Microsoft Word - Rise of Nonbindings

THE RISE OF NONBINDING INTERNATIONAL AGREEMENTS:

AN EMPIRICAL, COMPARATIVE, AND NORMATIVE ANALYSIS

Curtis A. Bradley,

Jack Goldsmith

& Oona A. Hathaway

***

The Article II treaty process has been dying a slow death for decades,

replaced by various forms of “executive agreements.” What is only beginning to be

appreciated is the extent to which both treaties and executive agreements are

increasingly being overshadowed by another form of international cooperation:

nonbinding international agreements. Not only have nonbinding agreements become

more prevalent, but many of the most consequential (and often controversial) U.S.

international agreements in recent years have been concluded in whole or in

significant part as nonbinding international agreements. Despite their prevalence

and importance, nonbinding international agreements are not currently subject to

any of the domestic statutory or regulatory requirements that apply to binding

agreements. As a result, they are not centrally monitored or collected within the

executive branch, and they are not systematically reported to Congress or disclosed

to the public.

This Article focuses on three of the most important types of nonbinding

international agreements concluded by the United States: (1) high-level formal

agreements; (2) joint statements and communiques; and (3) nonbinding agreements

concluded by administrative agencies. After describing these categories and their

history, the Article presents the first empirical study of U.S. nonbinding agreements,

Allen M. Singer Professor of Law, University of Chicago Law School.

Learned Hand Professor, Harvard Law School.

***

Gerard C. and Bernice Latrobe Smith Professor of International Law, Yale Law School.

For excellent research assistance, we thank Josh Asabor, Tilly Brooks, Patrick Byxbee,

Yilin Chen, Ben Daus-Haberle, Eliane Holmlund, Simon Jerome, Tori Keller, Madison Phillips,

Annabel Remudo, Nathan Stull, Danielle Tyukody, and Kaylee Walsh. We also thank Ayoub

Ouederni for his outstanding assistance analyzing and presenting the data. We thank the many

scholars, lawyers, and government officials from around the world who provided us with insights into

the process for making nonbinding agreements. For assistance with the FOIA requests to more than

twenty federal agencies and lawsuits against the Departments of State, Defense, and Homeland

Security, we thank Daniel Betancourt, Jackson Busch, Charles Crain, Roman Leal, Abby Lemert,

David Schultz, Sruthi Venkatachalam, Kataeya Wooten, Brianna Yates, and especially Michael

Linhorst of the Media Freedom and Information Access Clinic at Yale Law School. For helpful

comments and suggestions, we thank Helmut Aust, Jean Galbraith, Duncan Hollis, Thomas

Kleinlein, Tim Meyer, Kal Raustiala, Michael Reisman, Ryan Scoville, David Zaring, and

participants in faculty workshops at the University of Chicago Law School and Yale Law

School.

The Rise of Nonbinding International Agreements [1-Feb-22

drawing on two new databases that together include more than 2100 nonbinding

agreements. Based on this study, the Article argues that many of the concerns that

prompted Congress to regulate binding executive agreements starting in the 1970s

also apply to nonbinding agreements. Finally, drawing in part on insights obtained

from a comparative assessment of the practices and reform discussions taking place

in other countries, the Article suggests legal changes designed to enhance

coordination and accountability.

CONTENTS

INTRODUCTION ............................................................................................................ 3

I. NONBINDING INTERNATIONAL AGREEMENTS IN U.S. LAW AND PRACTICE ... 9

A. What is a Nonbinding International Agreement? ....................................... 9

B. Nonbinding International Agreements in U.S. Law ................................. 12

C. The Modern Forms of Nonbinding International Agreements ................. 21

II. AGENCY NONBINDING AGREEMENTS: AN EMPIRICAL ANALYSIS ................ 29

A. Why Agencies Conclude Nonbinding Agreements ................................. 29

B. Analyzing Agency Nonbinding Agreements ........................................... 36

III. A COMPARATIVE PERSPECTIVE ......................................................................... 50

A. The Global Rise of Nonbinding Agreements ........................................... 50

B. Coordination ............................................................................................. 53

C. Transparency ............................................................................................ 56

D. Legislative Participation ........................................................................... 58

E. Summary .................................................................................................. 60

IV. LEGAL REFORM ................................................................................................... 61

A. Nonbinding Agreements and Domestic Delegation ................................. 62

B. Accountability Issues: Coordination, Reporting, Transparency .............. 63

C. International Best Practices ...................................................................... 75

CONCLUSION .............................................................................................................. 77

1-Feb-22] The Rise of Nonbinding International Agreements

INTRODUCTION

The only process specified in the Constitution for making international

agreements is the one set forth in Article II, which requires that presidents obtain the

advice and consent of two-thirds of the Senators present.

This treaty process,

however, has been dying a slow death for decades.

It has been replaced in part by

various forms of “executive agreements” that are authorized by a statute, a prior

treaty, or the president’s independent constitutional authority.

Executive

agreements, like treaties, are binding under international law. While executive

agreements are increasingly used as substitutes for treaties, their numbers too have

declined in recent years.

As treaties and executive agreements have declined, another form of

international cooperation has grown in prominence: nonbinding international

agreements. A nonbinding international agreement is an agreement between two or

more sovereign states, or between a state and an international organization, that is

not governed by international law. Whether an agreement is “binding” or not

determines whether it triggers a range of formal rules under international law,

including rules relating to compliance and state responsibility for breach. This is

why international lawyers sharply distinguish the two categories. Yet because the

international legal system often lacks centralized adjudication or enforcement, in

practice nonbinding agreements operate in ways functionally similar to many

binding agreements—that is, they rely on informal enforcement mechanisms such as

coordination, reciprocity, and reputation.

See U.S. CONST. art. II, § 2.

See Oona A. Hathaway, Curtis A. Bradley & Jack L. Goldsmith, The Failed Transparency

Regime for Executive Agreements: An Empirical and Normative Analysis, 134 HARV. L. REV. 629,

632 (2020) (noting that the Clinton administration submitted approximately 23 treaties per year, the

George W. Bush administration submitted around twelve per year, the Obama administration

submitted around five per year, and the Trump administration submitted only five treaties in Trump’s

first three and a half years in office). In the first year of his presidency, President Biden submitted one

treaty to the Senate—the Kigali Amendment to the Montreal Protocol on Substances that Deplete the

Ozone Layer. That year, the Senate consented to no treaties.

See, e.g., Oona A. Hathaway, Presidential Power Over International Law, 119 YALE L.J.

140 (2009) (documenting “a little noticed transformation during the last half-century in the way

international law is made in the United States” from Article II treaties to executive agreements).

President Bill Clinton made an average of 257 executive agreements per year, President

George W. Bush 230 per year, and President Barack Obama 148 per year. President Donald Trump

concluded just 68 from when he took office until February 2018. See Jeffrey S. Peake, The Decline of

Treaties? Obama, Trump, and the Politics of International Agreements 40 tbl.1 (Apr. 6, 2018)

(unpublished manuscript).

The Rise of Nonbinding International Agreements [1-Feb-22

It is difficult to overstate how important nonbinding agreements have become

to U.S. foreign policy. Almost all of the most consequential (and often

controversial) international agreements made by the last three presidential

administrations were nonbinding.

Recent examples of U.S. nonbinding agreements

include the EECD/G20 agreement on global tax reform; the Glasgow Climate Pact

among almost 200 nations that aims to reduce fossil fuel emissions; the Artemis

Accords, an eight-nation framework for interpreting and implementing provisions of

the binding Outer Space Treaty; the New Atlantic Charter, a United States-United

Kingdom agreement for promoting democratic values and institutions; and the

agreement between the United States and the Taliban concerning the withdrawal of

U.S. forces from Afghanistan, which President Biden controversially implemented in

August 2021. These high-profile agreements are the tip of the iceberg of a vast

agreement-making practice that largely flies under the radar of public and

congressional scrutiny.

The executive branch has many incentives to make agreements nonbinding

rather than binding. In contrast to Article II treaties and executive agreements that

are authorized by statute, the executive branch can make nonbinding international

agreements without the need to obtain congressional authorization or approval.

And

in contrast to executive agreements, which most commentators believe are limited to

matters that relate to the president’s independent constitutional authority, the

executive branch maintains that it can make nonbinding agreements on practically

any topic.

Nonbinding agreements also permit the executive branch to avoid

accountability and transparency mandates. The executive branch has a legal duty to

See infra Section I.C.

See infra Part II.

In responding to a question from a member of Congress about why the Obama

Administration did not use the Article II treaty process when concluding an important nuclear

agreement with Iran, Secretary of State John Kerry explained, “I spent quite a few years trying to get

a lot of treaties through the United States Senate and quite frankly, it has become physically

impossible. That is why. Because you can’t pass a treaty anymore.” Secretary of State John Kerry,

Testimony Before the Committee on Foreign Affairs, House Foreign Affairs Committee (July 28,

2015), https://www.c-span.org/video/?327359-1/secretaries-kerry-moniz-lew-testimony-iran-nuclear-

agreement.

By contrast, under most accounts, the president’s constitutional authority to conclude

binding executive agreements (a) must be tied to an independent presidential authority, (b) is narrower

than the power to enter into Article II treaties and congressional-executive agreements, and (c)

generally encompasses discrete issues such as the recognition of other governments and the settlement

of claims. See Hathaway, Bradley & Goldsmith, supra note 2, at 639-41; see also, e.g., Medellin v.

Texas, 552 U.S. 491, 532 (2008) (referring to “the Executive’s narrow and strictly limited authority to

settle international claims disputes pursuant to an executive agreement”). For a broader view of the

president’s independent authority, see Harold Hongju Koh, Twenty-First-Century International

Lawmaking, 101 GEO. L.J. ONLINE 1 (2012).

1-Feb-22] The Rise of Nonbinding International Agreements

report to Congress all binding agreements other than Article II treaties and to publish

the important ones.

But it can skirt these duties altogether by making a nonbinding

agreement, which it need not report or publish. In a world where foreign policy

challenges persist but Congress is gridlocked, it is no surprise that the executive

branch is drawn to a form of agreement that it can make on any topic, and without

congressional approval or accountability.

By all indications, the executive branch’s use of nonbinding agreements has

been growing for some time. In 2005, amidst the decline in binding agreements, a

lawyer in the State Department Legal Adviser’s Office observed that nonbinding

agreements had shown a “marked increase.”

As this Article documents, that trend

has accelerated since that observation. Moreover, many other countries have

witnessed a similar shift from binding to nonbinding arrangements.

For example,

the legal adviser to Germany’s Federal Foreign Office has noted that “[a]s seemingly

everywhere else, the significance of non-legally binding agreements has consistently

been rising in our practice.”

And the head of the treaty division of Mexico’s

foreign ministry has reported that approximately 70% of Mexico’s international

agreements are now nonbinding.

Previous U.S. scholarship has highlighted the growing phenomenon of

nonbinding international agreements. In 1977, in the wake of the nonbinding

Helsinki Accords, Oscar Schachter wrote a brief comment analyzing the “twilight

existence” of such nonbinding international agreements.

In 2002, Kal Raustiala

observed that agencies were using non-legally binding “Memoranda of

Understanding” to structure “transgovernmental cooperation.”

In 2009, Duncan

See Hathaway, Bradley & Goldsmith, supra note 2, at 645-51.

Robert E. Dalton, National Treaty Law and Practice: United States, in NATIONAL TREATY

LAW AND PRACTICE: DEDICATED TO THE MEMORY OF MONROE LEIGH 765, 767 (Duncan B. Hollis et

al. eds., 2005).

This finding is the result of a survey we conducted of experts and officials from more than

a dozen countries, as well as publicly available materials relating to a number of other countries.

Christophe Eick, Legal Adviser, German Federal Foreign Office, Welcome and Opening

Remarks, https://rm.coe.int/0-1-c-eick-cahdi-ws-opening-remarks/1680a23543.

See Alejandro Rodiles, ITAM School of Law, Mexico, Survey for University of Chicago

Law School Conference on “Non-Binding International Agreements: A Comparative Assessment”

(submitted Sept. 1, 2021) (on file with authors).

See Oscar Schachter, Editorial Comment: The Twilight Existence of Nonbinding

International Agreements, 71 AM. J. INT’L L. 296 (1977).

Kal Raustiala, The Architecture of International Agreements, 43 VA. J. INT’L L. 1, 22-23

(2002). See also Kal Raustiala & David G. Victor, Conclusion, in THE IMPLEMENTATION AND

EFFECTIVENESS OF INTERNATIONAL ENVIRONMENTAL COMMITMENTS (David G. Victor, Kal Raustiala

The Rise of Nonbinding International Agreements [1-Feb-22

Hollis and Joshua Newcomer emphasized the importance of nonbinding agreements

(which they called “political commitments”) in U.S. practice, provided a taxonomy

of these agreements, and considered their constitutional implications.

In 2012,

Harold Koh explained how nonbinding agreements operated in conjunction with

binding instruments to effectuate “layered diplomacy” in U.S. practice.

And in

2014, Jean Galbraith and David Zaring offered a foreign relations law perspective on

nonbinding agreements concluded by administrative agencies (which they called

“soft law agreements”).

These studies provide a valuable starting point for understanding nonbinding

agreements. But none seeks to discern the extent and nature of the U.S. practice of

concluding nonbinding agreements. Nonbinding agreements are extraordinarily

difficult to study because they operate in a law-free zone—no international or

domestic legal rules govern them—and because the systems that track international

agreements do not include nonbinding agreements. Article II treaties are published

by the Senate,

listed in the Treaties in Force compilation prepared by the State

Department,

and registered with the United Nations;

executive agreements are

collected by the State Department and reported to Congress under the Case-Zablocki

Act,

and are published in both public and private databases.

Nonbinding

& Eugene B. Skolnikoff, eds., 1998) (advocating the use of nonbinding agreements in international

environmental cooperation).

See Duncan B. Hollis & Joshua J. Newcomer, “Political” Commitments and the

Constitution, 49 VA. J. INT’L L. 507 (2009).

Koh, supra note 8, at 14.

See Jean Galbraith & David Zaring, Soft Law as Foreign Relations Law, 99 CORNELL L.

REV. 735 (2014). A related literature explores the reasons why nations might conclude nonbinding

agreements (sometimes characterized as “informal” or “soft law” commitments) rather than binding

ones. See, e.g., Anthony Aust, The Theory and Practice of Informal International Instruments, 35

INT’L & COMP. L. Q. 787 (1986); Charles Lipson, Why are Some International Agreements Informal?,

45 INT’L ORG. 495 (1991); Gregory Shaffer & Mark A. Pollack, Hard and Soft Law: What Have We

Learned?, in INTERNATIONAL LAW AND INTERNATIONAL RELATIONS: INSIGHTS FROM

INTERDISCIPLINARY SCHOLARSHIP (Jeffrey L. Duffoff & Mark A. Pollack, eds., 2012).

See Congress.gov (“About Treaty Documents”), https://www.congress.gov/search?

q=%7B%22source%22%3A%22treaties%22%7D.

U.S. Department of State, Treaties in Force, https://www.state.gov/treaties-in-force/.

United Nations Treaty Collection, https://treaties.un.org.

See 1 U.S.C. § 112b.

See KAV Agreements, HEINONLINE, https://home.heinonline.org/titles/World-Treaty-

Library/KAV-Agreements [https://perma.cc/P745-KWCP]; Treaties and Other International Acts

Series (TIAS), U.S. Dep’t of State, https://www.state.gov/tias [https://perma.cc/CFC8-BDSU]. We

1-Feb-22] The Rise of Nonbinding International Agreements

agreements, by contrast, have no central repository and are not subject to any rules

about transparency or publication. Most nonbinding agreements are thus never made

public. And the ones available to the public are scattered across the internet based

on the varying preferences of the dozens of agencies and departments that make

them. Most other countries similarly fail to make their nonbinding agreements

public. This has made nonbinding agreements challenging to study.

To overcome these problems, we took three steps. First, we built the first-

ever database of the nonbinding agreements concluded by U.S. administrative

agencies. Working with a team of research assistants, we gathered all public

information we could obtain on such nonbinding agreements. We then filed

Freedom of Information Act (FOIA) requests with more than twenty federal agencies

that we determined conclude nonbinding agreements, in order to obtain their

nonpublic records. When they failed to respond in a timely manner to our FOIA

request, we sued the Departments of State, Defense, and Homeland Security to

obtain their records of nonbinding agreements. In addition, we built a smaller (but

still substantial) database of a second form of nonbinding agreements—joint

statements and communiques issued following high-level international meetings or

conferences. These efforts have resulted in two databases that together already

include over 2100 nonbinding agreements that we have coded and analyzed to

provide an unprecedented quantitative empirical glimpse into the U.S. nonbinding

agreements practice.

Second, we interviewed government officials in several administrative

agencies about their experiences in connection with the drafting and conclusion of

nonbinding agreements. These interviews provided invaluable information about

why agencies choose to conclude nonbinding agreements and the processes that they

follow. Third, we reached out to experts and officials in other countries to learn

more about how their legal and regulatory systems address nonbinding international

agreements. This gave us a broader comparative perspective from which to view

U.S. practice than prior scholarship.

A key contribution of this Article, then, is to excavate and describe a growing

practice relating to international law. Nonbinding agreements, we show, are not just

an important part of the international agreement landscape; they are, increasingly, the

dominant part. The field of international law—in the United States and globally—

documented in earlier work that these databases are not complete, and we recommended reforms for

improving transparency. Still, the databases do exist. See Hathaway, Bradley & Goldsmith, supra

note 2.

We will continue to update a publicly available database on Dataverse that will include all

current agreements as well as additional agreements obtained through litigation or under FOIA.

The Rise of Nonbinding International Agreements [1-Feb-22

must reorient itself to this new reality. Increasingly, international cooperation is

shaped by commitments that claim not to be law at all. This development has

important ramifications for how international law is taught and studied, both in the

United States and elsewhere, and it raises fundamental questions about the nature of

the international legal system.

The growing importance of nonbinding agreements also raises the question—

largely unaddressed in prior scholarship—about whether and how such agreements

should be regulated domestically.

In the United States, nonbinding agreements in

many cases serve the role once reserved for Article II treaties and binding executive

agreements. And yet they are entirely exempt from the reporting and publication

requirements that apply to binding agreements. As we will explain, in some ways

nonbinding agreements should be treated differently, in part to preserve the

flexibility that such agreements offer. With this in mind, and drawing upon ideas

being discussed in other countries, we recommend reforms to enhance executive

branch accountability and transparency.

Unlike many issues of government accountability and transparency,

legislative reform here is plausible. Congress has often amended the Case-Zablocki

Act since the 1970s, and pending legislation proposes to amend it yet again—in part

to address reform proposals that we made in an earlier article.

The proposed

legislation, if enacted, would also begin to address some nonbinding

agreements. The proposed legislation does not go far enough, but it shows that

Congress is interested in improving accountability and transparency in this context.

Part I of this Article describes the rise of nonbinding international agreements

in U.S. practice and their current lack of legal regulation. Part II presents a novel

empirical account of the nonbinding agreements concluded by U.S. administrative

agencies based on numerous sources. Part III is a comparative analysis of how other

nations are addressing the regulatory challenges presented by nonbinding

agreements. Building on Parts II and III, Part IV considers potential legal reforms

for the United States, with a particular focus on executive branch coordination and

public transparency. The Article concludes with reflections on the implications of

the rise of nonbinding international agreements for the field of international law.

We addressed this question briefly in prior work. See Hathaway, Bradley & Goldsmith,

supra note 2, at 708-10. See also Ryan Harrington, A Remedy for Congressional Exclusion from

Contemporary International Agreement Making, 118 W. VA. L. REV. 1211, 1236-42 (2016)

(discussing how the Case-Zablocki Act could be construed to apply to nonbinding agreements).

See infra note 224.

1-Feb-22] The Rise of Nonbinding International Agreements

I. NONBINDING INTERNATIONAL AGREEMENTS IN U.S. LAW AND PRACTICE

A nonbinding international agreement can be bilateral or multilateral and can

take many forms.

A common element among all forms of these agreements is that

they are not governed by international law—a characteristic that has implications for

why nations make them and how they operate in practice. This Part provides

background on nonbinding agreements made by the United States to set the stage for

the empirical, comparative, and normative analysis that follows. It begins by

defining nonbinding agreements. It then explains the historical use of these

agreements by the United States and their place in the U.S. domestic legal system.

Finally, it examines contemporary U.S. practice concerning nonbinding agreements

and organizes the agreements into three categories for purposes of analysis.

A. What is a Nonbinding International Agreement?

A nonbinding international agreement can best be understood by comparison

to a binding international agreement, which in international law nomenclature is

called a “treaty.” A treaty is “an international agreement concluded between States. .

. and governed by international law.”

Important legal consequences of a legally

binding treaty include pacta sunt servanda (a duty to observe the terms of the treaty),

state responsibility for violations, and legal remedies for breach, such as reparations

and countermeasures.

A nonbinding international agreement is simply an agreement between

nations that is not governed by international law.

Such an agreement imposes no

Different terms have been used to capture what we call “nonbinding international

agreements,” including “political commitments,” “informal agreements,” “informal arrangements,”

“nonbinding arrangements,” “nonbinding documents,” “nonbinding instruments,” “nonbinding

arrangements,” “soft law agreements,” and (in an earlier era) “gentlemen’s agreements.” See, e.g.,

ANTHONY AUST, MODERN TREATY LAW AND PRACTICE 18 (3d ed. 2013); Robert Dalton, Asst. Legal

Adviser for Treaty Affairs, International Documents of a Non-Legally Binding Character, State

Department, Memorandum (Mar. 18, 1994) (copy on file with authors). Although some observers

might think that the word “agreement” connotes bindingness, we use “nonbinding agreements”

because it best reflects the role that these documents play in the international system. The term has,

moreover, been used in recent international discussions of the topic. See infra Part III.

Vienna Convention on the Law of Treaties, supra note 4, art. 2(1)(a) (emphasis added).

Under international law, all international agreements that are governed by international law, including

“executive agreements,” are considered treaties.

See, e.g., AUST, supra note 27, at 315-17.

It is also not governed by domestic law. States make contracts—for example with

corporations concerning investment matters and sometimes with other states—that are governed by

The Rise of Nonbinding International Agreements [1-Feb-22

international legal duty to comply with its terms, and breach or non-compliance with

the agreement implicates no international legal consequences. This does not mean

that nonbinding agreements lack any relationship to binding international law. To

the contrary, nonbinding agreements can serve as the basis for or precursor to

binding instruments made later;

provide interpretive guidance for binding

agreements;

clarify or expand upon the requirements of binding obligations;

embedded or incorporated into a binding obligation or instrument;

and influence

the development of customary international law.

But nonbinding agreements do

not create direct legal obligations and the attendant consequences.

The difference between a binding and a nonbinding agreement is easy to

articulate in theory but distinguishing between the two in practice can be challenging

because there is no universally accepted test for drawing the distinction. One test

domestic law rather than international law. See Organization of American States, Inter-American

Juridical Committee, Guidelines on Binding and Non-Binding Agreements 55-56 (2020),

http://www.oas.org/en/sla/iajc/docs/Guidelines_on_Binding_and_NonBinding_Agreements_publicati

on.pdf [hereinafter OAS Report]. Nonbinding international agreements exclude such contracts.

For example, the 1988 Baltic Sea Ministerial Declaration and the 1992 Baltic Sea

Declaration “paved the way for the 1992 Convention on the Protection of the Marine Environment of

the Baltic Sea Area,” a binding agreement. Andreas Zimmermann & Nora Jauer, Possible Indirect

Legal Effects of Non-legally Binding Instruments, Expert Workshop on Non-Legally Binding

Agreements in International Law (Mar. 26, 2021), at https://rm.coe.int/1-2-zimmermann-indirect-

legal-effects-of-mous-statement/1680a23584.

For example, investment tribunals “rely on non-binding rules . . . to establish procedures

through which to adjudicate disputes in a binding fashion,” and in legally binding decisions the

tribunals sometimes use “non-binding instruments to fill gaps in international investment

agreements.” Tim Meyer, Alternatives to Treaty-Making—Informal Agreements, in THE OXFORD

GUIDE TO TREATIES 59, 64 (Duncan Hollis ed., 2d ed. 2020).

For example, “space-faring states have favored legally nonbinding principles and technical

guidelines that are layered on top of . . . preexisting treaties” related to outer space. Koh, supra note

8, at 15. Similarly, in the environmental context, “decisions of treaty bodies, such as a Conference of

the Parties (COP), are often non-binding but can supplement or expound on binding obligations.”

Meyer, supra note 32, at 64.

In differing ways, this was true of both the Paris Agreement on climate change and the Iran

nuclear deal. See infra text accompanying notes 96-102.

See International Law Commission, Draft conclusions on identification of customary

international law with commentaries (2018), A/73/10, Conclusions 6, 9, and 10. Some scholars claim

that a nonbinding commitment might bind a country under the principle of estoppel, but the point is

not established in national practice. See OAS Report, supra note 30, at 126.

See OAS Report, supra note 30, at 123 (noting that nonbinding agreements do not “trigger

pacta sunt servanda nor any of the secondary international legal effects that follow treaty-making

(e.g., the law of treaties, State responsibility, specialized regimes)”).

1-Feb-22] The Rise of Nonbinding International Agreements

looks predominantly to the intent of the parties.

However, intent is not always easy

to discern. Some nonbinding agreements expressly state that they are nonbinding.

But many do not, in which case intent must be inferred from the language of the

agreement, the circumstances under which it was made, and other contextual

factors.

A second test turns on objective factors. On this view, “the agreement’s

subject-matter, text, and context determine its binding or non-binding status

independent of other evidence as to one or more of its authors’ intentions.”

The intent and objective tests often lead to the same conclusion about the

bindingness of an agreement. But uncertainties in the application of each test,

combined with the fact that different nations follow different tests, mean that nations

sometimes disagree about whether an agreement between them is binding or not.

Several prominent international tribunal cases involved disputes about whether

certain agreements were binding or not.

In the 1990s, the United States considered

certain defense-related memoranda of understanding to be binding agreements, but

its partners (Australia, Canada, and the United Kingdom) regarded them as

nonbinding political commitments.

Similarly, the United States viewed the nuclear

deal with Iran in 2015 as a nonbinding agreement, but Iran insisted that it was a

The intent test was embraced by the International Law Commission in its important mid-

century study of the law of treaties, see, e.g., II Yearbook of the International Law Commission 189

(1966), and by the delegates to the Vienna Convention on the Law of Treaties, U.N. Conference on

the Law of Treaties, Summary Records of Second Session, U.N. Doc. A/CONF.39/11, Add.1, 225,

13. It is the approach used by, among other countries, the United States, see Dalton, supra note 27,

and the United Kingdom, see AUST, supra note 29, at 31.

See, e.g., Dalton, supra note 27 (noting that the test for legal bindingness is “the intent of

the parties, as reflected in the language and context of the document, the circumstances of its

conclusion, and the explanations given by the parties”).

See OAS Report, supra note 30, at 80-81 and notes 127-134; Meyer, supra note 32, at 59.

See, e.g., Maritime delimitation and territorial questions between Qatar and Bahrain (Qatar

v. Bahrain), 1994 I.C.J. 112, ¶¶ 20–22 (Jurisdiction and Admissibility); Agean Sea Continental Shelf

Case (Greece v. Turk.), 1978 I.C.J. 3, ¶¶ 97–100 (Dec. 19); Delimitation of the Maritime Boundary in

the Bay of Bengal (Bangladesh/Myanmar), Case No. 16, Judgement of Mar. 14, 2012, ITLOS Rep. 4,

¶¶ 61–69.

See John H. McNeill, International Agreements: Recent US-UK Practice Concerning the

Memorandum of Understanding, 88 AM. J. INT’L L. 821 (1994).

The Rise of Nonbinding International Agreements [1-Feb-22

binding agreement.

And more recently, the United States and Mexico disagreed

about the bindingness of an agreement concerning migration.

The final definitional point is that, for our purposes, the fact that an

agreement is nonbinding does not necessarily mean that it is “soft law.” The two

concepts are sometimes used interchangeably, especially in scholarly discussions.

But soft law is often used as a broader term to capture agreements and international

policies that impose weak or uncertain obligations through some combination of

nonbindingness, vague or hortatory terms, shallow obligations, and a lack of

enforcement mechanisms.

For purposes of this Article, a nonbinding international

agreement is simply one that is not governed by international law, and it can include

agreements with vague or precise terms, shallow or deep obligations, and

enforcement mechanisms or no such mechanism.

B. Nonbinding International Agreements in U.S. Law

This section reviews how nonbinding agreements fit within the framework of

U.S. domestic law. It begins with a brief description of the history of such

agreements in the United States, and it then turns to the president’s domestic

See STEPHEN P. MULLIGAN, CONG. RESEARCH SERV., LSB10134, WITHDRAWAL FROM THE

IRAN NUCLEAR DEAL: LEGAL AUTHORITIES AND IMPLICATIONS 1 (2018) (stating the U.S. view); Eline

Gordts, Iran’s Foreign Minister to U.S. Senators: ‘The World is Not the United States’, HUFFINGTON

POST (Mar. 9, 2015), https://www.huffpost.com/entry/zarif-senators-letter_n_6834296 (explaining the

Iranian view).

See Joint Declaration and Supplementary Agreement Between the United States of

America and Mexico, Mex.-U.S., June 7, 2019, T.I.A.S. No. 19-607; Rachel Withers, Mexico

Releases the Full Text of Trump’s Immigration “Deal,” VOX, June 15, 2019,

https://www.vox.com/2019/6/15/18680129/us-mexico-immigration-deal-release-trump-tariff. In

response to a query from Senator Menendez, the State Department declared the agreement binding.

Letter from Mary Elizabeth Taylor, Assistant Secretary, Bureau of Legislative Affairs, U.S.

Department of State, to Robert Menendez, Ranking Member, Committee on Foreign Relations,

United States Senate (Sept. 9, 2019). Yet, according to a U.S. government lawyer, “the Government

of Mexico considers it non-binding.” Email from U.S. Government Lawyer to Oona A. Hathaway

(June 5, 2021).

See, e.g., Andrew T. Guzman & Timothy L. Meyer, International Soft Law, 2 J. LEGAL

ANALYSIS 171, 201-21 (2010); see also Dinah L. Shelton, Soft Law, in ROUTLEDGE HANDBOOK OF

INTERNATIONAL LAW 68 (David Armstrong ed., 2008).

See Kal Raustiala, Form and Substance in International Agreements, 99 AM. J. INT’L L.

581 (2005); Kenneth W. Abbott & Duncan Snidal, Hard and Soft Law in International Governance,

54 INT’L ORG. 421, 422 (2000); W. Michael Reisman, The Concept and Functions of Soft Law in

International Politics, in 1 ESSAYS IN HONOUR OF JUDGE TASLIM OLAWALE ELIAS 135 (Emmanuel G.

E & Prince Bola A. Ajibola, S.A.N. eds., 1992).

1-Feb-22] The Rise of Nonbinding International Agreements

authority to make them, their status in the domestic legal system, and their lack of

domestic regulation.

1. A Brief History of Nonbinding Agreements. The history of nonbinding

international agreements in the United States is murky. Diplomatic letters and other

papers effectuated informal agreements with other nations since the Founding. But a

distinct category of what we today mean by nonbinding international agreements did

not clearly emerge until the twentieth century.

Before then, the executive branch

made hundreds of agreements on its own authority. But there appears to have been

little discussion of whether these agreements were binding or nonbinding under

international law.

The issue became more salient in the early twentieth century as the Senate

began to complain about the executive branch’s increasingly ambitious use of the

executive agreement power.

The executive branch defended some agreements on

the ground that they lasted only as long as the executive branch chose to enforce

them and did not bind future administrations or the nation as a whole. Theodore

Roosevelt invoked this theory to justify the 1905 agreement he made with the

Dominican Republic for administering customs houses in Santo Domingo.

William Howard Taft made a similar argument, when he was Roosevelt’s Secretary

of War, to justify an agreement that defined the relative jurisdictions in cities at both

ends of the Panama Canal.

Taft described the agreement as a modus vivendi (or

temporary agreement) that was “revocable at will,” but it lasted beyond the

Roosevelt administration because subsequent administrations continued to enforce

There were concepts akin to nonbinding agreements much earlier. See, e.g., EMER DE

VATTEL, THE LAW OF NATIONS 355 (Béla Kapossy & Richard Whatmore eds., Liberty Fund 2012)

(1797) (distinguishing a “personal alliance” or “personal treaty,” which “expires with him who

contracted it,” from a “real alliance” or “real treaty,” which “attaches to the body of the state, and

subsists as long as the state, unless the period of its duration has been limited”).

For example, the Senate reacted testily to President McKinley’s use of an executive

agreement to “arrange[] the Spanish withdrawal from Puerto Rico, Cuba, and other former

possessions” at the termination of the Spanish-American war, and to early twentieth century

presidents’ agreements establishing U.S. policy in the far east, including the Open Door Policy, the

intervention in the Boxer Rebellion, and several agreements with Japan. See Bruce Ackerman &

David Golove, Is NAFTA Constitutional?, 108 HARV. L. REV. 799, 818 (1995). See generally Michael

D. Ramsey, The Treaty and Its Rivals: Making International Agreements in U.S. Law and Practice, in

SUPREME LAW OF THE LAND? DEBATING THE CONTEMPORARY EFFECTS OF TREATIES WITHIN THE

LEGAL SYSTEM OF THE UNITED STATES (Paul Dubinsky, Gregory Fox & Brad Roth eds., 2017)

(describing “enormous changes in U.S. foreign relations” by end of nineteenth century that led to

“new forms of agreement-making,” including “the rise of nonbinding agreements”).

See THEODORE ROOSEVELT, THEODORE ROOSEVELT, AN AUTOBIOGRAPHY 551 (1913).

See WILLIAM HOWARD TAFT, OUR CHIEF MAGISTRATE AND HIS POWERS 112 (1916).

The Rise of Nonbinding International Agreements [1-Feb-22

it.

Similarly, Secretary of State Robert Lansing explained that the 1917 Lansing-

Isgii agreement—which resolved various U.S.-Japan issues relating to China—

lacked “any binding force” on the United States, and was “simply a declaration of . .

. the policy of this Government, as long as the President and the State Department

want to continue that policy.”

Despite these early precedents, commentators in the first third of the

twentieth century disagreed about which of hundreds of other agreements made by

the executive branch were binding on the nation rather than simply a policy of a

particular administration. Quincy Wright’s influential 1922 book, The Control of

American Foreign Relations, argued that executive agreements that settled claims

and possibly agreements made under the Commander in Chief power were binding

on the nation under international law.

Wright maintained that other types of

executive agreements—which he variously labeled protocols, modus vivendi,

“gentlemen’s agreements,” administrative agreements, or agreements that define

executive policy—are “binding only on the president that makes them.”

He noted

that “presumably the foreign government would have no ground for objection if a

subsequent President discontinued such an executive agreement.”

Wright’s

assessment was influential, but other commentators reached different conclusions.

Investigation of Panama Canal Matters: Hearing Before the S. Comm. on Interoceanic

Canals, 59th Cong. 2590 (1907) (cable of Secretary of War William Howard Taft to Secretary of

State John Hay); see also id. at 2741-42 (Senator Morgan noting that the jurisdictional boundaries are

“settled here temporarily and provisionally by a modus vivendi”).

Treaty of Peace with Germany: Hearing Before the S. Comm. on Foreign Rels., 66th

Cong. 219 (1919) (testimony of Secretary of State Robert Lansing). President Harding later described

the “so-called Lansing-Ishii agreement” as an “exchange of notes [that], in the nature of things, did

not constitute anything more than a declaration of Executive policy.” GREEN HAYWOOD

HACKWORTH, 5 DIGEST OF INTERNATIONAL LAW 431 (1943). President Wilson said it was “an

understanding,” not an agreement.

QUINCY WRIGHT, THE CONTROL OF AMERICAN FOREIGN RELATIONS 240-44 (1922).

Id. at 238; see also id. at 54, 235, 237, 243. It appears from context in these passages that

Wright was not using the term “binding” to suggest that international law governed these agreements,

but rather to suggest that whatever political or moral obligation they imposed applied only to the

administration that made them.

Id. at 238.

See, e.g., GEORGE SUTHERLAND, CONSTITUTIONAL POWER AND WORLD AFFAIRS 120-21

(1919) (distinguishing executive agreements binding on the nation from those that “constitute[e] only

a moral obligation”); CHARLES HENRY BUTLER, TREATY-MAKING POWER OF THE UNITED STATES §

463 (1902) (concluding that “protocols,” Butler’s term for many executive agreements, “are binding

in a moral sense upon the Executive department of the administration making them,” but do not bind

the legislature, and “[i]t is doubtful if they are binding even morally upon any administration other

than that which entered into them”); Harry Swain Todd, The President’s Power to Make International

Agreements, 11 CONST. REV. 160, 162 (1927) (noting that the “question as to the binding force of an

1-Feb-22] The Rise of Nonbinding International Agreements

The wide range of positions was possible because the executive branch was rarely

clear, beyond the few precedents mentioned above, about which agreements were

binding on the nation. The U.S. Supreme Court never addressed the issue.

The meaning and scope of nonbinding international agreements within U.S.

practice began to clarify in the middle decades of the twentieth century. The

increased use and importance of executive agreements starting in the 1930s sparked

a scholarly debate that highlighted the massive number and array of agreements

made on the President’s authority alone and raised anew questions about which were

binding.

In the 1940s, Presidents Roosevelt and Truman announced the Atlantic

Charter and the Yalta and Potsdam agreements (concerning aims and principles

relating to World War II and its aftermath) on their own authority. The United States

claimed that all three were nonbinding under international law, but some countries

and scholars disagreed about the latter two.

In 1949, the International Law

Commission began work on the law of treaties that would result in 1969 in the

Vienna Convention on the Law of Treaties. That Convention’s definition of a treaty

as “an international agreement . . . governed by international law” aimed to exclude

nonbinding international agreements.

The question of which U.S. international agreements were binding and which

were nonbinding assumed new importance with the passage of the Case-Zablocki

Act in 1972. That Act required the Secretary of State to transmit to Congress “the

executive agreement is not easy to discuss” and is “not entirely settled in the minds of jurists”);

Charles Cheney Hyde, Agreements of the United States Other Than Treaties, in 17 GREEN BAG 229

(1905) (an agreement made by or at direction of the president “is in most cases a binding one upon the

nation.”); John W. Foster, The Treaty-Making Power Under the Constitution, 11 YALE L.J. 69, 79

(1901) (concluding that that “there are certain acts of an international character, binding the

Government, which the President may perform without the interposition of the Senate”).

The closest case we have found is United States ex rel Angarica v. Bayard, 127 U.S. 251,

261 (1888), where the Court assumed that an exchange of letters between the Secretary of State and

the Mexican equivalent was not binding on successors.

Compare Myres S. McDougal & Asher Lans, Treaties and Congressional-Executive or

Presidential Agreements: Interchangeable Instruments of National Policy: I, 54 YALE L.J. 181, 197-

99, 198 nn.15 & 17, 318-23 (1945) (maintaining that with a few exceptions, all executive agreements

are presumptively binding on the United States under international law), with Edwin Borchard, Shall

the Executive Agreement Replace the Treaty?, 53 YALE L.J. 664, 678-80 (1944) (suggesting that most

executive agreements bind only the administration that makes them). This debate was also influenced

by the Supreme Court’s decisions in United States v. Pink, 315 U.S. 203 (1942) and United States v.

Belmont, 301 U.S. 324 (1937), which made clear that some sole executive agreements could be

binding and supreme federal law.

See Schachter, supra note 14, at 297-98 & nn.10-11 (collecting sources).

See Fritz Münch, Comments on the 1968 Draft Convention on the Law of Treaties: Non-

Binding Agreements, 29 ZAÖRV 1 (1969); Schachter, supra note 14; Dalton, supra note 27.

The Rise of Nonbinding International Agreements [1-Feb-22

text of any international agreement . . . other than a treaty, to which the United States

is a party.”

Disagreement immediately arose about what types of agreements were

included within this obligation. In 1976, the Legal Adviser to the State Department

established a five-part test for resolving this issue, the “central requirement” of

which was whether the parties to the agreement intended it to be binding under

international law.

These criteria were reflected in federal regulations beginning in

1981.

At least since that time, party intent has been the primary touchstone in U.S.

practice in determining whether an agreement is binding or nonbinding under

international law.

The State Department has issued modest guidance about

“formal, stylistic, and linguistic features” that an agreement should include and

exclude to ensure that it is nonbinding.

But the executive branch has never

explained in a comprehensive way which executive agreements are binding and

which are nonbinding.

2. Domestic Authorization to Make Nonbinding Agreements. In practice the

executive branch appears to assert the authority to make nonbinding agreements with

other countries on practically any topic. While few observers in modern times have

questioned this practice,

there is no settled account of the constitutional basis for it.

The text of the Constitution does not speak directly to the issue, and neither the

Supreme Court nor the Justice Department’s Office of Legal Counsel has addressed

it.

1 U.S.C § 112b(a).

Foreign Relations Authorization Act: Hearing on S. 1190 Before the Subcomm. on Int’l

Operations of the S. Comm. on Foreign Rels., 95th Cong. 294 (1977) (memorandum by Legal Adviser

of Dep’t of State Monroe Leigh to key department personnel) [hereinafter Leigh Memorandum]. The

secondary requirements were significance, specificity, two or more parties, and form. Id. at 293-94.

The regulations were promulgated pursuant to a 1979 amendment to the Case-Zablocki

Act and are codified today at 22 C.F.R. § 181.2(a)(1).

See Dalton, supra note 27 (providing examples from the 1970s and 1980s).

See U.S. Department of State, Guidance on Non-Binding Documents, https://2009-

2017.state.gov/s/l/treaty/guidance/index.htm [hereinafter State Department Guidance]. To take one of

several examples, the guidance states that “we advise that negotiators avoid terms such as ‘shall’,

‘agree’, or ‘undertake’ [in nonbinding agreements, and] “we have urged that terms such as ‘should’ or

‘intend to’ or ‘expect to’ be utilized in a non-binding document.”

Hollis & Newcomer, supra note 16, make normative arguments against the conventional

wisdom, see id. at 575, as does (briefly) Michael D. Ramsey, Executive Agreements and the

(Non)treaty Power, 77 N.C. L. REV. 133, 143 (1998). But see Michael D. Ramsey, Evading the

Treaty Power?: The Constitutionality of Nonbinding Agreements, 11 FIU L. REV. 371, 375-76 (2016)

(concluding that the “Constitution’s text and practice thus appear to allow Presidents to make

nonbinding agreements” but adding that “the President has a constitutional obligation to assure that a

purportedly nonbinding agreement is clearly and unequivocally nonbinding under international law”).

1-Feb-22] The Rise of Nonbinding International Agreements

The chief constitutional foundation for nonbinding agreements is the

president’s power to conduct the nation’s diplomatic relations and to speak on behalf

of the United States in the conduct of these relations.

This power derives in part

from the president’s textual authority (notably with Senate consent) to “make

Treaties” and to “appoint Ambassadors . . . and Consuls,” from the president’s power

to “receive Ambassadors and other public Ministers,” from the president’s status as

chief executive, and, sometimes, from the president’s duty to “take care to faithfully

execute the Law.”

The power has also been recognized in practice since the

Founding and flows from what Professor Louis Henkin described as the president’s

“control of the foreign relations ‘apparatus’”—the diplomatic machinery that

includes the State Department and other executive departments, U.S. ambassadors,

consuls, and ministers, and the president’s personal agents.

These sources of

authority—the implications of constitutional text, longstanding historical practice,

and control over the diplomatic machinery—provide the foundation for a number of

the president’s most important foreign relations powers.

The power to make

nonbinding international agreements is arguably best understood to flow from these

sources as well.

One way to view a nonbinding agreement is as a statement of U.S. foreign

policy, in coordination with other governments, that any party can opt out of

unilaterally. Viewed this way, the power to make such agreements falls within the

president’s power to announce U.S. foreign policy positions. Indeed, some

nonbinding international agreements might be viewed as a form of diplomatic speech

between the United States and foreign governments about how the parties intend to

act on matters that they have competence to execute. Such speech occurs countless

times every day in numerous contexts and in manifold forms. The president and his

or her subordinates could not exercise their diplomatic powers or meet their

See, e.g., Zivotofsky v. Kerry, 576 U.S. 1, 21 (2015) (president has “a unique role in

communicating with foreign governments”); United States v. Louisiana, 363 U.S. 1, 35 (1960)

(president is “the constitutional representative of the United States in its dealings with foreign

nations”).

U.S. CONST. art. II, § 1, cl. 1; id., art. II, § 2, cl. 2; id., art. II, § 3; Constitutionality of

Section 7054 of the Fiscal Year 2009 Foreign Appropriations Act, 33 Op. O.L.C., 2009 WL 2810454

(June 1, 2009) (describing various sources for the President’s authority to conduct diplomatic

relations).

LOUIS HENKIN, FOREIGN AFFAIRS AND THE UNITED STATES CONSTITUTION 41 (2d ed.

1996). A fourth possible basis is the Article II Vesting Clause. See Ramsey, Evading the Treaty

Power, supra note 65.

These powers include the power to announce U.S. foreign policy positions, to state the

U.S. interpretation of rules of customary international law, to assert rights on behalf of the nation and

its citizens and to claim reparations, and to recognize foreign governments and their territories. See

HENKIN, supra note 68, at 41-45.

The Rise of Nonbinding International Agreements [1-Feb-22

diplomatic responsibilities without communication of this sort. This communication

can be highly informal and unimportant (such as an email agreeing to meet to discuss

a small matter). It can be more formal and more important, such as a joint

communique stating common positions and aims on certain policy issues. And, at

the opposite end of the spectrum from the casual email, it can be a formal,

complicated, and important but nonetheless nonbinding agreement signed by heads

of state. The entire spectrum is encompassed by the president’s power over

diplomatic communications for the United States.

3. Nonbinding Agreements and Domestic Law. Nonbinding agreements do

not have the status of domestic federal law. They are by definition not “law.” And

they do not fit within the instruments identified in the Supremacy Clause—the

Constitution, treaties, or “laws of the United States . . . made in pursuance” of the

Constitution.

The Supreme Court has recognized that some “sole” executive

agreements operate as federal law that preempts state law.

But these decisions to

date have been limited to legally binding executive agreements.

And the Court has

emphasized, in the context of the president’s long-established power to settle claims

via executive agreement, that the power to make binding domestic law via executive

agreements is “narrow and strictly limited.”

The Court has also more generally

emphasized that the president in our system is not a lawmaker.

Given that the

scope of the president’s power to make nonbinding agreements is practically

limitless, it would be an unfathomable expansion of presidential power, and

disruption of the domestic legal system, if these instruments also had the status of

domestic law. These are some of the reasons why no one has ever seriously

suggested that nonbinding agreements have that status.

The president has sometimes even been described as the “sole organ of the federal

government in the field of international relations.” United States v. Curtiss-Wright Export Corp., 288

U.S. 378, 319 (1936). This description is now generally regarded as an overstatement. See Zivotofsky

v. Kerry, 135 S. Ct. 2076, 2090 (2015) (noting that “[i]t is not for the President alone to determine the

whole content of the Nation’s foreign policy”).

U.S. CONST. art. VI, cl. 2.

See United States v. Pink, 315 U.S. 203 (1942); United States v. Belmont, 301 U.S. 324

(1937).

For example, the Litvinov agreement that was at issue in both the Pink and Belmont

decisions was a binding sole executive agreement. See CONGRESSIONAL RESEARCH LIBRARY,

TREATIES AND OTHER INTERNATIONAL AGREEMENTS: THE ROLE OF THE UNITED STATES SENATE 88

(2001) [hereinafter CRS Study].

Medellín v. Texas, 552 U.S. 491 (2008).

See, e.g., Medellin v. Texas, 554 U.S. 759 (2008); Youngstown Sheet & Tube Co. v.

Sawyer, 343 U.S. 579 (1952).

1-Feb-22] The Rise of Nonbinding International Agreements

Nonbinding agreements can, however, influence or become part of domestic

law. First, executive branch officials often implement or comply with nonbinding

agreements within the executive branch bureaucracy.

Second, Congress can

incorporate nonbinding agreements into binding domestic legislation. For example,

Congress in the Clean Diamond Trade Act implemented the Kimberley Process

Certification Scheme, a nonbinding agreement that aims to remove conflict

diamonds from the global supply chain.

Third, administrative agencies can through

rulemaking and other instruments implement nonbinding agreements domestically.

For example, the international banking rules reflected in the nonbinding Basel

Accords “are the basis for binding domestic regulations of the banking industry.”

Fourth, it is conceivable that some elements of nonbinding agreements might

preempt state law under the theory of executive branch foreign policy preemption

suggested in American Insurance Ass’n v. Garamendi.

4. Lack of Domestic Regulation. Another remarkable characteristic of

nonbinding international agreements is that, despite their prevalence and importance,

they are not currently subject to any of the statutory or regulatory requirements that

apply to binding agreements. Congress long ago imposed transparency and

accountability requirements on the executive branch with respect to binding

international agreements. Under the Case-Zablocki Act, the executive must report to

Congress “any international agreement . . . other than a treaty” within sixty days after

it takes effect.

There is also a statutory obligation to publish important agreements

See Hollis & Newcomer, supra note 16, at 542 (noting that “[o]fficials regularly conform

U.S. foreign policy to existing political commitments”); cf. Richard R. Baxter, International Law in

“Her Infinite Variety”, 29 INT’L L. & COMP. L.Q. 549, 556 (1980) (“Bureaucrats follow through on

what they have said that they would do through force of bureaucratic habit”).

See 19 U.S.C. §§ 3901 et seq.; Interlaken Declaration on Kimberley Process Certification

Scheme for Rough Diamonds (Nov. 5, 2002), at https://www.kpcivilsociety.org/wp-

content/uploads/2019/10/KP-InterlakenDeclaration-KPCS-1102.pdf. See also Mallory Stewart, Are

Treaties Always Necessary? How U.S. Domestic Law Can Give Teeth to Non-Binding International

Commitments, 2010 ASIL PROCEEDINGS 189.

Meyer, supra note 39, at 64; see also Galbraith & Zaring, supra note 22.

539 U.S. 396, 415 (2003). The Court held in Garamendi that the executive branch foreign

policy reflected in a legally binding sole executive agreement that called for the establishment of a

fund to compensate victims of Nazi persecution preempted a California state insurance recovery law.

Some commentators read Garamendi as recognizing an “independent presidential power to override

state laws that interfere with executive branch foreign policy.” Michael D. Ramsey & Brannon P.

Denning, American Insurance Association v. Garamendi and Executive Preemption in Foreign

Affairs, 46 WM. & MARY L. REV. 825 (2004) (emphasis added). If so, a court might conceivably

derive such a policy from a nonbinding agreement.

1 U.S.C. § 112b(a).

The Rise of Nonbinding International Agreements [1-Feb-22

on the State Department’s website within 180 days after they take effect.

As we

have documented elsewhere, there are a number of deficiencies in this regime,

and

Congress is currently considering legislation that would bolster it.

Regulations adopted by the State Department to implement the requirements

make clear that they apply only if the parties to an agreement “intend their

undertaking to be legally binding, and not merely of political or personal effect.”

The regulations further state that “[d]ocuments intended to have political or moral

weight, but not intended to be legally binding, are not international agreements,” and

they give as an example the Helsinki Accords. More than twenty years ago, the

Congressional Research Service noted that “some believe these kinds of

[nonbinding] arrangements could represent a large loophole” in the reporting

regime.

Since then, the phenomenon of nonbinding agreements has grown

significantly.

Within the executive branch, the usual standards for approving and keeping

track of executive agreements do not apply to nonbinding agreements. The State

Department’s “C-175” process, named after a circular issued in 1955, is designed to

“facilitate[] the application of orderly and uniform measures to the negotiation,

conclusion, reporting, publication, and registration of U.S. treaties and international

agreements, and facilitate[] the maintenance of complete and accurate records on

such agreements.”

Pursuant to this process, before negotiating an agreement, an

executive agency must obtain pre-approval from the State Department.

After the

agreement is negotiated, the agency must receive additional C-175 approval from

State to conclude the agreement. Furthermore, after conclusion of the agreement, the

1 U.S.C. § 112a(d). For additional discussion of the reporting and publication obligations,

see Hathaway, Bradley & Goldsmith, supra note 2, at 645-54. Classified agreements are reported to

congressional committees but not published.

See generally id.

22 C.F.R. § 181.2(a)(1). Even before the adoption of the regulations, the State Department

had taken the position that the Case-Zablocki Act reporting obligation did not apply to nonbinding

agreements. See Schachter, supra note 14, at 302 (1977) (quoting from a memorandum by the State

Department Legal Adviser to “Key Department Personnel” dated March 12, 1976 on “Case Act

Procedures and Department of State Criteria for Deciding What Constitutes an International

Agreement”).

CRS Study, supra note 73, at 231.

11 U.S. Dep’t of State, Foreign Affairs Manual § 721,

https://fam.state.gov/fam/11fam/11fam0720.html.

Congress has similarly directed in the Case-Zablocki Act that “[n]otwithstanding any other

provision of law, an international agreement may not be signed or otherwise concluded on behalf of

the United States without prior consultation with the Secretary of State.” 1 U.S.C. § 112b(c).

1-Feb-22] The Rise of Nonbinding International Agreements

agency is supposed to transmit a copy to State for central collection.

But this

centralized approval and collection process applies only to binding agreements.

In sum, even though nonbinding agreements can be as consequential as

binding agreements and often resemble them in form and enforcement, they are not

subject to any of the legal regulations that apply to binding agreements.

C. The Modern Forms of Nonbinding International Agreements

Nonbinding international agreements arise in a wide variety of institutional

settings and come in a wide variety of forms. A major challenge to analyzing them

is defining their scope. One cannot hope to be comprehensive, since nonbinding

agreements can include all manner of informal diplomatic communication, including

emails, phone calls, and everyday cables that foster relatively trivial forms of

international cooperation and coordination, including about lunch dates and future

communications.

For purposes of the analysis in this Article, we consider three of the most

consequential forms of nonbinding agreements: (1) high-level formal agreements; (2)

joint statements and communiques issued by state representatives; and (3)

nonbinding agreements between U.S. administrative agencies and their foreign

counterparts that promote various forms of international regulatory cooperation.

While these categories capture three distinctive types, there is variation within them,

the lines between them are not always sharp, and they leave out less formal and less

consequential forms of nonbinding agreements.

Nonetheless, these categories

provide a framework for understanding the main forms of nonbinding agreements

concluded by the United States.

1. High-Level Formal Agreements. The first category involves formal and

often elaborate agreements, usually about important matters agreed to by senior

governmental officials, that commit the nation (as opposed to an agency or other

subunit) to a course of action. These agreements often have many of the trappings of

See 22 C.F.R. § 181.3(b).

See State Dep’t, Office of the Legal Adviser, Circular 175 Procedure, https://2009-

2017.state.gov/s/l/treaty/c175/index.htm (“The Circular 175 procedure does not apply to documents

that are not binding under international law. Thus, statements of intent or documents of a political

nature not intended to be legally binding are not covered by the Circular 175 procedure.”); see also 22

C.F.R. § 181.4. If there is a question about whether an agreement is binding, agencies are supposed to

submit the agreement to State no later than twenty days after signing it for a determination. See 22

C.F.R. § 181.3(c). But it is unclear how this obligation is enforced.

For example, our three categories exclude oral agreements, exchanges of letters that do not

result in a joint text, and standard-setting rules of international organizations.

The Rise of Nonbinding International Agreements [1-Feb-22

binding international agreements, such as content organized by articles, entry into

force and termination provisions, and sometimes even dispute resolution provisions.

But the parties nonetheless do not intend the agreements (or significant parts of the

agreements) to be binding under international law. They can be either bilateral or

multilateral.

While these types of nonbinding agreements have become more prevalent,

they are far from new. For example, the 1975 Helsinki Accords, which tempered

Cold War animosities between West and East and became a focal point for dissident

groups in the Soviet Union and its satellite nations that many believe were an

important cause of the fall of the Soviet Union, were nonbinding.

Recent examples

of important nonbinding high-level agreements include a multilateral agreement

known as the Artemis Accords that concerns the conditions for the safe and peaceful

exploration of space,

the OECD/G20 agreement on global tax reform,

and the

nonbinding agreement with the Taliban calling for the United States to withdraw all

forces by the end of May 1, 2021 (later extended to August 31).

This latter

agreement underscores the practical importance of nonbinding agreements even

though they are not enforceable under international law. President Biden explained

that the agreement protected U.S. persons during the withdrawal and emphasized that

if the United States missed the August 31 deadline, the Taliban likely would have

carried out attacks on U.S. troops.

See Daniel C. Thomas, The Helsinki Accords and Political Change in Eastern Europe, in

THE POWER OF HUMAN RIGHTS: INTERNATIONAL NORMS AND DOMESTIC CHANGE 205 (Thomas Risse

et al. eds., 1999). Relatedly, the Organization for Security and Co-operation (“OSCE”) in Europe,

“the world’s largest regional security organization,” grew out of the Helsinki Accords and is

constituted by nonbinding agreements. Who We Are, ORG. FOR SEC. AND COOP. IN EUR.,

https://www.osce.org/who-we-are (last visited July 31, 2021).

See The Artemis Accords (adopted Oct. 13, 2020), at https://www.nasa.gov/specials/

artemis-accords/img/Artemis-Accords-signed-13Oct2020.pdf.

Statement on a Two-Pillar Solution to Address the Tax Challenges Arising from the

Digitalisation of the Economy (July 1, 2021), https://www.oecd.org/tax/beps/statement-on-a-two-

pillar-solution-to-address-the-tax-challenges-arising-from-the-digitalisation-of-the-economy-july-

2021.htm.

See Agreement for Bringing Peace to Afghanistan

Between the Islamic Emirate of Afghanistan Which is Not Recognized by the United States as a State

and is Known as the Taliban and the United States of America (Feb. 29, 2020), at

https://www.state.gov/wp-content/uploads/2020/02/Agreement-For-Bringing-Peace-to-Afghanistan-

02.29.20.pdf.

Remarks by President Biden on the End of the War in Afghanistan, The White House

(Aug. 31, 2021), https://www.whitehouse.gov/briefing-room/speeches-remarks/2021/08/31/remarks-

by-president-biden-on-the-end-of-the-war-in-afghanistan/.

1-Feb-22] The Rise of Nonbinding International Agreements

Two high-level agreements concluded during the Obama administration—the

Iran nuclear deal and the emissions reduction pledge in the Paris Agreement on

climate change—generated controversy.

These agreements provoked controversy

in part because of the novel mechanisms the executive branch used to circumvent the

need for congressional consent. Many commentators argued that both agreements

required congressional approval because they were so consequential and because

they could not be fully justified by prior congressional authorization. Congressional

consent was a high hurdle to the deals, however, because there was significant

opposition in Congress.

The agreements posed additional challenges because both

made pledges that required domestic implementation. The United States in the Paris

Agreement agreed to undertake economy-wide emission reduction targets, and in the

Iran deal it agreed to eliminate certain sanctions against Iran.

The Obama administration took two innovative steps in concluding these

agreements. First, it made the Iran deal and the emissions pledge in the Paris

Agreement nonbinding. This allowed the administration to conclude the agreements

without seeking congressional approval. Second, it changed domestic law to meet

the commitments in these agreements by invoking pre-existing authority delegated

from Congress. For the Iran deal, the administration exercised the power that

Congress had given it to waive the sanctions in accordance with the national

interest.

And for the Paris Agreement, it made new regulations pursuant to

authority granted earlier in several domestic statutes.

The Paris Agreement itself was a binding agreement that was likely made pursuant to a

prior treaty—the United Nations Framework Convention on Climate Change (UNFCCC). See Curtis

A. Bradley & Jack L. Goldsmith, Presidential Control Over International Law, 131 HARV. L. REV.

1201, 1267-69 (2018). However, while the UNFCCC plausibly authorized most portions of the Paris

Agreement, it likely did not authorize the president to pledge binding emissions targets. Id. at 1268-

69. The Obama administration insisted that the emissions reduction targets in Article 4.4 of the Paris

Agreement be made nonbinding, which avoided the need for congressional or Senatorial approval. Id.

at 1251 & n. 232, 1268-69.

Majorities in both houses of Congress voted against approval of the Iran deal but were

unable to stop the agreement from taking effect under the terms of the Iran Nuclear Agreement

Review Act. See Jennifer Steinhauer, Democrats Hand Victory to Obama on Pact with Iran, N.Y.

TIMES, Sept. 11, 2015, at A1. For evidence of congressional opposition to the Paris Agreement, see

David M. Herszenhorn, Votes in Congress Move to Undercut Climate Pledge, N.Y. TIMES (Dec. 1,

2015).

See Kenneth Katzman, Cong. Research Serv., Iran Sanctions (2021). In addition, the

agreement was the basis for, and incorporated by reference into, a U.N. Security Council resolution

that terminated the international sanctions against Iran. See U.S.S.C. Resolution 2231 (20 July 2015).

For an overview of the domestic regulations that supported the nonbinding commitment in

the Paris Agreement, see Cass R. Sunstein, Changing Climate Change, 2009–2016, 42 HARV. ENVT’L

L. REV. 231 (2018).

The Rise of Nonbinding International Agreements [1-Feb-22

2. Joint Statements and Communiques. A second category of nonbinding

agreements are statements issued following high-level international meetings or

conferences that memorialize what the national representatives agreed to, their

intended subsequent courses of action on matters of mutual concern, or their

common positions growing out of the meeting. We will refer to these statements as

“joint statements and communiques.”

We define this category as follows: Often after a meeting or conference, the

representatives of at least two sovereign states issue a joint text (that text may be

issued jointly or separately, simultaneously or non-simultaneously) that does not

purport to create a legal obligation,

100

although it may (indeed, often does) suggest a

promise or intention to carry out future action. Such a text may also be issued by an

international organization that represents a group of sovereign states.

101

Like high-

level agreements, joint statements and communiques are generally very visible—

indeed, joint statements and communiques are almost always specifically intended

for public consumption. Unlike high-level agreements, however, joint statements

and communiques tend not to have many of the trappings of binding international

agreements, such as content organized by articles, entry into force and termination

provisions, or dispute resolution provisions.

102

They often read more like press

statements than international agreements.

A notable example of a joint statement or communique is the Atlantic

Charter, the 1941 “joint declaration” about postwar aims issued by Franklin

Such statements may also be referred to as “joint statement,” “joint declaration,” “joint

communique,” “ministerial statement,” “summit statement,” or “statement of intent.”

100

There are instances where states issue non-identical, but coordinated, press statements. In

2015, for example, President Obama and President Xi concluded a nonbinding agreement on

cybersecurity cooperation, announced by the White House in a “Fact Sheet.” Fact Sheet: President Xi

Jinping’s State Visit to the United States, WHITE HOUSE: OFF. SEC’Y (Sept. 25, 2015),

https://obamawhitehouse.archives.gov/the-press-office/2015/09/25/fact-sheet-president-xi-jinpings-

state-visit-united-states. China announced the same agreement in a read-out of President Xi’s visit.

We do not include such statements in this category, but the underlying agreement, still undisclosed, is

likely best categorized as a high-level agreement.

101

See, e.g., NATO, Summary of the NATO Artificial Intelligence Strategy (Oct. 22, 2021),

https://www.nato.int/cps/en/natohq/official_texts_187617.htm?utm_source=POLITICO.EU&utm_ca

mpaign=ff4d08d052-EMAIL_CAMPAIGN_2021_10_28_11_46&utm_medium=%20

email&utm_term=0_10959edeb5-ff4d08d052-190730104 (agreeing on principles relating to the

development and use of artificial intelligence in defense and security).

102

Some joint statements, however, do have the trappings of more formal agreements. See,

e.g., U.S.-EU Joint Declarations and Annexes (Nov. 3, 2009), at

https://obamawhitehouse.archives.gov/the-press-office/us-eu-joint-declaration-and-annexes

(containing three detailed annexes).

1-Feb-22] The Rise of Nonbinding International Agreements

Roosevelt and Winston Churchill following a series of meetings.

103

Another famous

example is the 1972 Shanghai Communique, in which the United States and China

pledged to work towards normalization of relations and to conduct relations on the

principles of respect for sovereignty, nonaggression, noninterference in internal

affairs, equality and mutual benefit, peaceful coexistence, and peaceful settlement of

disputes.

104

This paved the way for normalization of relations between the two

countries during the Carter administration, marked by the issuance of another joint

communique.

105

More recently, the United States, Israel, and the UAE used a joint

statement to announce the normalization of relations between Israel and the UAE in

2020.

106

And in 2021, the United States and China issued a “Joint Glasgow

Declaration on Enhancing Climate Action in the 2020s,” which sets out a plan for the

two countries to work together to address climate change.

107

Most joint statements and communiques are not particularly momentous. It

is standard practice for the White House, the State Department, and other agencies to

issue a joint statement announcing points of agreement and cooperation following a

meeting between the President or a high-level State Department official and a high-

level foreign official.

108

The United States is also a regular party to nonbinding joint

103

Atlantic Charter, U.S.-U.K., Aug. 14, 1941, 55 Stat. 1603. The same is true of the “New

Atlantic Charter” concluded by President Biden. See The New Atlantic Charter (June 10, 2021),

https://www.whitehouse.gov/briefing-room/statements-releases/2021/06/10/the-new-atlantic-charter/.

104

See Joint Communique, P.R.C.-U.S., Feb. 28, 1972, 66 DEP’T ST. BULL. 435, reprinted in

10 I.L.M. 443. The Communique also contained numerous unilateral pledges in addition to

cooperative ones. For example, and famously, the United States also stated that it was not challenging

the existence of one China and said that it “reaffirms its interest in a peaceful settlement of the Taiwan

question by the Chinese themselves.” Id.

105

See U.S.—China Joint Communiqué Establishing Diplomatic Relation, CURRENT

HISTORY VOL. 77, at 81, 81-86 (1979). Three years after that, during the Reagan administration, the

two countries issued another Shanghai Communique, in which the United States pledged to gradually

reduce its arms sales to Taiwan. See U.S. China Joint Communique on United States Arm Sales to

Taiwan (1982), available at https://www.ait.org.tw/our-relationship/policy-history/key-u-s-foreign-

policy-documents-region/u-s-prc-joint-communique-1982/.

106

Joint Statement of the United States, the State of Israel, and the United Arab Emirates

(Aug. 13, 2020), https://trumpwhitehouse.archives.gov/briefings-statements/joint-statement-united-

states-state-israel-united-arab-emirates/.

107

U.S.-China Joint Glasgow Declaration on Enhancing Climate Action in the 2020s (Nov.

10, 2021), https://www.state.gov/u-s-china-joint-glasgow-declaration-on-enhancing-climate-action-in-

the-2020s/. The Conference of the Parties to the Paris Agreement on Climate Change (nearly 200

countries) also agreed to a nonbinding Glasgow Climate Pact. See https://unfccc.int/sites/default/

files/resource/cma2021_L16_adv.pdf.

108

See, e.g., U.S.-India Joint Leaders’ Statement: A Partnership for Global Good, WHITE

HOUSE BRIEFING ROOM (Sept. 24, 2021), https://www.whitehouse.gov/briefing-room/statements-

releases/2021/09/24/u-s-india-joint-leaders-statement-a-partnership-for-global-good/

The Rise of Nonbinding International Agreements [1-Feb-22

statements or communiques following multilateral diplomatic conferences, such as

the Group of Seven, the Group of Twenty, and the North Atlantic Council.

109

There are likely many thousands of joint statements and communiques

currently in effect; we therefore did not attempt to develop a comprehensive

database. We did, however, gather over seven hundred documents that we obtained

primarily through public sources.

110

With the help of a team of research assistants,

we coded key features of the statements. We found that the content of these joint

statements varies widely, in terms of length, tone, specificity, and significance of the

commitments. Some merely state a shared understanding of a situation, shared

values, or general goals,

111

whereas others contain concrete and measurable pledges,

including pledges about how the parties will implement obligations under prior

binding agreements, or a framework to continue to ensure mutual compliance (for

example, an action plan on the ministerial level or a follow-up meeting to assess

progress).

112

In some of the statements, the U.S. executive branch pledges to seek

congressional action, such as appropriations. Strikingly, some of these statements

entail bold new commitments to future action,

113

and, as Table 1 shows, they often

use language recommended against by the State Department for nonbinding

agreements.

114

For instance, almost eighty percent use “will” or “agree,” and over

109

See, e.g., G7 Foreign and Development Ministers’ Meeting: Communiqué, EUR.

EXTERNAL ACTION SERV. (May 5, 2021), https://eeas.europa.eu/headquarters/headquarters-

homepage/97842/g7-foreign-and-development-ministers%E2%80%99-meeting-

communiqu%C3%A9_en (G7); G-8 Leaders Communique, WHITE HOUSE: OFF. SEC’Y (June 18,

2013), https://obamawhitehouse.archives.gov/the-press-office/2013/06/18/g-8-leaders-communique

(G7, formerly known as G8); Communique: Second G20 Finance Ministers and Central Bank

Governors Meeting, G20 (Apr. 7, 2021), https://www.g20.org/wp-

content/uploads/2021/04/Communique-Second-G20-Finance-Ministers-and-Central-Bank-Governors-

Meeting-7-April-2021.pdf (G20); London Declaration, NATO (Dec. 4, 2019),

https://www.nato.int/cps/en/natohq/official_texts_171584.htm (NATO).

110

We primarily gathered the documents from the Public Papers of the President of the

United States, https://www.govinfo.gov/app/collection/ppp/, and from the State Department online

archives. We obtained a small number through our FOIA requests to the agencies.

111

See, e.g., G8/Africa Joint Declaration: Shared Values, Shared Responsibilities (May 27,

2011), https://obamawhitehouse.archives.gov/sites/default/files/uploads/g8_africa_joint_declaration

_final_-eng.pdf.

112

See, e.g., U.S.-China Joint Statement Addressing the Climate Crisis (April 17, 2021),

https://www.state.gov/u-s-china-joint-statement-addressing-the-climate-crisis/.

113

These include the date on which the statement was concluded, participating countries,

title, subject area, short description, future plans or commitments, and whether it was signed or not.

This dataset will be made public on Dataverse upon publication of this Article.

114

See Guidance on Non-Binding Documents, supra note 67. See, e.g., Joint Statement by

President George W. Bush and Prime Minister Junichiro Koizumi: Partnership for Security and

Prosperity (June 30, 2001), https://www.govinfo.gov/content/pkg/PPP-2001-book1/pdf/PPP-2001-

1-Feb-22] The Rise of Nonbinding International Agreements

fifty percent use “agreement.” It is possible that the informality of the format—they

generally read more like press releases than international agreements—frees the

parties to make bold declarations of intent and use terms that might otherwise signal

a binding agreement.

115

Such statements appear to serve multiple purposes: For example, they are a

signaling device to the world of a plan of cooperative action between states. They

can also be used to signal a united front in the face of potential conflict, likely as a

form of deterrence.

116

And they can establish mechanisms for ongoing dialogue,

book1-doc-pg754.pdf (announcing a “new bilateral economic initiative” for trade engagement, noting

that the governments “will engage in cooperative efforts to address other key issues”).

115

Occasionally, documents labeled as joint statements are intended to be binding. For

example, the executive branch treated a 2012 joint statement with Afghanistan as a binding agreement

and reported it under the Case-Zablocki Act. See Joint Statement—Enduring Strategic Partnership

Agreement Between the United States of America and the Islamic Republic of Afghanistan (May 12,

2012), at https://www.govinfo.gov/content/pkg/PPP-2012-book1/xml/PPP-2012-book1-doc-

pg553.xml; White House, Fact Sheet—The U.S.-Afghanistan Strategic Partnership Agreement (May

1, 2012), at https://obamawhitehouse.archives.gov/the-press-office/2012/05/01/fact-sheet-us-

afghanistan-strategic-partnership-agreement.

116

See, e.g., Joint Statement on U.S.-Ukraine Strategic Partnership (Nov. 16, 2018),

https://2017-2021.state.gov/joint-statement-on-u-s-ukraine-strategic-partnership/index.html; U.S.-

Gulf Cooperation Council Camp David Joint Statement (May 14, 2015),

The Rise of Nonbinding International Agreements [1-Feb-22

including agency-level working groups. They are a commitment tool, as states not

legally bound to live up to the commitments in the statements may still suffer

reputational harm for turning back on a bold public declaration. And they serve the

purpose of providing internal signaling and agenda setting for domestic actors.

117

Once the leaders of their nations have committed to specific goals, that likely

informs and empowers those within their own governments to take steps to make

those commitments a reality.

3. Agency Nonbinding Agreements. The third category involves agreements

between government agencies.

118

The agreements contain a wide variety of

commitments, including commitments to exchange information, cooperate on

enforcement measures, consult with the other party prior to taking certain actions,

and align regulatory standards.

119

The growth of nonbinding agreements that foster

regulatory cooperation is in part a response to the increasing globalization of goods,

services, and persons. One sign of the increased importance of these agreements is a

2012 Executive Order entitled Promoting International Regulatory Cooperation,

which through various means encouraged agencies to engage in international

regulatory cooperation “consistent with domestic law and prerogatives.”

120

Unlike high-level nonbinding agreements and joint statements and

communiques, such agreements often are concluded out of public view.

121

While the

https://obamawhitehouse.archives.gov/the-press-office/2015/05/14/us-gulf-cooperation-council-camp-

david-joint-statement; U.S.-Japan Joint Statement on North Korea (Feb. 19, 2005), https://2001-

2009.state.gov/r/pa/prs/ps/2005/42491.htm.

117

See, e.g., Sixth Ministerial Meeting of the Global Health Security Initiative (Nov. 18,

2005), http://ghsi.ca/ministerial-statements/rome-november-2005/; Exploration of Mars: Statement of

Intent (Oct. 24, 2000), https://www.dropbox.com/work/Nonbindings/Data/Joint%20Statements/

NASA%20Joint%20Statements%20from%20FOIA?preview=Statement+of+Intent+-+France.pdf.

118

See Bradley & Goldsmith, supra note 95, at 1218-19; U.S. GOV’T ACCOUNTABILITY OFF.,

GAO-13-588: INTERNATIONAL REGULATORY COOPERATION 10 (2013),

https://www.gao.gov/assets/660/656488.pdf.

119

Another type of nonbinding instrument increasingly used by agencies is nonbinding

agency guidance. See Nicholas R. Parrillo, Federal Agency Guidance and the Power to Bind: An

Empirical Study of Agencies and Industries, 36 YALE J. REG. 165 (2019) (describing the use of

nonbinding guidance). Some of the transparency and rule of law issues presented by nonbinding

agency guidance may overlap with the issues presented by nonbinding international agreements,

although the latter are distinct in that they involve commitments to other nations rather than merely

domestic directives.

120

Exec. Order No. 13,609, 77 Fed. Reg. 26413 (May 1, 2012).

121

We emphasize that, as noted above, nonbinding agreements can assume countless forms

and do not always fall neatly into one of our three organizing categories. For example, the U.S.-E.U.

Privacy Shield, which governs global digital data flows, is a nonbinding agreement but is constituted

in an unusual form. The U.S. Commerce Department unilaterally issued a policy after negotiations

1-Feb-22] The Rise of Nonbinding International Agreements

text of high-level nonbinding agreements is not always available to the public or

Congress, the fact of the agreement is generally public. And joint statements and

communiques are designed for public consumption. Agency nonbinding agreements,

meanwhile, are a significant part of U.S. nonbindings practice, but they are generally

not public. Even when they are, they are scattered across the internet, making

comparisons and generalizations difficult. Agency nonbinding agreements appear to

play a significant and growing role in U.S. diplomacy and foreign policy, as well as

in U.S. regulatory policy, and they appear to sometimes serve as substitutes for

binding agreements. For these reasons, we devote Part II of this Article to an

extensive empirical study of them.

II. AGENCY NONBINDING AGREEMENTS: AN EMPIRICAL ANALYSIS

Agency nonbinding agreements are notable both for their growing

importance and for their lack of public transparency. This Part seeks to bring agency

nonbinding agreements out of the shadows. It begins by examining why agencies

conclude nonbinding agreements, based primarily on extensive interviews with

agency personnel who conclude them. It then outlines the features of nonbinding

agreements concluded at the agency level based on an analysis of a new—and, we

believe, the first—general database of U.S. nonbinding agreements.

A. Why Agencies Conclude Nonbinding Agreements

Nonbinding and binding commitments often can serve similar purposes.

Moreover, the officials that conclude nonbinding agreements are generally the same

ones that conclude binding executive agreements. To understand why they choose to

conclude a particular agreement as a nonbinding agreement rather than as a binding

one, we conducted a series of interviews with officials in several executive agencies.

While the reasons for making nonbinding agreements varied across agencies, agency

officials consistently referenced five key considerations. There may of course be

other important reasons that were not mentioned in the interviews—and

considerations offered by officials in one agency do not necessarily apply in the

same way in other agencies or in different contexts. The observations offered by

with the European Union and in contemplation of an agreement, see EU-U.S. Privacy Shield

Framework Principles, at https://www.privacyshield.gov/servlet/servlet.FileDownload?

file=015t00000004qAg, and the European Commission issued a decision that included the Commerce

principles and deemed them “adequate” under E.U. privacy law. See Commission Implementing

Decision of July 12, 2016 Pursuant to Directive 95/46/EC of the European Parliament and of the

Council on the Adequacy of the Protection Provided by the EU-US Privacy Shield, 2016 O.J. (L 207).

We also remind the reader that, as noted above, governments and their agencies engage daily in

informal exchanges that may include nonbinding agreements of various sorts that are excluded from

our analysis altogether.

The Rise of Nonbinding International Agreements [1-Feb-22

officials nonetheless provide rare insight into why nonbinding agreements have

become increasingly common.

1. The Nature of the Commitment. Agency choice of whether to conclude an

agreement as a binding or nonbinding agreement can turn on the kind of

commitments the agreement entails. In particular, three key considerations came up

repeatedly in interviews: whether the type of commitment is required by law to be

done as a binding agreement; whether the agreement includes terms that suggest an

intent to create a binding commitment; and whether the agreement has enforcement

aims that require a binding agreement.

First, agency officials indicated that there are certain types of agreements

where an agency may be obligated under domestic law to conclude a given

agreement as a binding agreement. For example, certain forms of international

defense cooperation, including “personnel status, classified technology, property

rights and the like,” require a binding agreement.

122

At the FAA, if an agreement

requires an exchange of money or of personnel, then agency legal guidance provides

that it must always be concluded as a binding agreement with reimbursement of

costs. As the Deputy Director of International Affairs at the FAA explained, “If we

want to go to Rwanda and do a 1-week workshop on civil aviation safety, that would

come under one of these [binding] agreements. It provides that we’ll do this

seminar, this is what it will cost, and you can invite everyone from the region.”

123

Such an agreement, he explained, is done as a binding agreement both because of the

exchange of money and because the government requires binding liability waivers

and other protections for FAA personnel traveling to the country. By contrast, an

agreement to engage in coordinated research and development, where there is no

exchange of money, can be done as a nonbinding agreement.

Second, nearly every agency official with whom we spoke made clear that

they take account of whether the agreement includes terms that suggest an intent to

create a binding commitment. For this, they look to the State Department guidance

regarding what terms should be avoided in nonbinding agreements to avoid

confusion about the intended nature of the commitment.

124

If an agreement contains

terms that convey intent to create a binding commitment, it has to be changed or

done as a binding agreement. Indeed, many agencies have binding and nonbinding

agreements that differ primarily in that the nonbinding versions avoid the terms that

122

McNeill, supra note 41, at 823 (“By law the United States is authorized to carry out

certain forms of international defense cooperation only pursuant to international agreement.”).

123

Interview by Oona Hathaway with David S. Burkholder, Deputy Director, International

Affairs, U.S. Federal Aviation administration (July 20, 2021).

124

See supra note 64.

1-Feb-22] The Rise of Nonbinding International Agreements

are indicative of binding commitments. A negotiator in the Antitrust Division of

DOJ, for example, noted that nonbinding agreements must say “‘intend to,’ rather

than ‘shall,’ ‘will,’ or ‘agree.’”

125

Third, sometimes the choice of binding versus nonbinding agreement turns

on enforcement aims. For example, early antitrust agreements with foreign partners,

such as with Germany in 1976 and Australia in 1982, sought to provide assistance

with U.S. enforcement of the United States’ distinctive antitrust rules. These

agreements were done as binding agreements, because this was seen as giving the

United States greater leverage to insist on cooperation in law enforcement.

126

Today,

however, many countries have antitrust laws that are similar to those in the United

States. The main purpose of the agreements today is to facilitate joint enforcement

cooperation in antitrust, consumer protection, and data protection. For this, the FTC

uses nonbinding agreements. The agency personnel “just want to be able to talk to

colleagues” and to “have a basis to cooperate with them on enforcement.”

127

For

that, nonbinding agreements work well. Agreements that allow parties to share

confidential, protected information in their files, however, are done as binding

agreements.

128

Indeed, agency officials consistently reported that if there were

penalties of any kind in an agreement, then the agreement was always done as a

binding agreement (though binding agreements need not necessarily include

penalties or other enforcement measures). Binding agreements that obligate the

United States to assist in law enforcement by other nation states—whether through

providing information or otherwise—are made only with states that agency officials

trust to provide adequate rule of law protections.

One interviewee noted that the State Department prefers binding agreements

where the United States is concerned about performance by the foreign partner.

When an agency is providing money to a foreign partner, for example, it generally

prefers to do so as part of a binding agreement. That is because the receiving

country is considered more likely to “pay attention to the agreement”—which

usually places specific conditions on the use of the money—if it is binding.

129

125

Interview by Oona Hathaway with Caldwell Harrop, Assistant Chief International

Section, Antitrust Division, Department of Justice (April 16, 2021) [hereinafter Harrop Interview].

126

Interview by Oona Hathaway with Randy Tritell, Director of the Office of International

Affairs, FTC (Jan. 21, 2021) [hereinafter Tritell Interview].

127

Id.

128

The International Antitrust Enforcement Act requires antitrust mutual assistance

agreements when conducting this sort of cooperation.

129

Interview by Oona Hathaway with Former U.S. Government Lawyer (June 11, 2019).

The Rise of Nonbinding International Agreements [1-Feb-22

Other agency officials indicated that the absence of legally enforceable

obligations is often not a drawback. The Director of the Office of International

Affairs at the FTC explained that while the agency’s nonbinding agreements have no

formal enforcement mechanism, neither do many of the agency’s binding executive

agreements. “Of course, hopefully, governments feel compelled to honor them

anyways.” But, he added, the FTC does not “go around pointing out to other

governments that they’ve failed to follow this provision or that provision. If there’s

a problem, we will talk about it.”

130

A negotiator in the Antitrust Division of DOJ

echoed this view: “Some say it’s not binding because if they don’t do what they say

they will do, we can’t do anything about it. But for all of these—even those that we

call [binding] agreements—we don’t have any enforcement.”

131

Indeed, the absence of enforceable obligations can be a feature rather than a

bug, especially if an agency is uncertain of its capacity to meet its commitments.

The Deputy Director of the Office of International and Tribal Affairs for the EPA

noted that most of the agreements the EPA makes with foreign partners are

nonbinding. She explained, “That is usually because don’t have dedicated funding

or a legislative mandate. . . . We prefer to do it as a nonbinding because if we can’t

proceed because our funding is cut, for example, we aren’t bound to carry it out.”

She further explained that the agency does not see much substantive benefit to

concluding binding agreements: “Our partners basically act in good faith. If there

was some sort of agreement where we needed assurances, then we may need it to be

binding. . . . Most of the time, we are planning to share information on best practices

or see how we can develop common standards, or engage in research. In those cases,

we don’t need to do a binding . . . . If, however, we did a joint research project where

we need to know how it was done or make sure that certain procedures are followed,

then we might do it as a binding.”

132

2. Foreign Counterpart Preferences and Requests. Agencies sometimes rely

on nonbinding agreements to accommodate foreign counterparts’ preferences or

constraints. An official at the Office of International Programs at the Nuclear

Regulatory Commission explained that her office generally prefers to conclude

information sharing arrangements as binding agreements, because that “provides

greater emphasis on the commitment.”

133

Yet the Commission often relies instead

on a nonbinding agreement, because the other country requests it.

130

Tritell Interview, supra note 126.

131

Harrop Interview, supra note 125.

132

Interview by Oona Hathaway with Inga Barnett-Owens, Deputy Director of Office of

International & Tribal Affairs, Environmental Protection Agency (April 9, 2021).

133

Interview by Oona Hathaway with Susan Wittick, Office of International Programs,

Nuclear Regulatory Commission (May 25, 2021) [hereinafter Wittick Interview].

1-Feb-22] The Rise of Nonbinding International Agreements

Many foreign experts and officials we surveyed expressed a growing

preference in their countries for nonbinding agreements. They cited efficiency,

speed, flexibility, and avoidance of legal and political constraints that apply to

binding agreements as key reasons. For many foreign partners, binding agreements

are more difficult to conclude because they cannot be made at the agency level. As

the NRC official explained, “There are a lot of partners that cannot negotiate binding

agreements agency-to-agency. A lot of our partners can only sign a nonbinding

arrangement at the agency level. That’s true of all the common law countries—for

example Canada, Australia, India.”

134

In such cases, concluding a binding agreement

“means elevating it and a lot more process, which can take years.”

135

To conclude a

binding agreement with Colombia, for example, “they have to go to the highest

authority in their nation to get approval to sign it. It effectively takes an act of

Congress. So with them we do it as a nonbinding.”

136

Another interviewee agreed:

“A lot of it is driven by what our partner wants.”

137

This is a growing phenomenon, according to several agency employees with

whom we spoke. As the NRC official explained, “[T]he preference for nonbinding

agreements] seems to be broadening around Europe.”

138

In short, as foreign partners

increasingly request nonbinding agreements to accommodate their own legal

frameworks (and perhaps to avoid their own legal and regulatory constraints), the

United States finds itself relying more heavily on such agreements. We heard this

point echoed in our discussions with foreign counterparts, as described in Part III

below, although we also heard that other countries sometimes use nonbinding

agreements because the United States has requested them.

3. The Potential for Trust-Building. In varying ways, trust plays a role in

decisions to rely on nonbinding agreements. As noted above, binding agreements

may be favored by a U.S. agency when it has concerns about whether the partner

country will live up to its commitments. For this reason, insisting on a binding

agreement can sometimes be perceived as reflecting a lack of trust. In the early

1990s, for example, the United States insisted on using binding agreements in

defense arrangements with UK, Australian, and Canadian defense counterparts.

134

Id.

135

Id.

136

Id.

137

Interview by Oona Hathaway with Stacey Nathanson, Attorney-Advisor, NOAA Office of

the General Counsel, Fisheries and Protected Resources Section (July 14, 2021).

138

Wittick Interview, supra note 133.

The Rise of Nonbinding International Agreements [1-Feb-22

Concerned about U.S. motives, the partners initially responded by suspending

negotiations.

139

Nonbinding agreements may also be used in some circumstances to build

trust, perhaps with the goal of ultimately concluding a binding agreement. This is

particularly an issue where the agreement requires specific performance by the

United States, not just by the foreign partner. At the Department of Justice’s

Antitrust Division, for example, binding agreements may include collaboration in

law enforcement. Agreements where the United States agrees to assist in law

enforcement tend to be made with countries with which there are longer-standing

connections, collaboration, and trust.

140

Nonbinding agreements, which allow for

but do not obligate the United States to assist in enforcement, are more likely to be

used with newer partners or those it is less clear share the same substantive antitrust

and rule of law commitments. As a DOJ Antitrust official put it, “Nonbindings can

be thought of as trust-building exercises.”

141

He added, “We tend to use MOUs with

China, India, Russia or other countries newer to the business of antitrust enforcement

or where we have a less developed relationship. Usually you develop a relationship,

trust with each other, then you might later want to memorialize that relationship with

a binding agreement.”

142

In recent years, the United States has used more

nonbinding agreements in the antitrust context, because it is working with more

partners than in the past. “Increasingly today we are using MOUs. That’s because

we are concluding more agreements with countries that we have less well-established

relationships with.”

143

Sometimes the binding agreements are more detailed than

nonbinding agreements, but that is not always the case. Indeed, the two may be

nearly identical.

4. Simpler Process. While none of our interviewees stated that they

concluded nonbinding agreements to evade the legal and regulatory requirements

that apply to binding agreements, they did note that concluding nonbinding

agreements is simpler. To conclude a binding executive agreement, an agency needs

to request and receive approval from the State Department to initiate negotiations. It

then must submit the concluded agreement to the State Department. At each stage,

lawyers at the State Department may offer input—and the process of review may

139

McNeill, supra note 41, at 822-23.

140

The United States currently has binding competition agreements with Australia, Brazil,

Canada, Germany, Israel, Japan, and Mexico. See Email from Michael Shore, Federal Trade

Commission, to Oona Hathaway (Jan. 29, 2021).

141

Harrop Interview, supra note 131.

142

Id.

143

Id.

1-Feb-22] The Rise of Nonbinding International Agreements

take time. The final agreement must then be reported to Congress. While it is rare

for Congress to raise concerns, it could do so. None of these regulatory requirements

apply to nonbinding agreements.

Even those agencies that voluntarily share nonbinding agreements with the

State Department find that the consultation process is simpler. The Associate

Director of the Office of International Affairs of the FTC explained that the review

process itself “is pretty simple. We send an email to [the Office of Treaty Affairs],

and they sent an email back saying it’s ok, or maybe saying change ‘shall’ to ‘intend

to,’ and we go ahead.”

144

5. A Tool for Regularized Interaction, Collaboration, and Leadership.

Nonbinding agreements are often focused on ongoing cooperative activity—for

example, regular information sharing and regulatory cooperation.

145

Not only do the

agreements themselves create mechanisms for collaboration, but their negotiation—

and renegotiation—also serves to strengthen collaborative ties. Indeed, nonbinding

agreements frequently expire after a set period—often five years. Such expiration

dates appear to be more common than they are for binding agreements. At least one

agency uses these expiration dates as a mechanism to ensure ongoing collaboration

between agency personnel and their counterparts abroad. An official at the Nuclear

Regulatory Commission explained that the Commission uses the renewal process as

an opportunity to connect with other nuclear-power states: “The arrangements

provide for information exchange and allows us to have personnel exchanges, as

well. . . . It helps us establish thick relationships with our counterparts. The five-

year renewal requirement establishes relationships and high-level engagements.”

146

The signing ceremony provides an opportunity for engagement at the leadership

level, but staff-level engagement begins almost a full year in advance. As one staff

member put it, “There are people I still exchange Christmas cards with because we

spent a year talking about these issues.”

147

These relationships are valuable and

ensure that if there were a crisis, the countries would have ties at all levels of the

agency that would allow a quicker and more effective response.

Nonbinding agreements are often concluded, moreover, as part of a package

that may ultimately include both binding and nonbinding agreements. For example,

a Department of Defense official described a practice of concluding a binding

144

Interview by Oona Hathaway with Russell Damtoft, Associate Director, Office of

International Affairs, FTC (Jan. 21, 2021).

145

See infra Subsection II.B.3.

146

Interview by Oona Hathaway with Brett Rini, Nuclear Regulatory Commission (May 25,

2021).

147

Wittick Interview, supra note 133.

The Rise of Nonbinding International Agreements [1-Feb-22

umbrella agreement that forms the foundation for subsequent nonbinding

agreements.

148

Such a “Chapeau Agreement” can satisfy legal requirements for

matters such as logistical support, liability, and property rights, allowing subsequent

agreements to be done as nonbinding agreements.

149

Sometimes, however, the

umbrella agreement is itself nonbinding. At the Federal Aviation Agency, for

example, the agency frequently concludes an umbrella nonbinding agreement,

usually called a “Memorandum of Cooperation.” This MOC provides a structure for

ongoing cooperation. When a joint opportunity for research emerges, it is then

specified in an “annex” underneath the MOC. Then there is usually an “appendix”

under the annex that provides even more detail on how the agreement will be carried

out. The Deputy Director of International Affairs at the FAA explained: “We do

annexes underneath those MOCs. So if we wanted to do collaboration on aircraft de-

icing with Canada, for example, we’ll have an annex establishing an R and D

program on de-icing, and then we’ll have an appendix under that annex that says

we’ll commit to spend $200,000 and you will commit to spend $200,000 and then

we are going to share the information.”

150

B. Analyzing Agency Nonbinding Agreements

It is difficult for observers to come to informed conclusions about agency

nonbinding agreements because of the general lack of transparency. We have been

critical about the lack of government transparency in its use of binding executive

agreements, but public information about binding agreements is much more

accessible than it is for nonbinding agreements.

151

Indeed, there is not only no

public repository of nonbinding agreements, there is no nonpublic repository either.

This Section draws on data analysis of a collection that we have developed of

over 1400 agency nonbinding agreements—which represents the only general

database of such agreements currently in existence.

152

We built the database as

148

Interview by Jack Goldsmith with U.S. Government Lawyer (Jan. 29, 2021). One such

agreement is the Agreement between the United States of America and the United Kingdom of Great

Britain and Northern Ireland (June 1, 2007), https://www.state.gov/wp-content/uploads/2019/02/07-

601-United-Kingdom-Defense-Amend.EnglishOCR.pdf.

149

See McNeill, supra note 41, at 825.

150

Interview by Oona Hathaway with David S. Burkholder, Deputy Director, International

Affairs, U.S. Federal Aviation Administration (July 20, 2021) [hereinafter Burkholder Interview].

The FAA also concludes MOUs, but it generally does so as one-off arrangements: “The MOU doesn’t

have annexes and appendixes underneath it.” Id.

151

See Hathaway, Bradley & Goldsmith, supra note 2.

152

Nine hundred and twelve of these agreements were collected from public online sources.

In addition, we submitted Freedom of Information Act requests to twenty-three agencies, departments,

1-Feb-22] The Rise of Nonbinding International Agreements

follows: We began by examining eighty-two government agencies or offices of

agencies that we thought might conclude nonbinding agreements. We found that

eleven agencies have substantial online collections of their nonbinding

agreements.

153

Not every agency is so transparent, however. For several agencies,

we found indications that they conclude nonbinding agreements, but few or none

were publicly available. For example, NASA’s Advisory Implementing Instruction

indicates that NASA signs at least eight different types of “non-agreements,” but

none are published online.

154

We filed FOIA requests with twenty-three separate

agencies or departments between December 2020 and April 2021. By January 2022,

fourteen had produced documents, and nine had not yet done so.

155

Of these, only

or divisions thereof. We then checked all the nonbinding arrangements against the database of

binding agreements that we had compiled for Hathaway, Bradley & Goldsmith, supra note 2.

153

Those with significant online collections are Alcohol and Tobacco Tax and Trade Bureau

(U.S. Treasury), https://www.ttb.gov/publications/memorandum-of-understanding; Commodity

Futures Trading Commission,

https://www.cftc.gov/International/MemorandaofUnderstanding/index.htm; Department of Energy,

https://www.energy.gov/ia/international-energy-commitments-iec; Department of the Interior,

https://www.doi.gov/intl/International-Instruments; Department of Justice, Antitrust Division,

https://www.justice.gov/atr/antitrust-cooperation-agreements; Department of Transportation, Federal

Aviation Administration, https://www.faa.gov/space/additional_information/international_affairs/;

Equal Employment Opportunity Commission, https://www.eeoc.gov/mou/memoranda-understanding;

Federal Communications Commission, https://www.fcc.gov/general/international-agreements; Food

and Drug Administration, https://www.fda.gov/international-programs/international-arrangements;

Federal Trade Commission, https://www.ftc.gov/policy/international/international-cooperation-

agreements; and the Securities and Exchange Commission,

https://www.sec.gov/about/offices/oia/oia_cooparrangements.shtml.

154

NAT’L AERONAUTICS AND SPACE ADMIN., SPACE ACT AGREEMENTS GUIDE (2014)

https://perma.cc/FT8L-CQ3X at 20. NASA, Active International Agreements by Signature Date (as of

September 30, 2020), https://www.nasa.gov/sites/default/files/atoms/files/house_

approps_agreement_report_09-30-2020_international.pdf. Similarly, the Office of the Director of

National Intelligence Guideline 4 Report notes that, “In addition to legally binding international

agreements there are also Memoranda of Understanding between governments that are not legally

binding but are relatively detailed arrangements that support information sharing.” Guideline 4 Report

- Facilitate Information Sharing Between Executive Departments and Agencies and Foreign Partners,

at 2, available at https://www.dni.gov/files/ISE/documents/DocumentLibrary/guideline-4---sharing-

with-foreign-partners.pdf.

155

The agencies that produced documents or were otherwise responsive to our FOIA request

were the Department of the Treasury; Federal Trade Commission; Department of Justice (Antitrust

Division, Civil Rights Division & Tax Division); Department of Transportation (FAA); Department

of Labor; Department of Commerce (National Oceanic and Atmospheric Administration); US

Agriculture Department; Environmental Protection Agency; the Export-Import Bank; the Food and

Drug Administration; NASA; and the Nuclear Regulatory Commission. The unresponsive agencies

(as of January 9, 2021) are the Department of Commerce; Department of Defense, Department of

Homeland Security; Department of Justice (Criminal Division); Office of Management and Budget

(OIRA); State Department (Office of Information Programs and Services & Legal Adviser’s Office);

and USAID. All but the Department of State, Department of Defense, and Department of Homeland

The Rise of Nonbinding International Agreements [1-Feb-22

one—DHS—claimed that records were exempt from FOIA (though as of this writing

it has not identified the applicable exemption). We sued three of the least responsive

agencies—the U.S. Department of State, the Department of Defense, and the

Department of Homeland Security. That litigation is ongoing, and we will add to the

database any responsive documents we receive as a result (as well as those received

from the other outstanding requests).

With a team of research assistants, we coded all the nonbinding agreements

in our database to identify a range of characteristics. Unless otherwise noted, the

data below is based on this coding.

156

Although our database is the first of its kind, it

is important to emphasize that it is not comprehensive. Agencies vary a great deal in

whether they post the agreements online and in their responsiveness to our FOIA

requests. While we requested agreements back to 1989 in order to allow

comparisons to the database of binding agreements that we built for earlier work,

157

agency records are less accessible and comprehensive the further back in time we go.

Indeed, older agreements are less likely to be digitized, which may affect their

availability.

Nevertheless, as we will explain, there are reasons to believe that the picture,

which encompasses a wide variety of agencies, is generally indicative of the patterns

and trends in this area. Moreover, to the extent that our empirical account suggests

some reasons for concern (such as about lack of coordination and transparency), the

agreements not within our dataset would almost certainly, if known, heighten the

concerns. After all, the agencies that either already publicly disclose their

agreements or that were cooperative in the FOIA process are likely the agencies with

the least to hide. If we had access to the full range of nonbinding agreements that

agencies conclude, the case for reform would likely be stronger, not weaker. It is our

hope that the publication of this Article will help prompt greater transparency for

nonbinding agreements, either voluntarily by the executive branch or through

legislative mandate, which would in turn allow researchers to develop a more

complete account of this important government practice. But the account below is

much more extensive than anything previously published.

Security provided responses to queries and indicated that the requests were in progress, but they

nonetheless did not produce any documents as of January 2022. The requests sought all unclassified

nonbinding agreements and any internal agency guidance relating to such agreements from January 1,

1989 to December 31, 2019. To assist agencies in identifying non-binding agreements, the request

referenced the State Department Guidance, supra note 64, and specified that any agreement reported

under the Case-Zablocki Act was not included.

156

The dataset and codebook will be made available on Dataverse when this Article is

published.

157

See Hathaway, Bradley & Goldsmith, supra note 2.

1-Feb-22] The Rise of Nonbinding International Agreements

1. The Agencies that Conclude Nonbinding Agreements. As noted above, we

determined that forty-one agencies had concluded nonbinding agreements. Figure 1

represents the agencies in our database. Some rely heavily on these agreements,

especially the Department of Energy (308 agreements); Health and Human Services

(198); Export-Import Bank (148); Department of Transportation (125); Department

of Commerce (123); SEC (106); and Commodities Futures Trading Commission

(103). There are, moreover, a number of agency collaborations (for example, the

DOJ and FTC frequently collaborate on antitrust agreements). Absence from this

chart, however, does not mean that an agency does not use nonbinding agreements.

We filed a lawsuit against the Department of Defense for its failure to respond to our

FOIA request. In response, the Department identified over 6000 documents it claims

are responsive to our request—documents it has not yet disclosed. The Department

of Homeland Security, which we have also sued, has not yet identified how many

documents are responsive to our request.

FIGURE 1: AGENCY NONBINDING AGREEMENTS BY AGENCY

2. Foreign Partners. Agency nonbinding arrangements are widely used, but

they are particularly concentrated with several key states. Table 2 shows the top

partners for binding agreements, agency nonbinding agreements, and joint statements

and communiques.

158

There is significant overlap across the three types of

agreements, though the ordering is not identical. China, for example, is the second

158

Only bilateral agreements are included in a country’s total. Agreements with more than

one country are included as Multilateral.

The Rise of Nonbinding International Agreements [1-Feb-22

most significant partner for agency nonbinding arrangements, but the 20

most

frequent partner for binding executive agreements. Moreover, several

commonwealth countries—Canada, United Kingdom, India, and Australia—are

higher on the ranking for nonbinding agency agreements than they are for binding

agreements. This supports the observation of several interviewees that the

commonwealth countries are increasingly requesting nonbinding arrangements.

Although our separate dataset of Joint Statements and Communiques is smaller, and

thus we are cautious about drawing broad conclusions from it, they appear to mirror

more closely the patterns found in binding agreements—particularly in the

dominance of multilateral over bilateral statements.

3. Substantive Commitments Across Nonbinding Agreements. We identified

nine types of trans-substantive commitments that appear in nonbinding

1-Feb-22] The Rise of Nonbinding International Agreements

agreements.

159

The results appear in Table 3. The most common substantive

commitments—(1) regulatory cooperation and coordination and (2) information

exchange—are often intertwined. For many agencies, nonbinding agreements serve

as a vehicle for working with foreign partners to gather information required to carry

out their regulatory missions. Many of these agreements, moreover, include

confidentiality requirements. These nonbinding agreements allow for information to

be shared between agencies to help them perform their regulatory tasks, and the

agreement provides assurances that shared information will not be divulged. (Most

agreements included more than one type of commitment, hence the total sums to

well over 100%.)

Information-sharing agreements may be particularly well suited to a

nonbinding commitment. After all, if a state fails to live up to its side of the

commitment—by, say, failing to provide information as promised or by divulging a

piece of information that is supposed to be kept confidential—the counterpart can

159

For each type of commitment, we identified common terms of reference. We then

searched the text of all of the agreements and identified the number of unique agreements with at least

one of the relevant terms. For information exchange, for example, the terms were: “information

exchange; information exchanges; provide information; providing access to information; exchange of

technical information; providing the information; transmit the information; provide technical

information; information sharing; information-sharing; requests for information; sharing information;

sharing relevant nonproprietary information; sharing of information; transfers of personal data; share

knowledge; knowledge exchange; provision of information; information shall be provided;

information is shared in confidence; collect and share information; exchanging information; exchange

technical, commercial and financial information; exchanging knowledge; exchange of ideas and

information; exchange of publicly available scientific and technical information; exchange of data and

information; exchange information; exchange of information; exchanges of information; exchange of

scientific and technical information; exchange information.”

The Rise of Nonbinding International Agreements [1-Feb-22

simply cease its own performance in response. Though we found no agreement that

specified this informal enforcement measure, it is implicit in the nature of the

agreement. Neither side is bound to comply; hence if one side ceases to comply the

other state needs no justification to reciprocate.

This built-in informal enforcement measure is implicit in nearly all the

substantive commitments that cross subject areas. Consider, for example,

nonbinding agreements relating to research and technical cooperation. These

agreements often involve parallel research and development programs in which two

countries agree to invest roughly similar amounts of funding into a particular

research topic (for example, the best kind of asphalt for airport runways)

160

and then

promise to share the information that the research program produces. If one side

fails to live up to its commitment—by failing to invest or failing to share the

resulting information—the other side can respond by doing the same or by refusing

to cooperate going forward. In short, the nonbinding agreement may be particularly

well suited, and perhaps even preferred, in cases where formal enforcement tools are

unnecessary because each side benefits from ongoing joint performance. Simple

reciprocity serves to provide sufficient incentive for states to perform as promised.

The performance covered by nonbinding agreements, moreover, tends to be

ongoing cooperative activity—for example, regular information sharing. Hence, one

side does not risk making a large investment that is lost if the other side fails to

perform. (While characteristic of nonbinding agreements, this structure is not

exclusive to them.) If one state ceases providing information, the other side suffers a

minor loss. It is likely that where there is staged performance—one side gives a

large sum of money and then the other side performs an agreed task, for example, it

may be preferable for the agreement to be binding—because the state that performs

first loses leverage and requires some external tool to ensure performance by the

second party. Finally, failure to comply with these substantive commitments is not

subject to runaway effects that legally binding treaties often seek to restrict—such as,

for example, restrictive trade policies that can quickly spiral out of control in a

system that permits tit-for-tat responses.

161

4. Nonbinding Agreements Over Time. Based on the agreements in our

database, it appears that the number of nonbinding agreements has grown over time.

Given the partial nature of the data, we are cautious about drawing conclusions based

160

Burkholder Interview, supra note 150 (offering a joint research program into the best

asphalt to use on airport runways as an example of a topic on which the FAA might conclude a

nonbinding agreement with another country).

161

See Oona A. Hathaway & Scott J. Shapiro, Outcasting: Enforcement in Domestic and

International Law, 121 YALE L.J. 252, 327 (2011) (describing reasons for “adjudicated” outcasting).

1-Feb-22] The Rise of Nonbinding International Agreements

on these results alone. Nonetheless, there is good reason to think that this reflects a

real trend. Nearly all interviewees indicated that their agencies were relying more

heavily on nonbinding agreements. All but one of the surveys of comparative

scholars and practitioners (described below in Part III) similarly affirmed that they

had witnessed an increase in reliance on nonbinding agreements.

Figure 3 shows a gradual increase in nonbinding agreements in the database

over time, accelerating in the late 2000s. The spike in 2013 is due to jumps in

nonbinding agreements concluded by three agencies: the Department of Energy

(which concluded 49 nonbinding agreements in 2013, compared to 20 in 2012 and

22 in 2014); the CFTC (which concluded 30 nonbinding agreements in 2013,

compared to 0 in 2012 and 4 in 2014); and the SEC (which concluded 27 nonbinding

agreements in 2013, compared to 5 in 2012 and 1 in 2014). There does not appear to

be any precipitating cause for this bump, although it is possible that it was prompted

by the 2012 Executive Order that encouraged agencies to engage in international

regulatory cooperation “consistent with domestic law and prerogatives.”

162

The dip

at the end of the graph may represent a lag in posting agreements online, although it

is also possible that there has been a recent drop-off in the conclusion of agency

agreements, including during the Trump administration. The key point for our

purposes is that there has been an overall increase over time in the conclusion of

agency nonbinding agreements.

FIGURE 3: AGENCY NONBINDING AGREEMENTS BY DATE CONCLUDED

162

See Executive Order 13,609, supra note 120.

The Rise of Nonbinding International Agreements [1-Feb-22

Interestingly, nonbinding agreements concluded by agencies appear to have

become more common even as binding executive agreements have become less

common. Figure 4 compares binding agreements over time to nonbinding

agreements from 1989-2016.

163

(Notably, we are confident that our set of binding

agreements is fairly complete, as it represents all agreements reported to Congress by

the State Department; the set of nonbinding agreements, however, is not complete,

for reasons explained earlier.) Binding executive agreements peaked in 2006 and

began a gradual decline, whereas nonbinding agreements have continued to show

overall growth.

164

It is difficult to determine whether the rise of nonbinding

agreements is offsetting what would have otherwise been binding executive

agreements, but a general rise of nonbinding agreements during a decline in binding

executive agreements suggests that there may be some degree of substitution.

FIGURE 4: AGREEMENTS, BY DATE CONCLUDED

The fall of binding agreements and concomitant rise of nonbinding

agreements is even more apparent if we eliminate agreements in just two subject

areas (either primary or secondary) where we have had less success obtaining

nonbinding agreements from agencies—defense (as noted, we have sued the

Department of Defense for nonresponsiveness) and humanitarian (USAID has yet to

produce documents). The results, in Figure 5, demonstrate that in recent years the

163

For data on binding arrangements, we rely on the database we compiled for Hathaway,

Bradley & Goldsmith, supra note 2. Links to the data can be found at Hathaway, Bradley &

Goldsmith, supra note 2, https://harvardlawreview.org/executive-agreements-visualizations. Figure 3

ends in 2016, because our data on binding executive agreements end in 2016.

164

Peake likewise finds that binding agreements plateaued in 2005-06. See Peake, supra note

1-Feb-22] The Rise of Nonbinding International Agreements

number of agency nonbinding agreements in the database is close to, and sometimes

even exceeds, the number of binding agreements, even though the database of

agency nonbinding agreements is less complete than the database of binding ones.

While the number of joint statements in our database remains comparatively small,

they, too, have grown and, together, the two forms of nonbinding agreements have

begun to overshadow binding ones.

FIGURE 5: AGREEMENTS, BY DATE CONCLUDED, EXCLUDING DEFENSE AND

HUMANITARIAN

5. Subject Areas. The agency nonbinding agreements in our database are

used in a variety of subject areas, but, as Figure 6 shows, they are particularly

concentrated in Finance, Trade, and Investment (485, or 34% of the total);

Environment, Conservation, and Energy (440, or 31%); and Science, Space, and

Technology (171, or 12%).

165

It is difficult to know to what degree this reflects real

differences in practice and how much it instead reflects differences in transparency.

As noted below, it is likely a product of both. Joint statements largely conform to

this same ordering, though there are far more joint statements than agency

nonbinding agreements in the areas of Defense and Humanitarian—the very two

areas in which the agencies have so far refused to produce nonbinding agreements.

165

All of these figures are based on the primary subject area identified.

The Rise of Nonbinding International Agreements [1-Feb-22

FIGURE 6: AGENCY NONBINDING AGREEMENTS AND JOINT STATEMENTS BY PRIMARY

SUBJECT AREA

Table 4 shows the distribution of binding executive agreements by subject

area relative to disclosed agency nonbinding agreements and joint statements. The

greatest disparity between binding agreements and agency nonbinding agreements is

in Defense. (Notably, joint statements reflect an ordering that matches binding

agreements perfectly, with defense the leading subject.) It is almost certain that this

reflects a difference in what is public rather than a difference in propensity to

conclude binding versus nonbinding agreements. A U.S. government employee told

us that the Department of Defense does “many hundreds and hundreds” of

nonbinding agreements, and, indeed, the Department has identified 6000 responsive

documents in our FOIA litigation. The use of nonbinding agreements is a large and

growing model for the DoD, despite its current absence from our agency nonbinding

agreement database.

Some agencies, by contrast, are transparent precisely because they want the

coordination between agencies to be publicly known. In the field of antitrust, for

example, making public the nonbinding agreements signals a level of cooperation

that regulatory authorities consider potentially helpful in encouraging companies to

1-Feb-22] The Rise of Nonbinding International Agreements

adhere to regulatory requirements.

166

The larger number of nonbinding agreements

in our database on Finance, Trade, and Investment may therefore reflect not just

reliance on nonbinding agreements but willingness to disclose them.

6. Clarity About the Nonbinding Commitment. One concern about

nonbinding agreements is whether they clearly reflect the intent of the parties. Many

nonbinding agreements either specifically state that they are nonbinding or have

language that makes clear the intent not to create a binding agreement (e.g., “This

Memorandum of Understanding does not impose any legally binding obligation on

the Authorities or supersede domestic law”). Figure 7 shows that a growing portion

of agency nonbinding agreements are expressly nonbinding in this way.

166

Interview by Oona Hathaway with Russell Damtoft, Elizabeth Kraus, Stacy Feuer,

Michael Shore, and Randy Tritell, FTC (Jan. 14, 2021).

The Rise of Nonbinding International Agreements [1-Feb-22

FIGURE 7: IS IT EXPRESSLY NONBINDING?

Nonetheless, as can be seen in Table 5 below, most of the nonbinding

agreements also use terms commonly associated with binding agreements—terms the

Department of State’s Guidance specifically cautions against:

167

treaty, agreement,

parties, shall, agree, undertake, entry into force, is to come into operation, activities

are to commence, done at, concluded at, and will.

168

These recommendations are

frequently ignored. This suggests that the United States is not observing guidance

meant to avoid misunderstandings with foreign partners. It also indicates that

coordination within the U.S. government is imperfect. The Guidance states, “The

Office of Treaty Affairs encourages agencies and offices to share the texts of

proposed non-binding documents with the office, which is responsible by law for

determining whether a particular document is a binding ‘international agreement’ for

purposes of reporting to Congress.” Some of the agencies with which we spoke

indicated that they did so, but the failure of so many agreements to comply with the

Department’s Guidance suggests that this may not be generally true.

167

This table was constructed by using optical character recognition (Amazon Textract) to

extract the text of all the agreements. This enabled us to search the text of the agreements for the

terms in the table. Whether the agreement is “expressly nonbinding” or not is based on a coding of

the agreement by research assistants after reading the entire agreement.

168

State Department Guidance, supra note 64.

1-Feb-22] The Rise of Nonbinding International Agreements

Many nonbinding agreements also have other features normally associated

with binding agreements. As seen in Table 6, a significant percentage of nonbinding

arrangements reference implementation, provide for some form of dispute resolution,

designate a process for amending or revising the agreement, or include a termination

or withdrawal provision. While none of these features makes an agreement binding,

each has the potential to create some confusion about the nature of the agreement.

Interestingly, with just one exception (dispute resolution), these features are more

common in agreements that are expressly nonbinding. Perhaps agencies consider

express disclaimers to be sufficient to meet the State Department’s concerns.

The Rise of Nonbinding International Agreements [1-Feb-22

III. A COMPARATIVE PERSPECTIVE

Nonbinding agreements have become more important not just in U.S.

practice, but around the globe. The practice of other nations is relevant to the

analysis of U.S. practice for several reasons. These nations are potential partners

with the United States in concluding both binding and nonbinding agreements, and

how the United States approaches nonbinding agreements will affect its relations

with these nations, and vice-versa. Moreover, other nations may pursue reform

strategies concerning nonbinding agreements that are relevant to the United States as

it considers how best to address this growing phenomenon. This Part therefore

describes comparative practice in this area, with a focus on the practice of prominent

constitutional democracies. In addition to taking account of publicly available

materials, we have solicited information from government officials and scholars in

over a dozen countries, both through detailed written surveys of the practice in their

countries and through their participation in an online conference.

169

A. The Global Rise of Nonbinding Agreements

The rise in nonbinding agreements is not limited to the United States. Based

on accounts from scholars and practitioners from around the world, it appears to be a

169

See University of Chicago Law School, Conference on “Non-Binding International

Agreements: A Comparative Assessment,” https://www.law.uchicago.edu/events/non-binding-

international-agreements-comparative-assessment. The surveys addressed the laws and practices of

Argentina, Austria, Canada, the European Union, Finland, France, Germany, Mexico, Israel, the

Netherlands, Spain, the United Kingdom, South Africa, and Switzerland, and they are on file with the

authors. We also draw upon a survey conducted in 2019 by Canada’s treaty department in which

eight nations (including Canada) were asked to describe their laws and practices relating to both

binding and nonbinding agreements. In addition, the OAS Report, supra note 30, includes information

on the practices of thirteen OAS members.

1-Feb-22] The Rise of Nonbinding International Agreements

widespread—indeed global—phenomenon.

170

For example, the Legal Adviser to

Germany’s Federal Foreign Office has noted that “[a]s seemingly everywhere else,

the significance of non-legally binding agreements has consistently been rising in our

practice” and that “[i]ssues that would have formerly been the subject of a binding

treaty under international law are nowadays addressed through Joint Declarations of

Intent.”

171

A survey respondent from Canada reported that “[t]here has been

significant growth in the use of [nonbinding] arrangements” and that “Canada now

concludes hundreds of arrangements per year.”

172

Mexico has reported that 70% of

the agreements now submitted to its foreign ministry for review are nonbinding.

173

A prominent British authority on treaties who formerly worked in the British Foreign

and Commonwealth Office has observed that “the use of MoUs is now so

widespread, some officials may see the MoU as the norm, with a treaty being used

only when it cannot be avoided.”

174

When asked to explain why there has been an

increase in these agreements, our survey respondents attributed it to factors such as

increased international cooperation by regulators, the ease and speed by which such

agreements can be concluded, the greater flexibility offered by nonbinding

agreements, and the desire for confidentiality.

Perhaps not surprisingly, governments and international organizations are

increasingly turning their attention to this phenomenon. In 2016, the Inter-American

Juridical Committee of the Organization of American States launched an initiative to

identify state practices in the Americas regarding both binding and nonbinding

170

Countries use a variety of terms to describe what we are calling nonbinding international

agreements. In some countries, there is an effort to avoid using the word “agreement” in this context

because that word might suggest a binding commitment.

171

Eick, supra note 12.

172

Gib van Ert, Law—Ottawa and Vancouver, Survey for University of Chicago Law School

Conference on “Non-Binding International Agreements: A Comparative Assessment” (submitted

Aug. 11, 2021). Most of our survey respondents reported seeing an increase. Some indicated that

they thought there had been an increase but that it was difficult to know for sure given the lack of

publication of the agreements. The respondent from the Netherlands reported that she had not seen an

increase in nonbinding agreements. See Noortje van Rijssen, Legal Office, Netherlands Ministry of

Foreign Affairs, Survey for University of Chicago Law School Conference on “Non-Binding

International Agreements: A Comparative Assessment” (submitted July 28, 2021). In an earlier

survey conducted by Canada of the practices of eight countries, all respondents reported an increase in

both the frequency and importance of nonbinding agreements. See Treaty Law Division, Global

Affairs Canada, Working Group on Treaty Practice, Survey on Binding and Non-Binding

International Instruments 13, 31 (Sept. 18, 2019) (on file with authors) [hereinafter Working Group

Survey].

173

See Alejandro Rodiles, ITAM School of Law, Mexico, Survey for University of Chicago

Law School Conference on “Non-Binding International Agreements: A Comparative Assessment”

(submitted Sept. 1, 2021); see also Working Group Survey, supra note 172, at 31.

174

AUST, supra note 29, at 29.

The Rise of Nonbinding International Agreements [1-Feb-22

agreements. As the rapporteur (Duncan Hollis) explained, the initiative “found its

impetus in the rising number of non-traditional international agreements, including

non-binding agreements among States as well as agreements in both binding and

non-binding form concluded by government ministries and sub-national territorial

units.”

175

In 2020, the Committee published a Final Report on Binding and Non-

Binding Agreements, as well as a set of Guidelines and Commentary relating to

these agreements.

176

The OAS Guidelines suggest that nations adopt a number of

legal reforms, some of which we describe below. In addition, Canada’s treaty

department circulated a survey in 2019 in which eight nations (including Canada)

were asked to describe their laws and practices relating to both binding and

nonbinding agreements.

177

More recently, the Concept Note for a March 2021

meeting of the Council of Europe’s Committee of Legal Advisers on Public

International Law (CAHDI) stated that nonbinding agreements “are of increasing

prevalence in international relations” and cautioned that, although these instruments

“present a number of advantages for states as compared to treaties,” “the usage of

non-binding agreements is not without dangers.”

178

In some countries, recent events triggered increased attention to nonbinding

international agreements. There was significant controversy in Switzerland, for

example, over the government’s support for the Global Compact for Migration, a

nonbinding multilateral instrument endorsed by the UN General Assembly that sets

forth a variety of “guiding principles” and “objectives and commitments” relating to

migration.

179

The Switzerland constitution expressly gives the parliament a right to

participate in decisionmaking about foreign relations, and it requires parliamentary

approval of most treaties, and sometimes makes treaties subject to popular referenda.

Critics argued that unilateral executive branch approval of the Migration Compact

would be inconsistent with these requirements, despite the Compact’s nonbinding

status. In response to the controversy, the Swiss government decided to submit the

175

OAS Report, supra note 30.

176

See id.

177

See Working Group Survey, supra note 156.

178

Concept Note, Expert Workshop on Non-Legally Binding Agreements in International

Law (Mar. 26, 2021), https://rm.coe.int/concept-note-workshop-non-legally-binding-agreements-in-

international-/1680a2358d.

179

See Domnhall O’Sullivan, UN Migration Pact: A Hard Landing for “Soft” Law,

https://www.swissinfo.ch/blueprint/servlet/eng/politics/multilateralism_un-migration-pact--a-hard-

landing-for--soft-law-/44635712. Earlier debates about Parliament’s role in nonbinding agreements

had arisen in the contexts of banking, finance, and tax. See Anna Petrig, Democratic Participation in

International Lawmaking in Switzerland After the “Age of Treaties,” in ENCOUNTERS BETWEEN

INTERNATIONAL LAW AND FOREIGN RELATIONS LAW: BRIDGES AND BOUNDARIES (Helmut Aust &

Thomas Kleinlein eds., 2021).

1-Feb-22] The Rise of Nonbinding International Agreements

Migration Pact to the parliament for approval, while emphasizing that it was not

required to do so under the Constitution.

In considering the phenomenon of nonbinding agreements, nations and the

EU have been grappling with three basic issues: (1) How to ensure that there is

sufficient coordination within the executive with respect to the making of nonbinding

agreements, through mechanisms such as foreign ministry review and centralized

collection. (2) Whether and to what extent the transparency rules that apply to

binding agreements should also apply to nonbinding agreements—in particular,

whether these agreements should be made available to the public. (3) The extent to

which the legislature should be involved in or notified of these agreements. The

following sections describe national laws and practices in prominent constitutional

democracies relating to these three issues.

B. Coordination

Many nonbinding international agreements are made not by heads of state or

foreign ministries but rather by other departments and agencies of the executive

branch. Governments have found that this disaggregation of the practice presents

challenges with respect to the management of national foreign policy. The foreign

ministry might not know what commitments are being made on behalf of the

country, and it might not approve of them if it did know. In addition, some of those

commitments might unintentionally create binding obligations if not drafted

carefully. Moreover, without coordination, a commitment made by one department

or agency might conflict with a commitment made by another department or

agency.

180

Many foreign ministries have provided general guidance to executive

ministries and agencies about the drafting of nonbinding agreements—for example,

about terms that should be avoided to help ensure that the agreement will not be

considered binding. Some nations have gone further and have instituted centralized

foreign ministry review and approval of nonbinding agreements. In the UK, for

example, the Foreign Commonwealth and Development Office (FCDO) has issued a

guidance document emphasizing that, “[a]s with treaties, all draft MoUs should be

sent to the relevant FCDO department for clearance by their legal adviser, and

180

See generally OAS Report, supra note 30, at 114 (“When it comes to non-binding

agreements, States currently suffer from an information deficit. Both the number and contents of a

State’s political commitments, whether labeled as MOUs or otherwise, are often unclear.”).

The Rise of Nonbinding International Agreements [1-Feb-22

foreign language versions should be checked. Moreover, there should be the same

level of inter-departmental consultation as for treaties.”

181

Similarly, in Canada all departments and agencies are supposed to notify the

Treaty Section of the Department of Foreign Affairs, Trade and Development before

beginning agreement negotiations with another nation or an international

organization, in part so that “a proper distinction between treaties and other

international instruments that are not binding in public international law can be

maintained.”

182

All departments and agencies are directed “to avoid situations where

instruments that could reasonably be viewed as treaties . . . are not mistakenly

classified as non-binding instruments.”

183

In addition, nonbinding agreements

concluded by departments or agencies in Canada require centralized government

approval—usually through the Ministry of Foreign Affairs, although approval must

come from the Cabinet if the nonbinding agreement “would result in a major shift in

Canadian policy.”

184

In Canada’s Treaty Law Division, two lawyers are responsible

for reviewing nonbinding agreements and one of them serves as an “MOU

Coordinator.”

185

The Australian government has a less formal process, but it also encourages

centralized coordination. Its Guidance Note states that any agency that intends to

enter into a nonbinding agreement should consult with the Treaties Section of the

Department of Foreign Affairs and Trade when drafting and negotiating the text, and

it has set forth guidelines about the appropriate language to be used and avoided,

along with a model MoU.

186

The Guidance Note also states that all nonbinding

agreements should be sent to the Treaties Section for clearance prior to signature. It

directs agencies to retain the texts of these agreements, but it does not itself maintain

181

Treaties and Memoranda of Understanding (MoUs): Guidance on Practice and Procedures

(Mar. 19, 2013), https://www.gov.uk/government/publications/treaties-and-mous-guidance-on-

practice-and-procedures (underlining and bolding in original).

182

Government of Canada, Policy on Tabling of Treaties in Parliament, Sec. 6.1,

https://treaty-accord.gc.ca/procedures.aspx?lang=eng.

183

Id., Sec. 8.

184

Id., Annex C.

185

Working Group Survey, supra note 173, at 12. See also Survey by van Ert, supra note

155 (“The [Canadian] foreign ministry’s Treaty Law Division–MOU Unit must be consulted for

review of all non-legally binding instruments prior to their conclusion.”); OAS Report, supra note 30,

at 70, 103-04 (describing centralized review in Ecuador and Peru).

186

See Australian Government, Department of Foreign Affairs and Trade, Guidance Note:

Australia’s Practice for Concluding Less-Than-Treaty Status Instruments,

https://www.dfat.gov.au/international-relations/treaties/australias-practice-concluding-less-than-

treaty-status-instruments.

1-Feb-22] The Rise of Nonbinding International Agreements

any central collection of them. Administrative agencies in Germany follow a similar

process: they are supposed to send proposed nonbinding agreements to a division of

the Foreign Ministry for review and approval.

187

When reviewing agreements, the

Ministry “will look for trigger words usually used only in international treaties, and

‘soften’ them down to a non-legally binding alternative.”

188

The agreements, once

concluded, are supposed to be stored in the archives of the Foreign Ministry. In

other countries, such as Austria and Argentina, centralized foreign ministry review is

encouraged but not required.

189

Some countries have adopted registries of nonbinding agreements, though

most do not make them public. The Czech Republic, for instance, has recently

established a central, non-public registry within the executive branch for nonbinding

agreements, although the Czech Legal Adviser has noted that “not all MoUs in

practice reach my department and get registered.”

190

He also expressed support for

the idea of a public registry for nonbinding agreements, noting: “In most States, there

is an official register of published treaties (an official gazette), nevertheless, there is

a gap when it comes to MoUs, so such register makes sense.” Israel also has

centralized foreign ministry review of nonbinding agreements and maintains an

internal executive branch registry of nonbinding agreements.

191

In Finland, the

government is developing a new document management system that “could make all

Governmental soft law instruments available in one archive.”

192

South Korea’s

foreign ministry manages a central database of agency-to-agency agreements and

each agency is encouraged to input their agreements into the database.

193

Starting in

2014, Germany started storing nonbinding agreements in a central archive.

194

187

German Foreign Ministry Guidelines (2019) (translation on file with authors).

188

Eick, supra note 12, at 3.

189

See Nahuel Maisley, University of Buenos Aires and New York University, Survey for

University of Chicago Law School Conference on “Non-Binding International Agreements: A

Comparative Assessment” (submitted July 28, 2021); Michael Waibel, Universtat Wein, Survey for

University of Chicago Law School Conference on “Non-Binding International Agreements: A

Comparative Assessment” (submitted July 31, 2021).

190

Petr Valek, Director of the International Law Department, Ministry of Foreign Affairs,

Czech Republic (Mar. 26, 2021), https://rm.coe.int/1-3-p-valek-presentation-mous/1680a23585.

191

See Naomi Elimelech Shamra, Director, Treaties Department, Ministry of Foreign Affairs,

Survey for University of Chicago Law School Conference on “Non-Binding International

Agreements: A Comparative Assessment” (submitted July 25, 2021) (on file with authors).

192

Kaija Suvanto, Director General, Legal Service, Ministry for Foreign Affairs for Finland,

Survey for University of Chicago Law School Conference on “Non-Binding International

Agreements: A Comparative Assessment” (submitted July 7, 2021) (on file with authors).

193

See Working Group Survey, supra note 156, at 27.

194

Id.

The Rise of Nonbinding International Agreements [1-Feb-22

Canada reported in 2019 that its Treaty Law Division was “in the process of

developing a digital database of these [nonbinding] instruments” and was “reaching

out to lead divisions, department and agencies to ensure all signed instruments are

included in the database.”

195

The database is not public. Ecuador similarly has a

practice of central executive branch recording of nonbinding agreements.

196

While most of the above registries have been established through informal or

regulatory means, since 2014 a registry has been required by law in Spain. The

“Treaties and Other International Agreements Act” provides that when agencies

conclude nonbinding international agreements (which the Act refers to as “non-

normative agreements”), they must submit them to the Ministry of Foreign Affairs

for inclusion in a central, public registry.

197

As of 2021, however, this registry still

had not been established.

198

C. Transparency

Even if a country’s executive branch monitors and collects nonbinding

international agreements, the public does not necessarily have access. Yet these

agreements sometimes entail significant commitments by the government that can

affect national policy, or at least the interests of particular stakeholders.

Some nations make at least some nonbinding international agreements

available to the public. But as in the United States, publication is usually done

voluntarily rather than pursuant to a legal mandate, and it is typically not

comprehensive. For example, New Zealand maintains a public database that

includes its treaties as well as “a record of some of New Zealand’s legally binding

arrangements,” although it does not include “[m]inor or technical arrangements and

financial or commercially sensitive arrangements.”

199

Similarly, although there is no

legal requirement in Japan to publish nonbinding agreements, many such agreements

are apparently published.

200

195

Id. at 9.

196

See OAS Report, supra note 30, at 114 n.193.

197

See Carlos Esposito, Spanish Foreign Relations Law and the Process for Making Treaties

and Other International Agreements, in THE OXFORD HANDBOOK OF COMPARATIVE FOREIGN

RELATIONS LAW 213-17 (Curtis A. Bradley ed., 2019).

198

See Carlos Esposito, Universidad Autonoma de Madrid, Survey for University of Chicago

Law School Conference on “Non-Binding International Agreements: A Comparative Assessment”

(submitted July 31, 2021).

199

New Zealand Treaties Online, https://www.treaties.mfat.govt.nz/.

200

Email from Ryo Fukahori, Director, Treaties Division, International Legal Affairs Bureau,

to Curtis Bradley (Feb. 28, 2021) (on file with authors).

1-Feb-22] The Rise of Nonbinding International Agreements

In the 1990s, Australia adopted various reforms designed to increase the

transparency and accessibility of its treaties, but those reforms apply only to binding

agreements. Professor Andrew Byrnes has observed that, even though many

nonbinding agreements made by Australia are important, “the publication of these

agreements is sporadic and unsystematic, and the text of many such agreements is

not available to the public on government websites.”

201

Other nations, such as the

United Kingdom, apparently do not routinely publish nonbinding agreements.

202

Although Finland currently has no public registry system for nonbinding

agreements, the government is required by statute to provide public notice of

important foreign relations actions, and the legal adviser for its foreign ministry has

suggested publication “of non-legally binding instruments considered to be of

importance that are made between Governments.”

203

She also has observed that

establishing a more general public registry for nonbinding agreements would serve a

number of useful functions:

It is clear that this kind of a registry comparable to a treaty register

would make access to these political commitments easier and make

them more visible. This is positive from the point of view of

democracy, and transparency as well. Public access to non-legally

binding instruments could also serve the goal of using political

instruments only when they are an appropriate tool to reach the

intended purpose and when there is no need for legally binding

obligations. It could make the practice of using non-legally binding

instruments more coherent in an individual state as well as between

states. In the name of transparency, it would also be interesting to

collect the practice of different countries of publishing these

instruments, such as MoUs, e.g. in their treaty series.

204

201

Andrew Byrnes, Time to Put on 3-D Glasses: Is There a Need to Expand JSCOT’s

Mandate to Cover “Instruments of Less Than Treaty Status”?, 22 AUST. INT’L L.J. 1 (2015-16).

202

See Guidance on Practice and Procedures, supra note 163, at 11 (noting that MoUs are

“not usually published”); see also Arabella Lang, Public Law Project, Survey for University of

Chicago Law School Conference on “Non-Binding International Agreements: A Comparative

Assessment” (submitted July 30, 2021).

203

Suvanto, supra note 192.

204

Kaija Suvanto, Director General, Legal Service, Ministry for Foreign Affairs of Finland

(Mar. 26, 2021), https://rm.coe.int/2-3-suvanto-cahdi-panel-non-legally-binding-

agreements/1680a2358a.

The Rise of Nonbinding International Agreements [1-Feb-22

As noted above, the Czech Legal Adviser has expressed similar sentiments.

In France, although there is no publication system currently in place for nonbinding

agreements, a proposal has been made to require publication of such agreements

except where publication would be incompatible with “secret national defense” or

foreign policy requirements.

205

In South Africa, the government publishes both

binding and nonbinding agreements that have recently been concluded, although it is

unclear how comprehensive this is.

206

The OAS Guidelines recommend that “States

should maintain a national registry of all, or at least the most significant, political

commitments of the State and State institutions.”

207

D. Legislative Participation

Nations vary in the extent to which they involve their legislatures in treaty-

making.

208

Commonwealth countries normally do not require legislative approval,

although a number of these countries (and other countries, such as Israel) as a matter

of custom or statutory mandate inform the legislature about treaties before they are

ratified. In many countries outside the Commonwealth, formal legislative approval

is required for some or all treaties.

209

In countries in which parliamentary approval is

required, treaties typically can operate in some circumstances as domestic law; by

contrast, when parliamentary approval is not required, treaties typically must be

implemented by the legislature before they have domestic effect.

210

These requirements of parliamentary notice and approval, however, typically

apply only to binding agreements. For example, the UK statute that requires that

agreements be laid before Parliament for at least 21 days prior to ratification applies

only to binding agreements.

211

Noting this fact, the European Union Committee of

the British House of Lords observed in 2019 that “[a]ny future Treaties Committee

205

Mathias Forteau, University of Paris Nanterre, Survey for University of Chicago Law

School Conference on “Non-Binding International Agreements: A Comparative Assessment”

(submitted July 31, 2021).

206

See Hannah Woolaver, University of Cape Town, Survey for University of Chicago Law

School Conference on “Non-Binding International Agreements: A Comparative Assessment”

(submitted Aug. 24, 2021).

207

OAS Report, supra note 30, at 113.

208

See Pierre-Hugues Verdier & Mila Versteeg, Separation of Powers, Treaty-Making, and

Treaty Withdrawal, in THE OXFORD HANDBOOK OF COMPARATIVE FOREIGN RELATIONS LAW 138-42

(Curtis A. Bradley ed., 2019); Oona A. Hathaway, Treaties’ End: The Past, Present, and Future of

International Lawmaking in the United States, 117 YALE L.J. 1236, 1271-84 (2008).

209

See Verdier & Versteeg, supra note 191, at 140.

210

See id.

211

See Lang, supra note 192, at 41.

1-Feb-22] The Rise of Nonbinding International Agreements

may wish to consider proportionate means to remedy the resulting scrutiny gap.”

212

More recently, the International Agreements Committee of the House of Lords urged

the executive to report significant nonbinding agreements to parliament, and it

suggested some criteria for what would qualify as significant.

213

In Finland, the Constitution prescribes the process for concluding binding

agreements (which requires legislative approval for some but not all treaties) but is

silent about nonbinding agreements. Finland’s Legal Adviser recently noted that

“there could be merit to inform Parliament of the most significant non-legally

binding instruments,” but she indicated that this happens only on an ad hoc basis in

her country.

214

In some countries, there have recently been calls for more legislative

involvement in nonbinding agreements. For example, in the wake of the controversy

in Switzerland over the Migration Compact, the Foreign Policy Committee of the

Council of States in Switzerland asked the government to report on the “growing role

of soft law in international relations” and “the resulting creeping weakening of

Parliament’s democratic rights.” Switzerland’s Federal Council (a seven-member

executive council) responded by agreeing to increase parliamentary involvement in

the development of soft law, including in the conclusion of nonbinding agreements.

It noted, though, that “[g]iven that there are a large number of soft law instruments

and that they are usually issued under tight deadlines, it would be unfeasible for

Parliament to participate in the creation of these instruments across the board.”

215

But it made a commitment to the legislature that “members of Parliament are to be

consulted more frequently and provided with better documentation and regular

reports on relevant soft law projects,” something that it indicated would not require a

212

House of Lords, European Union Committee, 42nd Report of Session 2017-19, Scrutiny

of International Agreements: Lessons Learned 23 (June 2019),

https://publications.parliament.uk/pa/ld201719/ldselect/ldeucom/387/387.pdf. The Committee also

concluded that, for a variety of reasons, the 2010 law was not well designed to promote parliamentary

scrutiny of binding agreements.

213

See House of Lords, International Agreements Committee, 7th Report of Session 2021-22,

Working Practices: One Year On paras. 73-88 (Sept. 17, 2021),

https://publications.parliament.uk/pa/ld5802/ldselect/ldintagr/75/7502.htm (calling for reporting of a

nonbinding agreement if it either “(a) is politically or economically important; (b) imposes material

obligations on UK citizens or residents; (c) has human rights implications; (d) is directly related to a

treaty; or (e) would give rise to significant expenditure”).

214

Suvanto, supra note 204.

215

The Federal Council, Parliament to Be More Closely Involved in Soft Law Projects (June

27, 2019), at https://www.admin.ch/gov/en/start/documentation/media-releases.msg-id-75590.html.

See Petrig, supra note 179 (discussing “the efforts undertaken in Switzerland to associate Parliament

more closely in the making of ‘soft law’”).

The Rise of Nonbinding International Agreements [1-Feb-22

change in the law.

216

A Swiss parliamentary subcommittee continues to focus on

how to adapt existing procedures and practices in light of the increasing role of

nonbinding agreements.

217

Concerns about evasion of legislative prerogatives appear to be growing in

other countries as well. In Australia, Professor Byrnes has argued before a

committee in Parliament reviewing the issue that, “[b]ecause [non-binding

agreements] involve formal arrangements for the exercise of public power, their texts

should as a matter of principle be made public and thus subject to Parliamentary and

public scrutiny.”

218

Another commentator has observed that the Australian

government sometimes uses nonbinding agreements to avoid political constraints and

has argued for “some kind of accountability regime.”

219

Similar questions are being

raised with respect to the effect of the EU’s conclusion of nonbinding agreements on

that institution’s separation of powers.

220

E. Summary

The above account of comparative practice on nonbinding agreements, while

not comprehensive, shows that other nations are grappling with many of the same

regulatory questions faced by the United States. Around the globe, there is

216

Id.

217

See Roland Portmann, Swiss Federal Department of Foreign Affairs, Survey for

University of Chicago Law School Conference on “Non-Binding International Agreements: A

Comparative Assessment” (submitted Aug. 1, 2021).

218

Byrnes, supra note 183, at 6. Byrnes presented the argument in a paper in the 20th

Anniversary Seminar, Joint Standing Committee on Treaties (Mar. 18, 2016),

https://www.aph.gov.au/Parliamentary_Business/Committees/Joint/Treaties/20th_Anniversary.

219

Peter C. Lundy, International Law: The Netherworld of Nonbinding Agreements (July 17,

2017), https://www.internationalaffairs.org.au/australianoutlook/netherworld-nonbonding-

agreements/.

220

See Ramses A. Wessel, Normative Transformations in EU External Relations: The

Phenomenon of “Soft” International Agreements, 44 W. EUR. POL. 72 (2021); see also Mario

Mendez, Written Evidence (June 8, 2020), House of Lords, EU International Agreements

Subcommittee, https://committees.parliament.uk/writtenevidence/6561/pdf/ (“In the EU concerns are

also being raised about increasing recourse to political agreements given that they are not channeled

through the Article 218 TFEU framework that applies to legally binding agreements.”). Other

international organizations also face issues relating to the rise of nonbinding agreements. See, e.g.,

Statement by Mr. Miguel de Serpa Soares, Under-Secretary-General for Legal Affairs and United

Nations Legal Counsel (Feb. 26, 2021), https://rm.coe.int/0-2-cahdi-expert-workshop-statement-mss-

22march2021/1680a23544 (“[M]y office – as the centralized legal service of the Organization –

regularly reviews legally binding and non-legally binding draft agreements submitted by the various

United Nations Secretariat Departments, Offices, and Regional Commissions, including to avoid

misunderstandings and legal uncertainties.”).

1-Feb-22] The Rise of Nonbinding International Agreements

increasing awareness of a regulatory gap: the laws and practices that states have in

place to ensure coordination, transparency, and legislative involvement in the

making of international agreements are typically focused only on binding

agreements, but their executives increasingly are using nonbinding instruments.

The nations that have to date addressed these issues have mainly focused on

internal coordination. In the countries we surveyed, the foreign ministry has

provided some general guidance to ministries and agencies with respect to the

drafting of nonbinding agreements. These countries differ, however, on the extent to

which they centralize foreign ministry review of nonbinding agreements before they

are concluded, and over whether to have centralized collection of nonbinding

agreements. A number of countries are considering reforms to further improve

internal coordination. As for transparency, most of the nations that we surveyed do

not have any systematic publication of nonbinding agreements, although legal

advisers in some of these countries have stated that such publication might be

desirable. A number of countries are considering whether and how to provide public

access to nonbinding agreements. In most countries surveyed, there is little

legislative involvement with nonbinding agreements, and only a few countries are

entertaining proposals to change this.

IV. LEGAL REFORM

As we have shown throughout this Article, nonbinding agreements are

increasingly important and prevalent and are often used as a substitute for binding

agreements. They are also completely under the control of the executive branch,

entirely unregulated, and increasingly controversial. As discussed in the last Part,

this development has prompted concerns not only in the United States but also in a

number of constitutional democracies.

This Part considers several possibilities for legal reform in the United States.

In light of recent controversies over certain high-profile agreements such as the Iran

nuclear deal and the emissions pledge in the Paris climate change agreement, this

Part first considers how Congress might influence or check presidential uses of

nonbinding agreements that depend on pre-existing delegations of congressional

authority to implement the agreements. Moving to the more common (but less well

known) agency-level agreements, this Part then considers accountability mechanisms

and makes the case that the nonbinding agreements should be treated—in terms of

internal executive branch coordination, and external reporting and transparency—

largely like binding agreements. Finally, this Part examines whether the United

States should work with partner nations to develop international best practices for the

drafting of nonbinding agreements.

The Rise of Nonbinding International Agreements [1-Feb-22

A. Nonbinding Agreements and Domestic Delegation

We explained above how the executive branch—most notably in the Iran

Deal and the emissions reduction pledge in the Paris Agreement—has used

nonbinding commitments in combination with domestic regulations to forge

consequential international cooperation without the contemporary approval, or even

involvement, of Congress.

221

While controversial, this mechanism is generally

lawful. It merely combines two lawful presidential functions—making nonbinding

agreements and exercising statutorily-delegated regulatory authority—in novel

ways.

222

Nevertheless, there are reasons why Congress might want to regulate this

process more closely. First, it is unlikely that Congress contemplated that its

delegated authority to the president in these contexts would be used as the basis to

implement international agreements. Second, when a president relies on pre-existing

domestic delegations to implement new nonbinding international commitments,

Congress can block the agreement only by enacting a new statute, which in many if

not all cases would require it to overcome a presidential veto. Third, the use of

domestic delegations to implement nonbinding agreements portends a potentially

broad shift of agreement-making power to the president since there are no limits on

the president’s power to make nonbinding agreements, and since Congress has

delegated regulatory authority to the president in broad terms across a range of

topics. These three points taken together underscore that Congress is at a significant

structural disadvantage in the face of novel uses of extant delegated authority to

implement nonbinding agreements.

But Congress is not powerless. It can, if it wishes, and if it can overcome

possible vetoes, curb this presidential power through legislation that alters the prior

delegations. First, it can narrow or clarify discrete delegations to make them less

susceptible to use as the basis for implementing a nonbinding agreement. If

Congress were truly concerned about the Paris Agreement, for example, it could

amend the Clean Air Act to specify that it could not be the basis for the carbon

reduction elements of the Clean Power Plan, the primary regulatory vehicle for

implementing the emissions reduction pledge in the Paris Agreement. Second,

221

See supra text accompanying notes 95-98.

222

It is lawful, that is, as long as the agreement is in fact nonbinding and the president

properly exercises the authority delegated by Congress. Cf. Samuel Estreicher & Steven Menashi,

Taking Steel Seizure Seriously: The Iran Nuclear Agreement and the Separation of Powers, 86

FORDHAM L. REV. 1199, 1203, 1204 (2017) (arguing that the Iran nuclear agreement exceeded

presidential authority because “it is not clear that the [agreement] is a nonbinding political

commitment” and because “the President’s across-the-board exercise of waiver authority contradicts

the expressed intent of Congress in the sanctions statutes”).

1-Feb-22] The Rise of Nonbinding International Agreements

Congress could specify that particular delegations of authority cannot be the basis for

implementing a nonbinding international agreement without new congressional

approval. This is precisely the power that Congress leveraged when it passed the

Iran Nuclear Agreement Review Act, which suspended authority that Congress had

previously delegated to the president to waive U.S. sanctions on Iran while Congress

reviewed the draft nonbinding Iran Nuclear Deal. Third, and most aggressively,

Congress could enact a statute that made clear that none of its domestic delegations

to the president could be the basis for implementing a nonbinding international

agreement.

The last option obviously has the broadest implications; its practical impact

would depend on the unknowable extent to which future presidents wanted to build

on the Obama administration’s innovations to implement consequential nonbinding

agreements through domestic regulations. The policy desirability of the first two

options is similarly impossible to assess divorced from the particular application.

Our point is simply that if Congress decides that it wants to regain some of the

authority claimed by novel uses of nonbinding agreements, it has legally available

options. None of these options would interfere with the president’s power to make

nonbinding international agreements, or to conduct negotiations in connection with

those agreements, or to implement or enforce nonbinding agreements through

exercises of the president’s Article II power.

223

Rather, they would simply alter the

terms of domestic statutory delegations that clearly fall within Congress’s Article I

powers. Congress is not required to delegate these various forms of regulatory

authority in the first place and thus almost certainly has the authority to restrict the

uses to which the delegated authority is put.

B. Accountability Issues: Coordination, Reporting, Transparency

Despite their increasing importance, and in sharp contrast to binding

agreements, nonbinding agreements—especially agreements concluded at the agency

level—lack formal coordination or review within the executive branch, are not

regularly reported to the State Department, need not be reported to Congress or

223

They thus would not run afoul of the executive branch’s contestable claims of a very

broad exclusive power to conduct the nation’s diplomatic relations. See generally Jean Galbraith, The

Runaway Presidential Power over Diplomacy, 108 VA. L. REV. (forthcoming 2022) (analyzing and

criticizing this asserted power). Of most relevance here, congressional changes to the terms of

domestic law delegations would in no way interfere with the executive branch’s assertion of

“exclusive authority to determine the time, scope, and objectives of international negotiations or

discussions.” Constitutionality of Section 7054 of the Fiscal Year 2009 Foreign Appropriations Act

(June 1, 2009), Memorandum from David J. Barron, Acting Assistant Attorney General, Office of

Legal Counsel, to Acting Legal Adviser, Department of States, Regarding Constitutionality of Section

7054 of the Fiscal Year 2009 Foreign Appropriations Act, at 8 (June 1, 2009),

https://www.justice.gov/file/18496/download.

The Rise of Nonbinding International Agreements [1-Feb-22

published, and are not always recorded and organized even within the agency that

makes them. As in other countries, U.S. law has simply not caught up with how the

executive branch now makes most of its international commitments, including many

of its most important ones.

224

This Section makes the case for bringing the regulatory regime for

nonbinding agreements closer to the one that applies to binding agreements. The

analysis that follows focuses on what we see as the two biggest problems in this area:

(1) the lack of coordination of nonbinding agreements within the executive branch;

and (2) the inapplicability of the accountability mechanisms that apply to binding

agreements and the opportunity this gap creates for evading oversight by

characterizing an agreement as nonbinding.

1. Internal Coordination. This Subsection argues for greater internal

coordination of nonbinding agreements, then explains how the rules for internal

coordination should operate, and then discusses the difficult problem of how to

define the scope of nonbinding agreements for purposes of regulation.

a. Coordinating Nonbinding Agreements. As noted above, the State

Department’s “C-175” process that applies to binding agreements requires an

executive agency to (i) seek and obtain approval from the State Department before

negotiating an agreement, (ii) seek and obtain State Department approval before

concluding the agreement, and (iii) transmit a copy of the finalized agreement to the

State Department for central collection. These requirements serve many aims.

Among other things, they “facilitate the application of orderly and uniform measures

and procedures” for the negotiation and conclusion of binding agreements;

225

ensure

complete and accurate records of these agreements;

226

ensure that the agreements are

“carried out within constitutional and other appropriate limits” and do not conflict

224

Legislation was recently proposed that, if adopted, would implement various proposals

that we made in earlier work with respect to the transparency of binding agreements, see Hathaway,

Bradley & Goldsmith, supra note 2, and would also require reporting and publication of some

nonbinding agreements that either “could reasonably be expected to have a significant impact on the

foreign policy of the United States” or are “the subject of a written communication from the Chair or

Ranking Member of either of the appropriate congressional committees to the Secretary [of State].”

See S. 1260, § 3310 (approved by Senate on June 8, 2021). The fate of this proposed legislation is

unclear and, in any event, we argue in this Part for the adoption of broader accountability and

transparency measures. Nevertheless, the fact that the issue is being taken seriously in Congress

suggests that meaningful reform may be possible in this area.

225

11 U.S. Dep’t of State, Foreign Affairs Manual § 721(a),

https://fam.state.gov/fam/11fam/11fam0720.html.

226

Id.

1-Feb-22] The Rise of Nonbinding International Agreements

with other agreements or U.S. law;

227

and ensure that they can be properly reported

to Congress and published, as required by statute.

228

Despite their growing prevalence and importance, none of these rules apply

to nonbinding agreements. The State Department notes that nonbinding agreements

“may carry significant moral or political weight” in international relations and urges

executive branch agencies to avoid “ambiguity as to whether or not a document is

legally binding.”

229

To this end, as noted above, the Department provides

recommendations about how to draft nonbinding agreements to avoid ambiguity,

which can cause confusion about the nature of agreements and the consequences for

non-compliance, among other things. As also noted above, the State Department

encourages but does not require agencies to share the texts of proposed nonbinding

agreements with the State Department before making them.

Our database and interviews provide reason to believe that these thin

suggestions by the State Department are often ignored and are thus not serving their

goals. For instance, as noted above, almost 65% of the nonbinding agreements in

our database use the term “will” even though the State Department counsels against

doing so. The absence of mandatory terminological practice across agencies can

cause confusion among U.S. agreement partners. In addition, this uncertainty can

make it difficult for the State Department to ensure that the executive branch is

complying with its obligations under the Case-Zablocki Act to report binding

agreements. All of these problems are likely worse than described in Part II, because

the agencies that make available public information about their nonbinding

agreements and respond on time to FOIA requests are likely those with less to hide.

Indeed, some of the most troubling insights have come with regard to the most

nontransparent and recalcitrant agencies—particularly the DoD and DHS. Indeed, in

response to the complaint we filed against the DoD, it identified over 6000

documents that it believed may be responsive to the FOIA request, but it was unsure

whether the agreements it had identified were binding or nonbinding. In short, what

we have learned is suggestive of many more pathologies.

Nonbinding agreements are too important and too prevalent to justify their

radically differential treatment compared to binding agreements when it comes to

internal coordination. Nonbinding agreements entail foreign policy commitments to

other countries, and there should be some way to ensure that they are consistent with

the executive branch’s policy goals and with each other. We note, too, that, as we

227

Id. at § 722(1)-(2).

228

Id. at § 722(9); see also id. at § 711(a).

229

State Department Guidance, supra note 64.

The Rise of Nonbinding International Agreements [1-Feb-22

discussed in Part III, other countries are increasingly adopting internal registries of

nonbinding agreements for similar reasons. Indeed, according to our surveys, this is

the most prevalent reform currently under discussion abroad.

More formal coordination within the executive branch would serve many of

the most important aims of the coordination regime for binding agreements,

including ensuring uniformity in U.S. nonbinding agreements; facilitating a complete

and accurate record of these agreements, so that the executive branch knows what it

has committed to and can ensure that it is not undertaking problematic or

inconsistent commitments; ensuring that the nonbinding agreements are consistent

with U.S. binding obligations and with domestic U.S. law; ensuring that agencies do

not unintentionally create binding agreements; and creating a central repository of

nonbinding agreements so that the government can meet any reporting or publication

requirements that Congress might impose on nonbinding agreements, an issue we

discuss below. Importantly, internal coordination would also serve Congress’s aims

in the Case-Zablocki Act, since some under-reporting under the Act is likely a result

of incorrect assumptions by departments and agencies about whether an agreement is

binding.

230

These coordination goals can be achieved with a less robust scheme than the

one that applies to binding agreements. First, the State Department’s guidance on

suggested terminology and form for ensuring that agreements are not binding should,

going forward, be made mandatory. This will help make nonbinding agreements

uniform across the government and help avoid the problem of uncertainty within the

government about whether an agreement is binding or not. If the State Department’s

guidance is made mandatory, we think it unnecessary to require that the State

Department conduct pre-negotiation and pre-conclusion review, as it does for

binding agreements. The resource demands for such review would be enormous

given the high number of nonbinding agreements. And the need for such review is

not nearly as great for nonbinding agreements as for binding agreements, given that

nonbinding agreements do not raise the same complex legal issues about whether

binding agreements are supported by congressional authorization or fit within the

narrow constitutional scope of binding executive agreements. Of course, voluntary

review, especially in cases where there is uncertainty about whether an agreement is

binding or not, should continue to be an option.

Second, agencies should report nonbinding agreements to the State

Department upon conclusion of the agreement. This is so for reasons just stated: one

entity in the government, the same entity that is the repository for binding

agreements, should have a comprehensive eye on the international commitments

230

See Hathaway, Bradley & Goldsmith, supra note 2.

1-Feb-22] The Rise of Nonbinding International Agreements

made by various corners of the executive branch in order to ensure that they are

consistent with U.S. foreign policy aims and with one another.

231

We recognize that

many agencies will resist this requirement, because they do not want to invite State

Department involvement. But the benefits of internal coordination outlined above

outweigh bureaucratic self-interest in this instance.

b. Defining Nonbinding Agreements. One of the major challenges for

accountability reform for nonbinding agreements is how to define the category of

nonbinding agreements subject to regulation. Nonbinding agreements share

characteristics with other forms of diplomatic speech by the executive—for example,

a president’s oral commitments to a foreign leader in a phone call—that it would be

inappropriate to regulate. What is needed is a definition that captures the nonbinding

agreements that are akin to binding ones in function and form.

Before moving to our proposed definition, it is important to note that this

problem is not unique to nonbinding agreements. A similar definitional problem

arises with respect to the regulation of binding agreements. The line between the

binding agreements that under current law must be reported to the Congress and in

some instances published, and nonbinding agreements of various sorts, is not always

clear. To ensure proper internal coordination of binding agreements, and in order to

report and publish those agreements, the State Department has established five

criteria, plus other considerations, “to provide guidance” for identifying binding U.S.

agreements reportable to the State Department.

232

One criterion is the “significance”

of the arrangement, a judgment that is based on “the entire context” of the

agreement, and that excludes “minor or trivial undertakings.”

233

The 1976 State

Department memorandum that established these criteria noted that “difficult

judgments will have to be made in many cases.”

234

The same will be true of

nonbinding agreements. But as with binding ones, this objection should not be fatal,

and definitional problems can be worked out though practice and regulation.

Our experience with the FOIA process suggests that it is possible to define a

category of nonbinding agreements that is clear and meaningful. Initially, the

absence of an adequately clear definition of a “nonbinding agreement” was an issue

for some of our FOIA requests to the agencies. Several agencies asked for additional

231

We propose that these internal coordination mechanisms be implemented regardless of

whether the accountability steps proposed below are enacted.

232

See 22 CFR § 181.2(a). The five criteria are: the identity and intention of the parties; the

significance of the arrangement; the precision and specificity of the language; the existence of two or

more parties; and considerations relating to the form of the agreement. Id. at (1)-(5).

233

Id. at § 181.2(a)(2).

234

Leigh Memorandum, supra note 61.

The Rise of Nonbinding International Agreements [1-Feb-22

information on the ground that our initial requests were not sufficiently specific.

235

In every case where an agency gave this response, we provided clarification that

allowed the agency to find and report the agreements we hoped to receive. Our

proposed definition builds on this clarified request to the agencies.

236

Any definition of regulable nonbinding agreements should seek to capture

our first and third categories of nonbinding agreements: high-level formal

agreements and nonbinding agreements concluded by administrative agencies.

237

The definition should contain the following elements:

First, the agreement must be reduced to writing. In other words, there must

be a shared written text.

Second, the agreement must have at least two parties or participants, at least

one of which is the U.S. federal government or one of its departments, agencies, or

sub-entities (e.g., a federal agency or office within an agency), and at least one of

which is a sovereign state or sub-entity, or an international organization.

Third, if a document has been reported to Congress as an international

agreement under the Case-Zablocki Act, it is excluded.

235

See, e.g., Roberta Parsons, Acting Deputy Chief FOIA Officer and Acting Deputy

Director for FOIA/PA operations, Office of Privacy and Open Government, U.S. Department of

Commerce, to Oona A. Hathaway (Feb. 11, 2021) (“Specifically, we need more description about

what constitutes a non-binding arrangement or understanding for purposes of this request that

provides guidance for conducting a search, so that the level of effort required to locate responsive

documents is reasonable.”). The initial requests provided a background statement describing

nonbinding agreements and then requested “any and all unclassified nonbinding understandings or

arrangements with foreign countries or international organizations actually in” the agency’s

possession. The requests specified that they included “Any and all understandings or arrangements

that are nonbinding based on the text or drafting history, as described in the Background statement

above, or that comply with the definition set forth in the MOU Guidance document.” They also

included “Any and all understandings or arrangements transmitted to the Department of State and

determined to be nonbinding and therefore not reportable to Congress as an international agreement

under the Case-Zablocki Act.” The requests specifically excluded agreements reported to Congress

under the Case-Zablocki Act and agreements solely in oral form.

236

The clarifications included additional information such as common title language, the

presence of formal signature lines together with common accompanying language, common phrases

indicating the agreement is meant to be nonbinding, the presence of a date and common

accompanying language, and links to illustrative examples of nonbinding agreements.

237

What we have called category two agreements—joint statements and communiques—are

typically coordinated within the executive branch, and they are by definition transparent. As a result,

they do not need additional regulation.

1-Feb-22] The Rise of Nonbinding International Agreements

Fourth, the agreement must contain formal elements normally associated with

binding agreements, such as:

Title: The document’s title includes the following terms or similar terms:

“Memorandum of Understanding,” “Memorandum of Agreement,”

“Memorandum of Cooperation,” “Cooperative Arrangement,”

“Memorandum of Intent,” “Statement of Intent,” “Administrative

Arrangement,” “Declaration on Cooperation,” “Understanding,” “MoU,”

“Cooperation Arrangement,” “Framework Cooperation Arrangement,”

“Declaration of Principles,” “Joint Declaration,” “Administrative

Arrangement,” “International Plan of Action,” “Terms of Reference,”

“Joint Contingency Plan,” or “Arrangement.”

Body: The document is divided into “sections,” “articles,” or other

numbered or lettered parts.

Date: The document contains a date, sometimes expressly identifying an

“effective date,” date of “entry into force,” or something similar.

Signature Line: The document contains a signature line or similar

indication of conclusion:

• The phrase “Signed in,” “Signed at,” or “Done on,” or “Done at”

preceding a location and a date, such as “Signed in Washington,

D.C. on July 31, 2009,” “Signed in duplicate at Washington, D.C.,

the 1st day of September, two thousand and sixteen,” “Signed at

Washington, D.C., in duplicate, this 19th day of August, 2014,” or

“Done on the 23

Day of November, 2009,” or similar term.

• The signature line or similar indication of conclusion includes two

or more signature lines, including a signature line featuring “For the

Government of the United States of America”, “For the U.S.

Department of Commerce”, “For the Department of Commerce of

the United States”, or “For the Office of the General Counsel of the

U.S. Department of Commerce,” or similar term.

These criteria are far from determinant. But neither are the criteria for

identifying binding agreements that have governed for decades for purposes of

internal coordination and external reporting and publication. These criteria suffice to

capture the vast majority of nonbinding agreements for which there is currently no

regulation, and which should be coordinated with the State Department and possibly,

as we now turn to discuss, reported to Congress or published.

The Rise of Nonbinding International Agreements [1-Feb-22

2. Accountability and Evasion. The second major problem with the non-

regulation of nonbinding agreements is that they lack any regular accountability

outside the executive branch. This is so even though, as noted throughout this

article, nonbinding agreements are often as important as binding ones, and

sometimes more so. Neither Congress nor the public has any firm sense of what

nonbinding agreements the executive branch makes, or whether they are actually

nonbinding as opposed to binding, or whether they serve the national interest. The

accountability deficit is most severe in those instances where the executive branch

exploits the blurry line between binding and nonbinding agreements to avoid its

statutory reporting and publication duties under extant law. We have cited several

instances of this phenomenon in this Article, and such regulatory evasion through the

making of nonbinding agreements has been a repeated subject of concern on Capitol

Hill.

Our experience with seeking nonbinding agreements through the FOIA

process highlights the inaccessibility of these materials through ordinary

accountability tools. Under the current system, the public has little to no practical

access to information about nonbinding agreements.

We consider two possible reforms to reduce the accountability deficit and,

especially, the use of nonbinding agreements to evade the legal and regulatory

framework for binding agreements: (1) an ex post reporting requirement of defined

nonbinding agreements to Congress that tracks the requirement for binding

agreements; and (2) an ex post publication requirement of nonbinding agreements

that tracks the requirement for binding agreements.

238

a. Ex Post Reporting to Congress. The argument for an executive branch

duty to report to Congress all defined nonbinding agreements after they are

concluded is straightforward. When it enacted the Case-Zablocki Act in 1972 and

strengthened its provisions several times during the years since, Congress

demonstrated concern about secret international agreements made on behalf of the

United States that were not subject to any scrutiny outside the executive branch. It

believed that such scrutiny was needed to make sure that the agreements were lawful

and in the national interest, and that it would have disciplining effects on executive

officials making such agreements. And Congress believed that knowledge of such

238

It may be that ex ante review of nonbinding agreements would be warranted in a narrow

set of circumstances where the President seeks to evade congressional prerogatives, but if robust ex

post reporting and publication rules are adopted, such instances should be rare. If that prediction

proves wrong, Congress may consider targeted ex ante reporting mechanisms, as it has in related

contexts. See Hathaway, Bradley & Goldsmith, supra note 2, at 656-67 and accompanying notes.

Such mechanisms raise constitutional issues that we think can be overcome but that would require

careful consideration.

1-Feb-22] The Rise of Nonbinding International Agreements

agreements was necessary to carry out its constitutional duties related to foreign

relations and national security.

This logic now applies to nonbinding agreements as well. An ex post

reporting requirement would impose discipline on the making of nonbinding

agreements, for the agreement makers would know that Congress would potentially

be scrutinizing their work. It would allow Congress to know about what the

executive branch is doing in this area and hold hearings or draft legislation within

Congress’s authorities if it disagreed with executive branch policy. And it would

allow Congress to better ensure that the executive branch is meeting its reporting and

publication duties for binding agreements, for Congress would have full insight into

where and how the executive branch draws the line between binding and nonbinding

agreements, and whether it sometimes inappropriately crossed that line.

The costs of reporting nonbinding agreements to Congress would be

relatively modest if the State Department (or Congress) mandates that agencies

report defined nonbinding agreements to the State Department in accordance with

the definitions and criteria outlined above, a process that will be similar to the

existing one under the Case-Zablocki Act for binding agreements. The agency

offices that conclude binding agreements are nearly always the same ones that

conclude nonbinding ones—hence the agency officials are accustomed to the Case-

Zablocki Act reporting process. Indeed, eliminating the binding/nonbinding

distinction for internal reporting could lower decision costs for agencies. Agencies

would no longer have to puzzle over whether they are required to report a given

agreement to the State Department—and they no longer would have to tie up State

Department lawyers’ time with helping them determine whether an agreement is

binding or not. Offices that conclude binding and nonbinding agreements could

simply report them all to the State Department.

b. Ex Post Publication. A more aggressive accountability step would be to

require the executive branch to go beyond mere reporting to Congress and instead

publish all defined nonbinding agreements. Any such publication duty for

nonbinding agreements, we contend, should in material respects mirror the

publication duties for bindings.

239

Accordingly, the duty to publish would extend

only to nonbinding agreements as defined above, and, as with binding agreements,

239

In prior work we have called for broader publication duties for binding agreements that

would eliminate all carveouts other than for classified agreements. See Hathaway, Bradley &

Goldsmith, supra note 2, at 701-04. If these proposals were adopted, the same logic for the expanded

duty would apply to nonbinding agreements.

The Rise of Nonbinding International Agreements [1-Feb-22

would further exclude classified nonbinding agreements or those that the State

Department determines, with regulatory guidance, lack sufficient public interest.

240

The case for such publication begins with the case for publishing binding

agreements. There is currently a statutory duty to publish relatively important

binding agreements. In a prior article, we proposed that the Department of State

should publish all non-classified binding agreements. The arguments for doing so

rested on instrumental and noninstrumental grounds. The main instrumental

argument was accountability: public scrutiny on top of congressional scrutiny was

needed to ensure that the United States was making lawful international agreements

in the national interest. Public scrutiny was especially appropriate, we argued,

because Congress alone lacked the resources to examine the mass of agreements

made by the executive, and public scrutiny by journalists and interest groups could

identify and bring to light problematic nonbinding agreements.

241

The main

noninstrumental argument was that the publicity of law is a core requirement of the

rule of law, and secret law demands special justification.

Many of the same instrumental reasons for reporting binding agreements to

Congress—namely, scrutiny of the agreements by an entity outside the executive

branch to ensure that they serve the national interest and are truly nonbinding—apply

equally to nonbinding agreements. Public scrutiny of nonbinding agreements is

especially appropriate because Congress alone lacks the resources to examine the

mass of nonbinding agreements made by the executive, and public scrutiny by

journalists and interest groups can identify and bring to light problematic ones.

242

But we acknowledge that there are differences between binding and

nonbinding agreements related to publication. First, nonbinding agreements are by

definition not law, so they cannot count as secret law if not published. Traditionally,

the U.S. legal and foreign policy system has tolerated secret foreign policies more

than secret legal agreements. On the other hand, the State Department itself has, as

noted earlier, acknowledged that “[w]hile not binding under international law, a non-

binding instrument may carry significant moral or political weight.”

243

In addition,

as we have also explained, nonbinding agreements can have a variety of indirect

legal effects as a matter of both domestic and international law. Indeed, these effects

240

For a detailed explanation of the publication rules for binding agreements, including an

explanation for these carveouts, see Hathaway, Bradley & Goldsmith, supra note 2, at 646-47.

241

See Hathaway, id. at 701-04.

242

For additional discussion of the potential benefits of publication of nonbinding

agreements, for both domestic accountability and international relations, see OAS Report, supra note

30, at 115.

243

State Department Guidance, supra note 64.

1-Feb-22] The Rise of Nonbinding International Agreements

are part of what make nonbinding agreements attractive alternatives to binding ones.

Allowing these nonbinding agreements to remain hidden can be seen as a form of

secret law even though they do not create direct binding obligations, because of the

significant indirect effects these agreements can have on both domestic and

international law.

Second, one important justification for nonbinding agreements in the

scholarly literature and in some of our interviews with domestic and foreign officials

is that they, in contrast to binding agreements, are easier to keep secret. There is a

general presumption in international law, reflected in the U.N. Charter and other

instruments, that binding international agreements should be published, or at least

not kept secret.

244

There is presently no such international presumption for

nonbinding agreements (although, as noted in Part III, the issue of transparency for

nonbinding agreements is a matter of current debate in many countries). Imposing a

publication requirement on nonbinding agreements might change both their

effectiveness and import. Publicity after the fact might make some nonbinding

agreements harder or impossible to make, therefore diminishing their effectiveness

as a tool for cooperation. It might also lead nonbinding agreements to be treated

more seriously than if they were secret, thus diminishing another element of their

flexibility.

There would also be resource costs associated with extending reporting and

publication duties on a large class of additional instruments, including the costs

associated with determining which ones qualify. It is difficult to know how large

these costs would be. Several other nations have more centralized review of such

agreements than in the United States, so it is presumably feasible. Moreover, as we

have noted, the same bureaucratic personnel who would be responsible for

collecting, reporting, and publishing these agreements are already engaged in similar

tasks with respect to the binding agreements, so there would be an economy of scale.

Despite the tradeoffs, we conclude that there is no persuasive argument for

blanket exclusion of nonbinding agreements from transparency requirements, as is

presently the case.

245

We acknowledge that there may be instances where

244

Article 102 of the U.N. Charter states that “Every treaty or international agreement,

whatever its form and descriptive name, entered into by one or more Members of the United Nations

… shall as soon as possible be registered with the Secretariat in accordance with these regulations.”

This provision has been interpreted to apply to binding but not nonbinding agreements.

245

There are already a few isolated instances in which Congress has regulated U.S.

participation in nonbinding arrangements, at least indirectly. For example, for U.S. participation in

the Codex Alimentarius—which sets nonbinding international food safety standards—Congress has

required the FDA to give notice and an opportunity to comment on U.S. negotiating objectives. See

Jean Galbraith, From Treaties to International Commitments: The Changing Landscape of Foreign

Relations Law, 84 U. CHI. L. REV. 1675, 1693 (2017).

The Rise of Nonbinding International Agreements [1-Feb-22

confidentiality of nonbinding agreements is warranted for appropriate diplomatic

purposes. For binding agreements, Congress has accommodated the executive

branch’s desire for confidentiality by allowing it to report some agreements only to

the Senate Committee on Foreign Relations and the House Committee on

International Relations, under an injunction of secrecy, when “the immediate public

disclosure of [the agreement] would, in the opinion of the President, be prejudicial to

the national security of the United States.”

246

A similar accommodation should be

made for nonbinding agreements.

247

While additional exceptions to the defined class of nonbinding agreements

might be sensible, the burden should be on the executive branch to explain why they

are necessary. Currently, all nonbinding agreements are exempt from both reporting

and publication, even though the vast majority of them do not appear to raise any

serious confidentiality concerns. In short, the same basic rules regarding publication

should apply to nonbinding agreements as apply to binding ones. This will provide

better public accountability as well as reduce the incentives for the executive to use

nonbinding agreements merely to avoid legal and regulatory rules that apply to

binding agreements.

c. The Possibility of Circumvention. One advantage of our proposed reforms

is that they reduce the danger of executive branch circumvention of the regulatory

regime that currently governs binding agreements. This invites the question,

however, whether these reforms too might be circumvented. Perhaps agencies will

resort to telephone calls, unexchanged bullet points, and other informal means so that

their agreements fall outside our definition, and thus outside of the regulations?

This sort of danger of circumvention is inherent in any regulatory regime in

this area, but we think that the danger is not significant here, for several reasons.

First, and most importantly, agencies derive substantial benefits from memorializing

their agreements in common written texts. This gives them, their staff, and their

successors—and those of their counterparts—a common reference for cooperation on

an ongoing basis. Moreover, the contractual form likely helps give the agreements

more normative force, which promotes compliance, especially within a bureaucracy.

Second, based on our interviews of agency officials, the lack of transparency is often

due to the lack of any legal structure in place to mandate or provide a process for

246

1 U.S.C. § 112b(a).

247

Indeed, Congress could provide more categorical carveouts from publication. If it does

so, the agreements should nonetheless be reported, confidentially, to the Senate Foreign Relations

Committee and House Foreign Affairs Committee. Congress might also require that there be made

public some general information about confidential nonbinding agreements—for example, how many

have been concluded and on what topics.

1-Feb-22] The Rise of Nonbinding International Agreements

disclosure, not a conscious effort at secrecy. Third, the more informal arrangements

that could be adopted to evade the new transparency requirements on nonbinding

agreements are less likely to have the many indirect domestic and international legal

effects that make nonbinding agreements of particular concern.

Last, while we recognize that this transparency regime is a modest reform, it

will give rise to a great deal more information than currently exists regarding the use

of nonbinding agreements. That, in turn, will provide information on which

Congress could base further regulatory reforms, if warranted.

C. International Best Practices

The information on comparative practices described in Part III demonstrates

that many states are struggling with many of the same issues as the United States

when it comes to nonbinding agreements. Moreover, we uncovered some clearly

inconsistent assumptions about nonbinding agreements among foreign partners that

could give rise to misunderstandings about the nature of an agreement. To take one

example, Mexican negotiators avoid using the title “Memorandum of

Understanding” for nonbinding agreements, because for them it connotes that the

agreement is binding, whereas many other states have the opposite view. Such

differences can lead to misunderstandings and diplomatic tension. Indeed, we have

documented several instances where the United States and its partners did not have a

common understanding with its partners about whether an agreement is binding,

248

and of course there may be many examples that we are unaware of due to the lack of

transparency of these agreements.

To address these concerns, the U.S. government should consider working

with other nations to develop a set of international best practices for the drafting of

nonbinding agreements. Indeed, the Inter-American Juridical Committee of the OAS

has recently suggested some best practices, including with respect to the terminology

that should be used and avoided when drafting nonbinding agreements.

249

rapporteur Duncan Hollis noted, “[w]ithout further clarifications and elaboration,

there are legitimate concerns that existing agreement practices may lead to

inconsistent understandings, unaligned expectations, and even disputes among OAS

Member States, to say nothing of the international community as a whole.”

250

The

OAS guidelines could potentially serve as a foundation for future discussions on

drafting practices that extend beyond that region. At a recent meeting of the Council

248

See supra text accompanying notes 41-39.

249

See OAS Report, supra note 30, at 28-29.

250

Id. at 10.

The Rise of Nonbinding International Agreements [1-Feb-22

of Europe’s Committee of Legal Advisers on Public International Law (CAHDI), the

Director General of the Legal Service of Finland’s Ministry of Foreign Affairs

expressed the view that “uniform state practice in the field of non-legally binding

instruments [is] desirable,” and she expressed the “hope the OAS guidelines will

pave the way for a similar process” in Europe.

251

Many nations have their own

internal drafting guidance, and it might make sense as a first step to compile

examples of this guidance from around the world to identify commonalities and

differences. Indeed, this was one of the suggestions made at the CAHDI meeting.

252

In the absence of an international agreement on best practices, some countries

are borrowing from the guidelines developed in other countries. The director of the

international law department in the Czech foreign ministry has reported, for example,

that he borrows from guidance developed in the UK’s Foreign and Commonwealth

Office.

253

Therefore, another reason for the United States to become involved in

discussions of international best practices would be so that its own views, such as

those reflected in the current State Department Guidance,

254

can potentially influence

the resulting international standards.

To be sure, there may be instances in which the United States desires

ambiguity about whether an agreement is binding—perhaps for domestic political

reasons either in the United States or in the partner country. We have identified

some potential examples in this Article—for example, the Iran nuclear deal and the

migration agreement with Mexico, where the different positions about bindingness

may have stemmed more from domestic political considerations than from

ambiguities in drafting.

255

But the adoption of best practice guidelines would not

eliminate the possibility of such strategic ambiguity; in those instances, the United

States could simply decide not to follow the best practices. We acknowledge,

however, that ambiguity will be harder to maintain if best practices are widely

followed. In the vast majority of cases, including in essentially all of the agency-to-

agency agreements that have been a particular focus of this Article, it is in the

interest of the United States to have a common understanding with its partners about

whether an agreement is binding.

251

Suvanto, supra note 192.

252

See Eick, supra note 12 (suggesting this as a first step towards standardization).

253

See Valek, supra note 190.

254

See State Department Guidance, supra note 64. As we have noted, this guidance is

strangely located only in the archived portions of the State Department’s website.

255

See supra text accompanying notes 42-43.

1-Feb-22] The Rise of Nonbinding International Agreements

CONCLUSION

The current domestic legal framework for international agreements is a

product of the 1970s, when the central issue was how to address the shift from

Article II treaties to executive agreements. We are in the midst of another

fundamental change in how the United States makes international agreements.

While executive agreements remain an important part of U.S. foreign relations, they

are increasingly being eclipsed by nonbinding agreements—agreements that are

often identical in form and function to binding agreements, but not subject to any

legal regulation. As in other countries, it is time for the United States to begin

considering legal reforms that will ensure sufficient coordination, transparency, and

legislative oversight for this vital mechanism for making international commitments.

It is also time to reorient the field of international law to take account of the

rise of nonbinding agreements. The growing use of these agreements has potentially

profound implications for the future of the international legal system. Both the

teaching and study of international law need to be updated to more accurately reflect

how nations today are making commitments, and core assumptions underlying the

field—including most fundamentally the assumption that international law is

operating as law in constraining the behavior of nations—need to be revisited in light

of what appears to be an increasing shift toward nonbinding agreements in

international cooperation.

There are a range of changes to the field that should be prompted by the rise

of nonbinding agreements: First, the teaching of international law courses should be

updated to contain more emphasis on these agreements. At present, nonbinding

agreements are treated as an afterthought in most international law casebooks, if they

are discussed at all. Not only should nonbinding agreements be taught, but there

should be more attention to the ways in which such agreements interact with binding

agreements—and more generally to the ways in which law and diplomacy intersect.

Second, these agreements should receive significantly more scholarly

attention. We hope that the information in our databases of nonbinding agreements

will motivate scholars to further analyze them and the uses to which they are put.

There is a great deal of terrain still to explore. This Article has focused on the

nonbinding documents that are most akin to binding international agreements, but

there is a wide range of other nonbinding documents concluded during international

cooperation that remain to be documented and explored. This includes unilateral

statements, exchanges of notes or letters, and oral arrangements, to name a few.

Exploring and documenting this wider range of nonbinding documents would deepen

the field’s understanding of how such documents shape international diplomacy.

The Rise of Nonbinding International Agreements [1-Feb-22

Third, theories about international law and international cooperation need to

be revised to take account of nonbinding agreements. While there has been some

scholarship on why states choose to conclude nonbinding agreements over binding

ones, that scholarship remains sparse, and relatively little has been written on the

subject in recent years. There are a range of questions waiting to be answered,

including: Are their contexts or subject areas that are more (or less) amenable to

nonbinding agreements? To what extent are nonbinding agreements substitutes for

binding agreements, and to what extent do they pave the way to binding agreements?

Are states more likely to enter into nonbinding agreements with certain countries

and, if so, why? In addition, there are deeper theoretical issues to be explored. What

effect does the legal bindingness of a commitment have on the nature of the

commitment? Does the presence of a legal commitment affect state behavior in

some way that a nonlegal commitment (such as that in a nonbinding agreement) does

not?

In short, the rise of nonbinding international agreements opens up a vast

range of new questions for international law teachers, scholars, and practitioners to

explore.