juliano, What Does Trump’s Paris Climate Decision Mean? (Politico, May 31, 2017), archived at http://‌perma.cc/‌4MUW‌-JH5V (noting that “Trump apparently decided not to treat the Paris deal as a treaty and submit it for a Senate ratification vote, which would surely fail”).
• 242See House Executive Overreach Task Force Hearing, 114th Cong, 2d Sess at 43–44, 48–52, 59, 63–68 (cited in note 17) (testimony of Steven Groves).
• 243See id at 59, 64–68.
• 244Paris Agreement Art 28; UNFCCC Art 25, 1771 UNTS at 187.
• 245In addition, the EPA has institutional reasons to be attentive to congressional concerns in determining how far it is willing to go both in its legal interpretations and its uses of discretion. See Devin Henry, Spending Bill Keeps EPA Funding Flat in 2016 (The Hill, Dec 16, 2015), archived at http://‌perma.cc/‌AG4D‌-T95X (describing the failure of attempts to tie the EPA budget to the absence of climate action but noting that, overall, “[t]he bill keeps agency staffing levels at their lowest level since 1989”).
• 246See Bulman-Pozen, 102 Va L Rev at 966–67, 982–87 (cited in note 139) (describing the functional federalism in this rule). The Clean Power Plan is not the only action the EPA is taking with respect to climate change, but it is the most significant to date. See A Conversation with Gina McCarthy (cited in note 222).
• 247See generally Fact Sheet: Clean Power Plan by the Numbers (EPA), archived at http://‌perma.cc/‌F3GU‌-P33D; Environmental Protection Agency, Carbon Pollution Emission Guidelines for Existing Stationary Sources: Electric Utility Generating Units, 80 Fed Reg 64662 (2015), amending 40 CFR Part 60. This rule was thus finalized before the completion of the Paris negotiations.
• 248See Cary Coglianese, When Management-Based Regulation Goes Global (RegBlog, Dec 23, 2015), archived at http://‌perma.cc/‌T8LE‌-BM7K. It is possible that the United States could meet its NDC without the Clean Power Plan due to other factors, such as tax credits for renewables, state and local climate-mitigation measures, or a substantial economic contraction, but prospects for this are far from clear. Indeed, it is not even clear that the United States would succeed in meeting the NDC with the Clean Power Plan in place. See Warren Cornwall, United States Will Miss Paris Climate Targets without Further Action, Study Finds (Science, Sept 26, 2016), archived at http://‌perma.cc/‌3YM6‌-UN53.
• 249West Virginia v Environmental Protection Agency, 136 S Ct 1000, 1000 (2016).
• 250Executive Order 13783 § 4, 82 Fed Reg at 16095 (cited in note 163).
• 251Statement by President Trump on the Paris Climate Accord (cited in note 22).
• 252See Ashley Parker, et al, ‘I Was All Set to Terminate’: Inside Trump’s Sudden Shift on NAFTA (Wash Post, Apr 27, 2017), archived at http://‌perma.cc/‌2TLL‌-A5ZB; Jenna Johnson, Trump on NATO: ‘I Said It Was Obsolete. It’s No Longer Obsolete.’ (Wash Post, Apr 12, 2017), archived at http://‌perma.cc/‌M3WY‌-JJL6.
• 253Open Letter to the International Community and Parties to the Paris Agreement from US State, Local, and Business Leaders (We Are Still In, June 5, 2017), archived at http://‌perma.cc/‌EQM2‌-FWZ7 (containing the pledge of nine states and numerous cities and counties “to continue to support climate action to meet the Paris Agreement”).
• 254A lesser threshold would be needed for the United States to remain in the Paris Agreement and make meaningful though inadequate efforts to comply with it. The Trump administration could block even the achievement of this threshold, however, by withdrawing the United States from the Paris Agreement or potentially by impeding the efforts of states like California that are pursuing emissions reductions.