Jump to Content Jump to Main Navigation

Part I Foundational Issues, 3 Alternatives to Treaty-Making— Informal Agreements

Timothy Meyer

From: The Oxford Guide to Treaties (2nd Edition)

Edited By: Duncan B. Hollis

From: Oxford Public International Law (http://opil.ouplaw.com). (c) Oxford University Press, 2021. All Rights Reserved.date: 28 November 2021

International co-operation — International law and international relations — Sources of international law — Consensual arrangements other than treaties

(p. 59) Alternatives to Treaty-Making— Informal Agreements

In an Oval Office press conference on the status of trade negotiations with China, US President Donald Trump brought the subject of informal agreements new prominence.1 Chinese and American negotiators had spent the week working out the terms of an agreement to resolve the trade war between the two nations. They had labelled the agreement a Memorandum of Understanding, or ‘MOU’. Upon hearing this news, President Trump—seated across from the top Chinese negotiator as well as his own Trade Representative, Ambassador Robert Lighthizer—announced that ‘I don’t like MOUs because they don’t mean anything’.2 Ambassador Lighthizer attempted to persuade the President otherwise, arguing that ‘An MOU is a binding agreement between two people. It’s detailed. It covers everything in great detail’.3 To which the President responded, ‘By the way, I disagree’.4

Who was right? Is an MOU the same thing as a legally binding international agreement? How can one tell the difference between a legally binding and a non-binding agreement? If MOUs are legally non-binding, are they worth the paper on which they are written? Do they change how States behave?5 Why would negotiators use them?

Questions like these have fascinated international lawyers and scholars for decades,6 although they have received increasing attention since the turn of the century. The end of the Cold War and the spurt of institution-building in the 1990s caused scholars to focus on the role of legalization in the international system.7 In a decade that saw treaties like the Kyoto Protocol (as well as treaties establishing institutions such as the International Criminal Court and the World Trade Organization), some struggled to understand why governments do not use formal treaties to codify the outcomes of all of their negotiations. Yet States continued to rely on multilateral non-binding agreements in a wide range of areas, ranging (p. 60) from banking regulation8 to food safety9 to arms control.10 States exhibit similar diversity in concluding a range of non-binding bilateral agreements on matters such as competition law,11 diplomatic relations,12 and even wartime relationships.13 At the same time, scholars also began to notice that actors other than States participate in international norm-making in ways that bear some similarities to norm-making among States.14 These alternatives to formal treaty-making produced a wide range of instruments with many different kinds of labels, such as MOUs, soft law, and political commitments. For their part, scholars sought to understand both the kinds of tradeoffs that States or non-State actors face when deciding between formal treaty instruments and the range of informal instruments available to them, as well as how treaties and informal international agreements differ in their influence on State behaviour.

This chapter reviews the existing literature and practice on informal lawmaking. Section I begins by reviewing basic definitional issues, as well as three contexts in which international lawyers confront non-binding agreements. Section II canvasses the existing literature on why States and non-State actors turn to informal agreements. Section III turns to normative matters, arguing that non-binding agreements are not significantly less efficacious or accountable than binding international agreements. The chapter ends with a call for States to continue to develop guidelines for informal lawmaking—work already begun by the Inter-American Juridical Committee (IAJC).

I.  What is Informal Lawmaking?

Informal agreements are usually defined in opposition to formal agreements, that is, treaties.15 As such, there is an inherent fuzziness to the category.16 Pauwelyn, for instance, argues that informality can come in terms of outputs, processes, or actors involved.17 Abbott and Snidal describe international agreements as ‘soft’ along three dimensions: (i) legality, which corresponds to bindingness; (ii) the vagueness of the substantive obligations; and (iii) the degree of delegation to third parties, such as courts.18 Along with these conceptual differences comes a cacophony of terminology. Aside from ‘informal’ law and (p. 61) ‘soft’ law, authors also commonly use the term ‘political commitments’ to describe agreements that are not binding under international law.19

Given the lack of consensus about either terminology or definitions, authors must be careful to define their terms. This chapter therefore focuses on non-binding instruments, and uses the terms ‘informal’ and ‘soft’ interchangeably with, and essentially to mean, non-binding.20 This definition tracks the most common and influential use of these terms,21 as well as the common core of more expansive definitions. The bindingness of an international instrument under international law has also been a particular preoccupation of international lawyers and international institutions.22 Therefore, while there is no objectively correct use of either terms or definitions, a discussion of binding versus non-binding agreements may most helpfully shed light on key issues of contemporary State practice.

At the outset, we must ask why binding and non-binding agreements should be considered alongside each other.23 Precisely because non-binding agreements are not legally binding under international law, a number of scholars have objected that they are not, in fact, part of international law at all.24 These authors often prefer the term ‘political commitment’ or ‘pledge’ to better reflect the binary nature of legality.25 This objection is partially semantic, focusing on the use of terms like ‘informal law’ and ‘soft law’ to describe non-binding instruments. The objection has a substantive component as well, though. Commentators have worried that conflating the two categories risks undermining the normative pull of binding international law,26 or expanding the rule of lawyers to issues better left to the realm of politics.27

Binding and non-binding instruments are, however, closely related in three aspects of international practice—negotiation, legal effect, and identification—and therefore merit joint consideration.

A.  Negotiation

The relationship between binding and non-binding instruments is most obvious at the stage of legal practice that receives disproportionately small coverage in academic (p. 62) commentary: the negotiation of international instruments. Government officials frequently negotiate legal bindingness as a component of a larger agreement. Indeed, after President Trump objected to a trade agreement with China taking the form of a non-binding MOU, reporters overheard Ambassador Lighthizer lean over to Chinese Vice Premier Liu He to seek the latter’s assent to change the form of the agreement.28 Similarly, during the negotiations of the Joint Comprehensive Plan of Action (JCPOA) regarding Iran’s nuclear programme, the legal status of the agreement appeared up in the air. US Secretary of State John Kerry emphasized that the administration wished to pursue a non-binding agreement that focused on enforcement,29 while the Iranian government publicly argued that any agreement would be binding under international law.30 Looking further back, the United States and the United Kingdom (as well as other Commonwealth countries) have disagreed on whether agreements captioned as MOUs can be binding. The United States’ view is that MOUs can be legally binding (hence Ambassador Lighthizer’s statement in the Oval Office). The United Kingdom’s view, on the other hand, has historically been that they are not. In 1991, following a dispute about the legal status of MOUs between the two countries regarding military cooperation, the US Defense Department directed its negotiators to avoid any non-binding language, including the use of the term MOU, in agreements with the United Kingdom, Canada, and Australia.31 These examples attest to the fact that the bindingness of an agreement can be traded off against other features of an agreement, such as the language or precision of obligations or the existence of dispute settlement provisions.32 For this reason, non-binding agreements are appropriately analysed alongside binding international agreements. They are two possible outcomes of the same process of negotiating inter-State pacts.

Informal lawmaking might also include instruments or commitments negotiated by actors that do not traditionally participate in formal lawmaking. Some commentators, such as Pauwelyn, include in this group government officials that are not normally engaged in diplomatic work.33 This view finds support in the practice of countries like France that do not recognize as treaties agreements concluded by administrative agencies.34 Governance through networks of domestic regulators, such as the Basel Committee on Banking Supervision, are also sometimes considered informal.35

Distinctions based on the role of government officials who are not diplomats can be overstated, however. Delegations negotiating formal treaties are regularly staffed by officials from interested ministries and departments, regardless of whether those officials would ordinarily be considered diplomats. Indeed, many countries, such as the United States, (p. 63) recognize agreements between domestic agencies as binding commitments when the other requirements for such status are satisfied. Nor are the procedures used in bodies consisting of domestic agencies necessarily less formal than those used in diplomatic conferences.36 On the other side of the coin, diplomats often participate in relatively unstructured negotiations among governments. Indeed, precisely because technological changes have made it easier for a wider range of government officials to communicate with their foreign counterparts, distinguishing formal from informal lawmaking on the basis of who the State chooses to represent it may make little sense.

The emergence of instruments or commitments negotiated, promulgated, or joined by non-State actors, however, does mark a notable departure from formal lawmaking.37 To be sure, non-State actors frequently participate in and influence the contours of treaties.38 But under traditional rules of international law, such non-State actors do not have the capacity to create international law at all. This absence has not stopped non-State organizations from creating a wide range of international instruments. The climate change area has a particularly rich array of such instruments, including the Greenhouse Gas Protocol, created by the World Resources Institute and the World Business Council for Sustainable Development, and rules issued by the Climate Disclosure Standards Board, a consortium of business and environmental groups.39

Such ‘transnational’ governance is an important alternative—as well as complement—to treaties. Private transnational governance can mitigate or render irrelevant lax regulatory positions taken by States. Again, in the climate context, private transnational governance has provided an important alternative to State-led initiatives that have addressed climate change haltingly at best. The regulation of cyberspace also accords an important role to non-binding instruments, such as the Paris Call for Trust and Security in Cyberspace.40 Such instruments can more easily include the multiple transnational stakeholders relevant to governing cyberspace. States can also encourage private transnational governance, or at least choose not to pre-empt it.41 For instance, the International Accounting Standards Board (IASB) is a private non-profit organization that develops international accounting standards. The majority of its funding, however, comes from governmental contributions.42 Likewise, nations remain free to adopt any accounting standards they like, meaning they can pre-empt or ignore the IASB’s work. In this way, transnational governance can fill gaps left by State-based regulation, can push States to tackle new regulatory problems, or can solve regulatory problems that States prefer not to deal with themselves.

(p. 64) B.  Legal effects

Binding and non-binding agreements can also have similar effects, although there are important differences. The similarities are clearest if one differentiates between purely political commitments and non-binding commitments intended to have some legal effect (ie soft law). Some line must separate such informal lawmaking from acts that have no legal significance whatsoever. Pauwelyn, Wessel, and Wouters distinguish informal lawmaking from non-lawmaking based on whether the instrument constrains a State’s behaviour in some way.43 Under their view, a political commitment could seemingly count as informal lawmaking if it limits the freedom a State enjoys going forward. In this way, Pauwelyn and other’s definition seems to treat any instrument that has behavioural effects as ‘legal’.

This view is problematic, however, because it equates effectiveness—the degree to which an agreement constrains or changes State behaviour44—with legality. On the one hand, many formal international agreements (that is, treaties) may have little to no effect on State behaviour. A wealth of literature examines so-called ‘shallow’ agreements that require States to do little to change their behaviour.45 Some studies on compliance have argued that certain treaties, such as human rights agreements, appear not to prompt States to make efforts to comply with the legal standards established by the agreements.46 On the other hand, many kinds of joint action that do effectively constrain or change State behaviour do not seem legal at all. For instance, joint statements by heads of State at a press conference, or the decision to hold or cancel a summit, might change State behaviour going forward, yet seem to have little indicia of legality.47

For this reason, other authors take a different view. I have argued that soft law agreements are those that, while non-binding under international law, have consequences in some legal systems.48 These legal consequences distinguish soft law agreements from entirely non-legal acts.49 For instance, the Basel Accords, which establish international banking standards, are non-binding as a matter of international law, but they are the basis for binding domestic regulations of the banking industry.50 Soft law agreements might also have legal consequences in international law. This can occur in a number of different ways. In dispute (p. 65) resolution, for instance, international arbitrators or judges may rely on soft law instruments to resolve disputes in a binding fashion.51 As Kaufmann-Kohler explains, international investment tribunals regularly rely on non-binding rules—such as the International Bar Association’s Rules of Evidence and Guidelines on Conflicts of Interest—to establish procedures through which to adjudicate disputes in a binding fashion.52 And Schill documents the use by investment tribunals of non-binding instruments to fill gaps in international investment agreements.53 Non-binding obligations may expand on or interpret a binding legal obligation. Binding legal instruments, for instance, often contain non-binding obligations that may clarify or provide context for binding obligations in the same instrument. For instance, the UN Framework Convention on Climate Change contains a non-binding target of reducing greenhouse gas emissions to 1990 levels—an obligation relevant to understanding how States should approach the binding commitments established by the treaty. Similarly, decisions of treaty bodies, such as a Conference of the Parties (COP), are often non-binding but can supplement or expound on binding obligations.54 By way of example, the Paris Agreement on Climate Change, like many agreements, empowers its COP to ‘make, within its mandate, the decisions necessary to promote its effective implementation’.55 These decisions could shape how member States understand the obligations of the Paris Agreement itself, thereby giving them legal effect.

Of course, non-binding agreements are not subject to the law of treaties and the law of State responsibility would not appear to apply to them.56 And, as I explain below, although reputational sanctions attach to non-conformance with both binding and soft law instruments, they are weaker for the latter. But soft law agreements do often have enforcement mechanisms that give them meaningful bite. The JCPOA, for instance, includes a verification programme and the possibility of sanctions if Iran breached the terms of the agreement.57 The Organisation for Economic Co-operation and Development’s (OECD’s) non-binding Guidelines for Multilateral Enterprises can be interpreted prospectively by the Committee on International Investment and Multinational Enterprises through a dispute-resolution-like process.58 The mere fact of non-bindingness thus does not preclude States from building in enforcement mechanisms similar to those found in binding instruments and designed to create legal effects.

(p. 66) C.  Identification

Lastly, faced with existing international instruments, international lawyers must also take on the difficult task of determining whether an agreement is legally binding or not. Such a determination can affect the application of the law of treaties, the law of State responsibility, the jurisdiction of international tribunals, and the application of domestic approval processes associated with different kinds of international commitments. Such a task is, of course, a necessarily comparative exercise. What kinds of evidence indicate an instrument is governed by international law so as to qualify as a treaty? Inter-governmental organizations like the IAJC,59 as well as scholars,60 have bent themselves to the task of formulating an approach to identifying binding versus non-binding commitments. International tribunals have also developed a jurisprudence aimed at distinguishing binding and non-binding commitments. In Qatar v Bahrain61 and Pulp Mills,62 for instance, the International Court of Justice (ICJ) found that the minutes of meetings between officials from the governments in question constituted binding agreements. In these cases, the ICJ applied an objective test, looking to the text of the instruments in question to determine whether the parties had made commitments to each other. Once it determined that they had made such commitments, the ICJ did not ask whether the parties intended those commitments to be legally binding, instead equating the existence of commitments with the existence of legal commitments.63 In other cases, such as the Iron Rhine and the South China Seas Arbitration, tribunals have considered the intent of the parties, asking whether the parties intended to make binding legal commitments or merely non-binding (or ‘political’ or ‘aspirational’) commitments.64 I return to this subject in subsequent parts of this chapter.

II.  Why Do States Make Law Informally?

Informal lawmaking has a number of potential drawbacks. Precisely because it is informal, it may exert less of a compliance pull on States.65 Under a ‘broken windows’ theory of international lawmaking, the increased use of informality could reduce the normative pull of formal international law.66 Informality might also be used as a way to short circuit the (p. 67) ordinary mechanisms of accountability that comes with formal lawmaking.67 Why, then, do States rely on informal lawmaking?

The traditional answer to this question has been flexibility. As Lipson describes informal agreements, ‘[t]hey are willows, not oaks’.68 On Lipson’s account, informal agreements are thought to be more flexible for several reasons. First, they are, it is claimed, more easily amended.69 Second, they need not be as specific as formal agreements, which means they make fewer informational demands on parties during negotiations.70 Third, informal agreements tend to be less prominent and subject to fewer legal formalities in their approval, reducing the costs of entering into informal agreements.71 Taken together, the lack of precision and reduced costs of entering into and amending the agreement make it easier for States to walk away from their commitments if circumstances change.

This flexibility rationale has an intuitive appeal to it, but on closer inspection certain aspects of the flexibility rationale do not hold up, while others require greater specification. More specifically, we might ask: how do informal agreements create flexibility, and when is that flexibility valuable?

A.  Reduced penalties for violation

Informal agreements’ flexibility flows primarily from reduced penalties associated with violating informal agreements. These reduced penalties have both a formal legal dimension and a practical one. As noted above, informal agreements are not subject to the law of State responsibility.72 One does not speak of a ‘breach’ of a non-binding agreement in the same doctrinal sense that one speaks of a breach of a treaty.73 Indeed, the Vienna Convention on the Law of Treaties (VCLT) would simply not apply to non-binding agreements.74 In this sense, when States conclude non-binding agreements they are making a (non)choice of law. Treaties are agreements between States governed by international law. States can also conclude contracts governed by some nation’s domestic laws.75 Or they can conclude non-binding agreements, which opt out of both of these sets of laws.

Beyond the absence of sanctions associated with international law as such, non-binding agreements also often carry fewer real-world penalties for violation. In part, this can come from the absence of rules authorizing retaliation—or countermeasures—for violation. Although a State could consent in a soft law instrument to the withdrawal of a benefit if it fails to conform to the soft law instrument,76 background institutions and rules providing these mechanisms do not exist for non-binding instruments. But it also flows from the (p. 68) reduced reputational sanctions for violation. Binding and non-binding agreements can be compared to contracts and pledges in the domestic system.77 A contract creates firm expectations in the mind of a counterparty. A pledge, by contrast, creates considerably weaker expectations. Put in everyday terms, think of two different responses to an invitation to a party. ‘I will be there’ creates a firm expectation that you will attend the event. ‘I will try to attend’ creates a much weaker expectation.

This distinction matters when expectations are disappointed. Just as people have a reputation for abiding by their word, so too do States. Failing to live up to the expectations one has created hurts that reputation. A good reputation for compliance with one’s commitments is valuable to a State because it allows other States to more effectively rely on the State’s promises.78 This reliance makes international agreements more valuable, especially in the international system in which no central enforcement of legal obligations exists, and may help a State extract more valuable concessions from its partners. As a State’s reputation gets worse, it loses the ability to extract those valuable promises from other States. Other States cannot rely as accurately on its word, and so will demand more, either in terms of substantive concessions or guarantees, before they are willing to enter into an agreement.

Entering into a binding agreement thus represents a risk for States. They are placing their reputation at risk if they later breach the agreement. Non-binding agreements involve a less serious pledge of a State’s reputation. If a State fails to honour a non-binding commitment, other States will not update their view of the ‘violating’ State in the same way, precisely because they never had the same kind of firm expectation that the State would behave in a particular way.

Non-binding agreements thus create flexibility by allowing States to violate their commitments at reduced cost. Significantly, though, this is true for both parties to an agreement. It is thus not obvious that States should want this kind of flexibility. The benefits from one’s own flexibility, in other words, come with the costs of other States’ flexibility. Only in some circumstances will this reciprocal flexibility make sense. These circumstances can arise either because of compliance concerns or as a way to ease negotiations. In economic terms, compliance flexibility will make sense when the marginal costs from higher sanctions are more than the marginal benefits from the sanctions. This, in turn, will occur when States face a high risk of unavoidable violations. Reputational costs, after all, are not like monetary damages. What one party loses represents a joint cost to the parties. States may rationally seek to avoid these joint costs. This ‘loss avoidance’ rationale explains why non-binding commitments are more prevalent than it might seem at first blush.

The reduced costs associated with violating a non-binding agreement also have implications for amending such agreements. States, after all, are not only the subjects of international law but also its authors. They can use non-compliance with international agreements as a renegotiation tactic.79 Non-compliance creates costs for other States. That is true for binding commitments, as well as non-binding commitments. A failure by one country to adhere to the Basel Accords, for instance, could implicate the stability of financial institutions in other countries. In some circumstances, these costs can give States the leverage to renegotiate existing norms.

(p. 69) Non-compliance can be used to renegotiate treaties and informal agreements alike. President Trump, for instance, likely put the United States in violation of its trade obligations by imposing tariffs on a wide range of Chinese products during 2018–19.80 Those tariffs led to negotiations with China that could, in principle, change trade law. But lowering the costs of non-compliance by making an agreement non-binding makes this renegotiation tactic easier. Non-compliance becomes cheaper, so States are more willing to engage in it as a vehicle to change the law.

Sometimes States may choose soft law precisely because they wish to enable this kind of non-compliance. In such situations, the choice of soft law may enable a powerful State to act as a leader in updating norms.81 For instance, the non-binding nature of the Nuclear Suppliers Group Guidelines—rules on trade in technology that can be used to make nuclear weapons—has allowed the United States to push the boundaries of the rules in order to normalize civilian nuclear trade with India. Such trade was not permitted under the Guidelines because India was not a member of the Nuclear Nonproliferation Treaty. Using non-binding guidelines to coordinate State policies on nuclear trade, though, allowed the United States to act as a leader in revising its own rules, which ultimately caused the Nuclear Suppliers Group to update multilateral rules.82

Similarly, Mark Pollack and Greg Shaffer have argued that States may choose to use soft law precisely because it conflicts with a pre-existing hard law agreement.83 Soft law makes it easier for States to strategically establish conflicting rules as a means of changing the law—a process they refer to as hard law and soft law operating as antagonists. These conflicts work through essentially the same mechanism as the threat of violation. By reducing the penalty for violation, soft law agreements make it easier for States to create conflict that States then resolve by changing hard law agreements through amendment or interpretation. For instance, Iceland and Norway spearheaded the formation of the North Atlantic Marine Mammal Commission (NAMMCO). They did so because they were unhappy with the binding moratorium on commercial whaling imposed by the International Whaling Commission (IWC).84 But NAMMCO does not promulgate binding rules that conflict with the moratorium. Instead, NAMMCO publishes non-binding recommendations and guidelines on the harvesting of whales and small cetaceans.85 These guidelines aim to show that responsible rules can permit sustainable harvesting of whales without the IWC’s moratorium. More generally, the establishment of a rival regime like NAMMCO is easier for States precisely because the existence of non-binding agreements lowers the costs associated with conflict among legal rules.

(p. 70) B.  Procedural flexibility

Non-binding agreements are also regularly thought to create more flexibility as a procedural matter. If an agreement is non-binding, the thinking goes, States do not have to go through the same formalities in order to obtain approval for the agreement.86 While this difference in formalities is real in some circumstances, the difference does not drive behaviour in the way the literature often supposes.

The claim that non-binding agreements create procedural flexibility usually focuses on the different domestic procedures that governments use to approve international agreements. The US Constitution, for instance, requires that the Senate—the upper chamber of the legislature—gives advice and consent prior to the ratification of treaties.87 The Senate has been the graveyard for treaties for years. It famously failed to approve the Treaty of Versailles at the end of the First World War, and foundational treaties like the VCLT and the 1982 UN Law of the Sea Convention have languished for decades. More recently, President Clinton felt he could not win approval of the Rome Statute of the International Criminal Court or the Kyoto Protocol, prompting his successor—President Bush—to ‘unsign’ the treaties.88

Legislative approval is not the only roadblock for binding agreements. In some countries, government agencies other than the foreign ministry lack the authority to conclude binding agreements. For instance, France does not recognize arrangements administratifs as binding agreements under international law.89 Similarly, a survey of American States done by the IAJC, acting through Special Rapporteur Duncan Hollis, revealed variation both in whether agencies other than the foreign ministry could conclude binding agreements, as well as the domestic procedures States use to conclude such agreements.90 In countries in which agencies lack the capacity as a matter of domestic law to enter into binding agreements, the willingness and resource capacity of the foreign ministry—which may have little experience or interest in the subject matter of a particular agreement—can constrain a State’s ability to enter into a binding agreement. In such situations, a non-binding agreement may be an attractive—and in some cases unregulated—alternative.

Although domestic hurdles to creating binding agreements can sometimes push States to use non-binding agreements, one must be careful not to overstate the role domestic procedural flexibility plays in choosing non-binding agreements over a binding alternative. In the United States, for instance, the impediment posed by the US Senate does not meaningfully constrain the executive branch from entering into binding international agreements. Under US law, the executive branch can make binding commitments in a very wide range of circumstances without Senate approval. Examples include major trade agreements, such as the North American Free Trade Agreement and the Marrakesh Agreement Establishing (p. 71) the World Trade Organization, and major environmental agreements, like the Minamata Convention on Mercury and the Paris Agreement on Climate Change.91 Some of these agreements are approved by the full Congress, while others—like the Minamata Convention and the Paris Agreement—are not submitted for legislative approval at all. This flexibility in domestic procedures means that the US executive branch can avoid cumbersome legislative approval processes while still securing a binding international commitment.92

Tighter regulation of non-binding agreements can also diminish the procedural flexibility such agreements afford, making them less attractive. The French government appears to discourage the use of non-binding agreements among administrative agencies.93 States like Columbia, Mexico, and Peru often have their Foreign Ministry participate in the formation of non-binding international agreements. Meanwhile, Panama refers most non-binding agreements to its legislature for approval, meaning that concluding such an agreement does not allow the government to evade legislative oversight.94

The procedural flexibility of non-binding agreements plays a more important role at the international level. The creation of binding international law is bedevilled by the problem of State consent.95 While customary international law does not require a State to consent before being bound,96 treaties do. As a consequence, getting agreement among States on the terms of cooperation can be quite difficult. States can use various ways to pressure each other to join agreements,97 but they have little legal recourse if a State continues to hold out. As a result, it is often said that States face a tradeoff between the breadth and depth of international agreements.98 States can make the substantive obligations of an agreement weaker in order to attract more participants, or they can have a tough agreement that binds only a smaller number of States.

Non-binding instruments offer a way to soften this breadth–depth tradeoff. They do so in several ways. First, soft law agreements can complement hard law agreements.99 A number of scholars, for instance, argue that informal or soft law instruments can help build consensus among States.100 On this account, informal lawmaking can allow States to develop shared ideas, build trust, and ultimately form non-binding rules that can harden into (p. 72) binding rules at some future date.101 Much of this scholarship views soft law as a second-best option, relative to hard law,102 except insofar as soft law makes an eventual hard law agreement easier.103

Others, however, argue that soft law can be a valuable alternative to hard law in its own right.104 By allowing States to reach agreement in situations in which they would not otherwise be able to, informal lawmaking expands the range of issues over which States can cooperate.105 In this sense, informal lawmaking complements formal lawmaking on the micro scale when it allows States to use soft law to develop interpretations of binding agreements. But it also complements hard law on a macro scale by giving States more tools to use in their pursuit of cooperative endeavours. This tool can soften the breadth–depth tradeoff that formal lawmaking often faces.

Informal lawmaking can also soften the breadth–depth tradeoff by circumventing the consent requirement. International institutions often have the power to make determinations about the interpretation or implementation of international agreements. In certain circumstances, these determinations are binding. For instance, the IWC, discussed earlier, has the power to set legally binding catch limits for whales.106 Likewise, the decisions of the ICJ or the WTO’s Dispute Settlement Body are binding on the parties to the dispute.

Beyond those fairly limited circumstances, though, the decisions of international institutions tend to be non-binding. As a result, most of the decisions of COPs, and indeed of international tribunals (at least with respect to parties not before the tribunal), are non-binding. For instance, the Kyoto Protocol to the UN Framework Convention on Climate Change contemplated the creation of an enforcement system. That enforcement system, the treaty suggested, could be binding or non-binding.107 If the COP adopted a system ‘entailing binding consequences’, that system would have to be approved by the members as an amendment to the treaty, meaning that each State would have to consent before it could be bound. Instead, the COP established the Kyoto compliance mechanism by a simple decision of the parties.108 That procedure allowed the adoption of a compliance system that could result in member States losing rights under the treaty (in this case, emissions rights) without each State individually having to approve the amendment.

Decisions and opinions of bodies like the UN Human Rights Committee, an expert body convened under the auspices of the International Covenant on Civil and Political Rights, have a similarly non-binding effect. The Committee, and others like it under a variety of human rights agreements, issue interpretations of the treaties that create them. (p. 73) These interpretations are non-binding, but they shape the expectations States have as to what counts as compliance with the underlying binding obligation. In this way, expert committees operate as a backdoor vehicle around the consent requirement. A treaty body may be able to adopt interpretations of a treaty that the parties themselves would not agree to through the ordinary treaty-making process. Nevertheless, if States’ views about what constitutes compliance are informed by these interpretations, then they have some legal significance.

The same is true of non-binding decisions from international tribunals. Of course, advisory opinions from the ICJ are not binding. Moreover, as noted above, decisions of most tribunals are not binding with respect to parties not before the tribunal. Despite this fact, many States may not be in a position to resist the development of a non-binding jurisprudence that constitutes a gloss on their binding legal obligations. Indeed, part of the current crisis at the WTO—in which the United States, under both Presidents Obama and Trump, have blocked the appointment of new members of the WTO’s Appellate Body—stems from the United States’ belief that the Appellate Body had made decisions creating obligations to which it had not agreed and had applied a system of effectively binding precedent despite the lack of authorization to do so.109 In principle, the objection of a powerful State like the United States can check the growth and influence of this form of non-binding law. However, even a State as powerful as the United States has not been able to dictate Appellate Body reform, and weaker States are in an even less enviable position in that regard. In short, despite the absence of formally binding effect, the WTO’s Appellate Body has managed to significantly change how States understand their WTO obligations.

III.  The Challenge of Informal Lawmaking

Informal lawmaking faces a number of potential objections. Here I want to consider two: (i) the claim that informal lawmaking undermines the efficacy of formal lawmaking; and (ii) the claim that informal lawmaking is less accountable than formal lawmaking.

A.  The efficacy of international law

The idea that informal international law undermines formal international law rests on the claim that the existence of informal lawmaking will either (i) cause States to conclude fewer binding agreements, thereby reducing their ability to coordinate behaviour effectively, or (ii) muddy the waters with respect to what counts as a binding agreement, thereby again reducing the incentive to comply with binding agreements.

On close inspection, both of these arguments fall short. As discussed above, States are often reluctant to agree to formally binding obligations. The availability of informal obligations allows States to coordinate their behaviour in a range of circumstances in which they otherwise would not. To be sure, if given a choice between formalized cooperation and no (p. 74) cooperation, in some situations States might choose formal cooperation. But these circumstances are likely to be less in number and of less consequence than situations in which, in a world without informal cooperation, States simply forego cooperation entirely. Such a choice, after all, would reflect the status quo.

More importantly, though, the idea that removing informal cooperation would boost the efficacy of formal law overlooks the fact that States do not bargain only on the dimension of legality. When negotiating international agreements, States have a host of tools that they can change in order to reach agreement. They can, for instance, make obligations vaguer or more precise; they can cover a topic comprehensively or leave issues unregulated or the subject of future negotiations; and they can include dispute resolution or some other form of delegated monitoring.110 Removing the legality of the agreement would simply mean that more tradeoffs are made on the other issues. In practice, that would mean that international legal obligations would often require less of States than a comparable informal agreement and would be less likely to include dispute resolution. Put differently, informal lawmaking can promote more meaningful and deeper cooperative terms precisely because it does not force States to make concessions on substance in order to reach agreement.

Equally importantly, informal lawmaking allows States to update the law when circumstances have changed. The status quo bias of formal treaty-making, in which each State must agree before it can be bound by amendments, means that a world with only formal lawmaking risks the irrelevance of that law.111 Instead, informal lawmaking acts as grease to the international legal system, making its overall functioning more fluid.

This fluidity is especially valuable in situations in which States are uncertain about the future State of the world, such as in the presence of scientific uncertainty, and States have common interests.112 Where States have common interests, choosing non-binding agreements does not necessarily carry with it a compliance cost, because States wish to comply with international norms anyway. But where States are uncertain about the future State of the world, they may wish to amend rules in response to new information. Formally, of course, a single State can prevent such an amendment, even when the amendment is in the interest of most States. Non-binding agreements allow individual States to violate the existing agreement as a way to pressure holdouts to agree to amendments. In effect, it reduces the veto power of holdouts when States have a common interest in amendment. In the absence of common interests, on the other hand, States may—even in the presence of uncertainty—prefer a binding agreement that is more difficult to amend. The relative difficulty of using violation to induce change prevents States from being coerced into undesirable amendments.

The existence of informal international law, and the related uncertainty about whether an obligation is binding might, in principle undermine the efficacy of formal international law. After all, the fact that not all agreements are legally binding will surely increase doubt about a range of instruments, some of which would not necessarily have had their legality in doubt. Indeed, States may strategically inject doubt on the subject.

(p. 75) This concern, although a real one, is not as serious as it might first seem. After all, international lawyers regularly are called upon to determine the legality of an instrument or obligation. Indeed, identifying the legal bindingness of an instrument is one of the core tasks of foreign ministry lawyers. The fact that lawyers must make judgment calls, and that the existence of informal lawmaking may increase the number of circumstances in which they must do so, is not an objection to informal lawmaking writ large (although as I discuss below, some harmonization in approaches to identifying non-binding agreements could also reduce the role of lawyers in making difficult interpretive judgments). And, in cases such as Qatar v Bahrain,113 Pulp Mills,114 and the Iron Rhine,115 international courts and tribunals have had little reluctance identifying which agreements are binding and which are not. Outcomes may, of course, differ. In part, these differences may reflect differences between an objective approach (Qatar v Bahrain, Pulp Mills) versus an intent-based standard (Iron Rhine, South China Sea). But both objective and intent-based standards will accord weight to the instrument itself, meaning that the kinds of evidence at issue may not be as starkly different under these two standards. Tribunals may also differ on whether they equate commitments among states with treaties,116 or whether they allow greater space for states to make non-legally binding commitments. Moreover, international instruments often contain non-binding obligations within them. For instance, Article 4.2 of the UN Framework Convention on Climate Change—surely a binding agreement—contains a non-binding commitment on governments to return their greenhouse gas emissions to 1990 levels.117 At least to the extent that informal lawmaking equates to non-binding commitments, then, informal lawmaking is inescapable.

B.  The accountability of informal lawmaking

The questions of whether informal lawmaking—understood, again, to refer to non-binding instruments—is less accountable presents a somewhat more difficult challenge. Clearly, governments sometimes choose the form of international agreements to avoid domestic oversight, especially through legislatures. At the same time, though, informal international lawmaking must be compared to other forms of lawmaking, as well as realistic expectations for what accountability means. Viewed in this light, non-binding instruments are not, as a category, less accountable than more formal means of lawmaking.

At the outset, one must define what one means by accountability. Pauwelyn, for instance, focuses on the role of oversight mechanisms, transparency, and participation, while also highlighting the important temporal relationship between the lawmaking act and the opportunity for accountability to take place.118 Grant and Keohane have broken the (p. 76) concept down even further, distilling seven different types of accountability: hierarchical, supervisory, fiscal, legal, market, peer, and reputational.119 These different methods of accountability typically involve different processes and different stakeholders. International institutions that depend on the financial support of governments, for example, are accountable to those governments that provide the bulk of such support.

Once such broad views of accountability have been adopted, evaluating informal lawmaking as a category becomes extremely difficult. The extent to which particular rules or institutions are ‘accountable’ will vary based on the characteristics of the institution or rules, as well as the issue area, more than on the formality of the rules. The non-binding (in the view of the United States) JCPOA regarding Iran’s nuclear programme received considerably more attention, and was therefore considerably more ‘accountable’ to public will than are run-of-the-mill extradition treaties that are clearly binding.

Moreover, the accountability of informal lawmaking should be compared to the alternative. Formal international law and international legal institutions are often perceived to suffer from a democracy deficit of their own.120 This view tends to privilege a notion of accountability to the people from whom national governments are thought to derive the right to rule. On this account, international law suffers from an accountability problem because it is further removed from the people, and in some versions because people of one nation are subject to rules negotiated by governments representing other peoples and nations.

These criticisms do not flow from the characteristics of international law, but rather from its supranational nature. As such, these criticisms should apply equally to formal and informal international law. Indeed, on this view, informal international law may actually be more accountable. For instance, proponents of a strong view of national sovereignty sometimes object to international institutions because they constrain democratic choice at the national level. International law, in other words, becomes a way to tie the hands of future elected representatives.121 As discussed above, however, informal agreements are more easily broken. They therefore remain more accountable to the people because later governments are less constrained. A group of US senators relied on this very feature of the JCPOA in arguing that a future US president would be able to undo the deal.122

Even those with a more favourable view of international law might worry about informal lawmaking due to the different domestic procedures associated with its creation.123 The lack of legislative review, in particular, could lead to less democratic monitoring of the kinds of informal commitments into which States enter. In this sense, the rise of non-binding agreements internationally would reinforce an ongoing trend in democratic countries like the United States or the EU in which power is increasingly concentrated in an executive that (p. 77) may be relatively less sensitive to the multitude of interests within the country than a legislature would be.124

This concern is a real one. As the IAJC’s ongoing study of non-binding commitments demonstrates, many countries do not have established procedures governing the formation of non-binding commitments.125 As such, a whole category of international commitments may evade legislative oversight. Once again, though, this concern is not unique to informal commitments. Formal international commitments frequently do not receive legislative review, either. Many countries are, of course, not democracies and thus no international commitments, formal or not, receive meaningful legislative review. In Westminster-style governments, the executive rises from the legislature and thus the latter can usually be relied upon to approve agreements negotiated by the former, often with minimal oversight.126 And, as noted above, in countries like the United States the executive branch enjoys some discretion in choosing how and whether formal agreements will be legislatively approved, minimizing the differences between formal and informal agreements.

Informal agreements can also be subject to legislative oversight, or other forms of scrutiny, if the obligations are domesticated. Legislatures might enact statutes codifying non-binding international commitments, as some countries have done for the Kimberly Process.127 In other cases, non-binding commitments may be domesticated through ordinary administrative rulemaking. The United States, for instance, has used administrative rulemaking—subject to legislative oversight, if not express approval—to domesticate the banking rules coming from the Basel Committee.128

Finally, the subject matter of an agreement may do more to determine the kind of oversight it receives. As the fate of the JCPOA—the agreement regarding Iran’s nuclear programme—in the United States illustrates, simply taking a high-profile agreement away from the legislature does not insulate it from oversight. The US Congress passed the Iran Nuclear Review Act, which provided for some legislative review of the JCPOA.129 Despite these procedures, the JCPOA remained highly controversial in terms of its subject matter, and the Obama Administration’s decision to frame it as a non-binding agreement not requiring congressional approval even became a political liability for the agreement. Ultimately, Donald Trump campaigned for and won the presidency with a promise to roll back the agreement if elected—the ultimate kind of accountability.

Ultimately, the accountability problem for international agreements of all stripes stems from the growing scope and detail of international cooperation. Globalization has increased the demand for international cooperation on an ever-broader range of subjects. Democratic institutions, to say nothing of voters, frequently lack the attention or resources to monitor all that is done in their name. The result is, indeed, a lack of accountability, but one that flows more from the modern world and the demands it places on governments than from the legality of the commitments chosen.

(p. 78) C.  Guidelines for informal agreements?

Having said that, much can be done both to boost the efficacy of informal agreements and to ensure that that they do not become an end-run around political accountability. The IAJC’s draft OAS Guidelines for Binding and Non-binding Agreements offers a good start in thinking about how to clarify issues around non-binding agreements. The proposed guidelines cover a number of issues. Here, I want to touch on two: (i) the development and transparency of domestic norms regarding international agreements; and (ii) a set of principles on distinguishing binding from non-binding instruments.

First, the proposed Guidelines reaffirm that States are free to organize their domestic authorities as they like. At the same time, the Guidelines also call on States to develop and publish procedures governing the creation of non-binding instruments.130 The Guidelines similarly call for governments to make public these authorities (as well as similar authorities for binding agreements), and also to track and publicize at least the most important non-binding commitments.131 The Guidelines also put forth an important call for governments to train their officials on how to handle international agreements of all stripes, because ultimately the effectiveness of any internal rules depend on those who must implement them.132

These proposals are eminently sensible. Non-binding agreements still entail consequences, both for a State’s reputation and its relationships with other States; in many cases they can entail legal consequences for the State itself or for market participants. Developing internal rules to limit this exposure can help a State avoid finding itself in a difficult situation. Publicizing those rules, as well as the resulting agreements, can avoid misunderstandings with other States.

In developing internal rules, however, States must also be sensitive to the considerations that may make non-binding agreements attractive in the first place. If governments do, in fact, use non-binding agreements in part to avoid the procedures associated with binding agreements, imposing procedures on binding agreements may have unintended consequences. The result may be that non-binding agreements become unattractive and governments or agencies develop yet a new alternative. This danger is particularly acute if the role of coordinating non-binding agreements is given to already-overtaxed foreign ministries. Those foreign ministries may have difficult relationships with agencies handling foreign affairs on specific issues (defence, trade) that would be further burdened by putting the foreign ministry in the position of supervising how other ministries interact with their counterparts. Worse, a foreign ministry without adequate resources for this new task might find itself overwhelmed, leading either to a backlog of agreements or circumvention of the new domestic procedures.

One way to ameliorate this difficulty lies in the second issue, how to distinguish non-binding and binding agreements. The Draft Guidelines do an admirable job of canvassing textual indications of an agreement’s binding or non-binding status.133 Agreement among governments on how these textual clues should be interpreted would be a step forward. But, as the Guidelines themselves note, merely agreeing on these indicators would only narrow the possibility for confusion. States may still look at conflicting evidence and reach different (p. 79) conclusions, a problem compounded by the fact that there are at least two approaches—a subjective test focused on intent, and an objective test focused on text and structure—for assessing the bindingness of an international instrument.134

For this reason, the Guidelines make an important call for States to do what domestic contracting lawyers have long been taught to do: include an explicit statement as to the legal status of an agreement.135 Doing so would have a number of salutary effects. Most obviously, it would be the single biggest step States could take to remove the confusion about an instrument’s legal effect.136 In some instances, of course, States may wish to be deliberately vague about the legal status of an instrument, perhaps because they have difficulty reaching agreement on its status, or because one side wishes to claim for domestic purposes that it is binding and the other side wishes to claim it is non-binding.137 In such situations, States would remain free to rely on existing standards for evaluating legal status. But in most situations, removing the possibility of confusion would greatly increase the efficacy of both binding and non-binding instruments.

Including what amounts to a choice-of-law clause in international agreements would also circumvent disputes about how different countries interpret evidence of bindingness.138 Resolving disagreements about how to interpret particular pieces of evidence is fraught. Governments will typically have followed one approach for a number of years and may worry that changing their approach as part of a harmonization exercise may destabilize existing agreements negotiated under prior understandings. A choice-of-law clause would allow all governments to retain their existing views with respect to their existing agreements.

Third, a choice-of-law clause would also alleviate the internal pressure on governments. While some central monitoring of both binding and non-binding agreements would remain advisable, giving government lawyers standard language they could use in their agreements with their foreign counterparts would go a long way towards standardizing approaches within and across governments. It would reduce monitoring costs internally by limiting the discretion of government lawyers to the selection of a clause (binding or not), rather than allowing lawyers to make a range of choices that might affect how one views bindingness in a subsequent interpretative effort.

The benefits of a choice-of-law clause all are improved if governments are able to agree on standard language for such clauses. Absent such agreement, the interpretation of such clauses could in principle recreate existing disputes about how to interpret the binding effect of text. Moreover, having agreed choice-of-law language internationally would also make it easier for governments to adopt internal procedures. Instead of each country generating its own language, countries could simply copy the international model. For this reason, the OAS draft Guidelines or similar projects should debate various formulations in an effort to secure widespread support among countries for specific clauses.

(p. 80) Finally, it is worth considering whether States should agree on a default rule for interpreting bindingness. Like a choice-of-law clause, a default rule would reduce reliance on interpretative methods that increase the uncertainty as to how an agreement will be viewed later in time. Klabbers has argued for such a rule and, in fact, sees one in the practice of certain international tribunals.139 His proposed default is that agreements between States (at least those evincing consent by the relevant actors) are binding unless a presumption in favour of bindingness can be rebutted. Such a presumption has much to recommend it. It preserves the normative nature of the legal system, puts States on notice they are likely to be expected to honour their commitments under the rule of pacta sunt servanda, and increases incentives to honour international commitments.

At the same time, though, the opposite rule—a presumption in favour of non-bindingness—has benefits as well. Most importantly, such a rule prevents States from unintentionally agreeing to more than they intended. Put differently, it places the burden on the State wishing to call the binding force of law into effect to make plain its intention. By so doing, leaders are not taken off-guard if informal documents such as communiqués are treated as binding,140 and lower level officials do not as easily run afoul of domestic restrictions on the creation of binding agreements. Such a default rule also preserves the benefits of informal lawmaking, benefits that have been widely recognized by States and in academic commentary. A presumption in favour of bindingness risks eliminating those benefits by treating considerably more international agreements as binding, reducing the flexibility States enjoy in bargaining over the terms of international cooperation. A presumption in favour of non-bindingness allows States to build agreements from the ground up, and ensures that more serious commitments receive more serious attention to their legal effects.


Informal international law is neither new, nor is it likely to go out of style any time soon. As States expand the range of issues on which they cooperate, they are likely to need an ever-larger array of tools. Non-binding agreements are an important part of a State’s toolkit. At the same time, though, the prevalence of such agreements also means that there are benefits from harmonizing how States approach them. International lawyers should not fear this push to regularize informal agreements. The presence of binding international agreements does not require a subordinate role for non-binding agreements. Indeed, given the pressures on States to use non-binding instruments, the efficacy of binding international law may increasingly depend on standardizing how alternative instruments are used.

Recommended Reading

  • KW Abbott and D Snidal, ‘Hard and Soft Law in International Governance’ (2000) 54 IO 421
  • A Aust, ‘The Theory and Practice of Informal International Instruments’ (1986) 35 ICLQ 787
  • Case Concerning Maritime Delimitation and Territorial Questions between Qatar and Bahrain (Qatar v Bahrain) (Jurisdiction and Admissibility) [1994] ICJ Rep 112
  • (p. 81) Case Concerning Pulp Mills on the River Uruguay (Argentina v Uruguay) (Judgment) [2010] ICJ Rep 128
  • J d’Aspremont, ‘Softness in International Law: A Self-Serving Quest for New Legal Materials’ (2008) 20 EJIL 1075
  • AT Guzman and T Meyer, ‘Soft Law’ (2010) 2 J Legal Analysis 171
  • AT Guzman and T Meyer, Goldilocks Globalism (OUP, Oxford 2020)
  • DB Hollis, Preliminary Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, 91st Regular Session, OEA/Ser.Q, CJI/doc.542/17 (6–16 August 2017)
  • DB Hollis, Second Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, 92nd Regular Session, OEA/Ser. Q, CJI/doc.553/18 (6 February 2018)
  • DB Hollis, Binding and Non-Binding Agreements: Third Report, Inter-American Juridical Committee, 93rd Regular Session, OEA/Ser.Q/, CJI/Doc.563/18 (15 July 2018)
  • DB Hollis, Fourth Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, Annex I, 94th Regular Session, OEA/Ser. Q, CJI/doc.580/19 (11 February 2019)
  • DB Hollis and JJ Newcomer, ‘ “Political” Commitments and the Constitution’ (2009) 49 VJIL 507
  • In the Arbitration regarding the Iron Rhine (‘Ijzeren Rijn’) Railway between the Kingdom of Belgium and the Kingdom of the Netherlands (Award) (24 May 2005) RIAA XXVII 35
  • In the Matter of the South China Sea Arbitration (Award on Jurisdiction) (29 July 2015) PCA Case No 2013-19
  • JJ Kirton and MJ Trebilcock (eds), Hard Choices, Soft Law: Voluntary Standards in Global Trade, Environment, and Social Governance (Routledge, London 2004)
  • J Klabbers, ‘Informal Agreements in International Law: Towards a Theoretical Framework’ (1994) 5 Finnish Ybk Int’l L 267
  • J Pauwelyn, RA Wessel, and J Wouters (eds), Informal International Lawmaking (OUP, Oxford 2012) (and chapters therein)
  • K Raustiala, ‘Form and Substance in International Agreements’ (2005) 99 AJIL 581
  • GC Shaffer and MA Pollack, ‘Hard Law v Soft Law: Alternatives, Complements and Antagonists in International Governance’ (2010) 94 Minn L Rev 706
  • D Shelton (ed), Commitment and Compliance (OUP, Oxford 2000) (and chapters therein)


1  J Jacobs and J Sink, ‘Trump’s Trade Chief Lectures His Boss and Gets Earful in Return’ Bloomberg (22 February 2019) at <https://www.bloomberg.com/news/articles/2019-02-22/trump-s-trade-chief-lectures-his-boss-and-gets-earful-in-return>.

2  Ibid.

3  Ibid.

4  Ibid.

5  In some contexts, including perhaps the context in which Ambassador Lighthizer used it, a ‘binding’ agreement might be one that spells out how States are to behave and encourages them to change their behaviour. However, I use the term ‘binding’ to refer to a legal obligation under international law, and thus use the term ‘non-binding’ to refer to anything that is not a legal obligation under international law. This distinction allows for us to think about how casting an obligation as one under international law (ie making it binding) can change behaviour. Using ‘binding’ to refer to any instrument that might change behaviour conflates the instrument with its effects, making this kind of analysis more difficult.

6  P Weil, ‘Towards Relative Normativity in International Law?’ (1983) 77 AJIL 413, 414–15 n7.

7  See eg KW Abbot and others, ‘The Concept of Legalization’ (2000) 54 IO 401.

8  See eg Basel Committee Guidelines on Banking Supervision, ‘International Convergence of Capital Measurement and Capital Standards: A Revised Framework’ (2004) at <http://www.bis.org/publ/bcbs107.htm>.

9  See eg Food and Agricultural Organization of the United Nations, ‘About Codex Alimentarius’ at <http://www.fao.org/fao-who-codexalimentarius/about-codex/en/>.

10  See eg Wassenaar Arrangement on Export Controls for Conventional Arms and Dual-Use Goods and Technologies, WA-DOC (17) PUB 001 (2017) at <https://www.wassenaar.org/app/uploads/2015/06/WA-DOC-17-PUB-001-Public-Docs-Vol-I-Founding-Documents.pdf>.

11  See eg M Matsushita, ‘International Cooperation in the Enforcement of Competition Policy’ (2002) 1 Wash U Global Stud L Rev 463.

12  DB Hollis and JJ Newcomer, ‘ “Political” Commitments and the Constitution’ (2009) 49 VJIL 507, 511 (describing the Shanghai Communiqué, a non-binding accord that normalized relations between the United States and China).

13  Ibid (describing the Atlantic Charter, which laid the foundation for the Allies’ cooperation during the Second World War).

14  See eg PS Berman, ‘Global Legal Pluralism’ (2007) 80 USC L Rev 1155.

15  J Pauwelyn, ‘Informal International Lawmaking: Framing the Concept and Research Questions’ in J Pauwelyn, RA Wessel, and J Wouters (eds), Informal International Lawmaking (OUP, Oxford 2012) 13, 15.

16  O Perez, ‘Fuzzy Law: A Theory of Quasi-Legality’ in P Glenn and L Smith (eds), Law and the New Logics (CUP, Cambridge 2016).

17  Pauwelyn (n 15) 15–20.

18  KW Abbott and D Snidal, ‘Hard and Soft Law in International Governance’ (2000) 54 IO 421.

19  See eg Hollis and Newcomer (n 12) 517; CA Bradley and JL Goldsmith, ‘Presidential Control over International Law’ (2018) 131 Harvard L Rev 1201, 1217–18.

20  More specifically, as I explain later, I define soft law to mean a subset of non-binding agreements—those that, despite being non-binding, have legal consequences of some kind. See n 48 and accompanying text.

21  A Aust, ‘The Theory and Practice of Informal International Instruments’ (1986) 35 ICLQ 787.

22  DB Hollis, Preliminary Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, 91st Regular Session, OEA/Ser.Q, CJI/doc.542/17 (6–16 August 2017) at <http://www.oas.org/en/sla/iajc/docs/INFOANUAL.CJI.2018.ENG.pdf>.

23  J Pauwelyn, ‘Is it International Law or Not, and Does it Even Matter?’ in Pauwelyn and others (n 15) 125; M Goldmann, ‘We Need to Cut Off the Head of the King: Past, Present, and Future Approaches to International Soft Law’ (2012) 25 LJIL 335.

24  See eg K Raustiala, ‘Form and Substance in International Agreements’ (2005) 99 AJIL 581, 586; J Klabbers, ‘The Redundancy of Soft Law’ (1996) 65 Nordic J Int’l L 167; Weil (n 6) 414–15.

25  Hollis and Newcomer (n 12) 520–521; Raustiala (n 24) 582. However, many authors that use the term ‘soft law’ also view legality as a binary distinction. See generally Abbott and Snidal (n 18); AT Guzman and T Meyer, ‘Soft Law’ [2010] 2 J Legal Analysis 171. For authors like Abbott and Snidal or Michael Reisman, ‘soft law’ often refers to other aspects, aside from legality, of an agreement. WM Reisman, ‘The Supervisory Jurisdiction of the International Court of Justice: International Arbitration and International Adjudication’ (1997) 258 RDC 1, 180–1

26  See generally Klabbers (n 24); Weil (n 6).

27  J d’Aspremont, ‘Softness in International Law: A Self-Serving Quest for New Legal Materials’ (2008) 20 EJIL 1075.

28  Jacobs and Sink (n 1).

29  AB Lerner, ‘State Department: Iran Deal “Non-binding” ’ Politico (11 March 2015) at <https://www.politico.com/story/2015/03/State-department-iran-deal-nonbinding-115978>.

30  JL Goldsmith and M Lederman, ‘The Case for the President’s Unilateral Authority to Conclude the Impending Iran Deal is Easy Because it Will (Likely) be a Non-binding Agreement Under International Law’ Lawfare (11 March 2015) at <https://www.lawfareblog.com/case-presidents-unilateral-authority-conclude-impending-iran-deal-easy-because-it-will-likely-be>.

31  JH McNeill, ‘International Agreements: Recent US–UK Practice Regarding the Memorandum of Understanding’ (1994) 88 AJIL 821, 825.

32  See generally Raustiala (n 24); Guzman and Meyer (n 25); Abbott and Snidal (n 18). See also Section I.B. (providing examples of non-binding instruments that contain verification and dispute resolution provisions).

33  Pauwelyn (n 15) 20.

34  Ibid.

35  Ibid; see also AM Slaughter, A New World Order (Princeton University Press, Princeton 2004) (discussing regulation by networks).

36  See A Bradford, ‘The Brussels Effect’ (2013) 107 Nw L Rev 1, 59 n 292.

37  See eg GC Shaffer and MA Pollack, ‘Hard Law v. Soft Law: Alternatives, Complements and Antagonists in International Governance’ (2010) 94 Minn L Rev 706, 719; Berman (n 14); A Boyle and C Chinkin, The Making of International Law (OUP, Oxford 2007) 41–5. I use the term ‘non-State actor’ to exclude intergovernmental organizations, which are the subjects of separate chapters in this Volume. See Chapters 5 and 6.

38  See Chapter 8 in this Volume; see also MJ Durkee, ‘The Business of Treaties’ (2016) 63 UCLA L Rev 264.

39  For a more complete description of non-State actors working on climate governance, see KW Abbott, ‘The Transnational Regime Complex for Climate Change’ (2012) 30 Env & Planning C: Government & Pol’y 571.

40  See M Finnemore and DB Hollis, ‘Beyond Naming and Shaming: Accusations and International Law in Cybersecurity’ EJIL (forthcoming 2020) 11, draft available at <https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3347958>.

41  See KW Abbott, C Kauffmann, and J Lee, The Contributions of Trans-Governmental Networks of Regulators to International Regulatory Cooperation (OECD Regulatory Policy Working Papers, No 10, 2018).

42  International Financial Reporting Standards (IFRS) Foundation, <https://www.ifrs.org/about-us/who-we-are/#funding>. The IASB is a part of, and funded by, the IFRS Foundation.

43  Pauwelyn (n 15) 16.

44  L Martin, ‘Against Compliance’ in JL Dunoff and MA Pollack, Interdisciplinary Perspectives on International Law and International Relations: The State of the Art (CUP, Cambridge 2012) 591.

45  See eg G Downs, D Rocke, and P Barsoom, ‘Is the Good News About Compliance Good News About Cooperation’ (1996) 50 IO 379.

46  O Hathaway, ‘Do Human Rights Treaties Make a Difference’ (2002) 111 Yale LJ 1935; but see BA Simmons, Mobilizing for Human Rights (CUP, Cambridge 2009). States may agree to obligations that they already are in compliance with, or which they have no intention of complying with, for a variety of reasons, such as appeasing domestic constituencies or gaining access to certain benefits, such as foreign assistance, not directly related to the agreement.

47  As Pauwelyn rightly notes, sometimes such seemingly informal interactions can be legally binding. Pauwelyn (n 15) 16. Unilateral Statements can, for instance, be legally binding, and oral agreements can also be treaties if States indicate an intent to be bound (although such treaties are not subject to the VCLT, which by its terms only applies to written agreements). What makes these instruments seem legal, however, is precisely the intent to be bound. Where that is absent, treating decidedly non-legal actions or instruments as forms of informal lawmaking conflates the legal with the political.

48  T Meyer, ‘Soft Law as Delegation’ (2009) 32 Fordham Int’l LJ 888, 891.

49  To be clear, this distinction accepts the view that bindingness itself is a binary distinction. Rather, it seeks to distinguish among different kinds of non-binding commitments based on how they interact with the legal system. This distinction among non-binding commitments is also the reason this chapter does not widely use the term ‘political commitments’, which lumps all non-binding instruments into the same category.

50  See C Brummer, Soft Law and the International Financial System (CUP, Cambridge 2015).

51  See eg AK Bjorklund and A Reinisch (eds), International Investment and Soft Law (Edward Elgar, London 2012); but see J Klabbers, ‘International Courts and Informal International Law’ in Pauwelyn and others (n 15) 219 (arguing that international tribunals treat ostensibly non-binding instruments as presumptively binding if the text evinces an intent to enter into commitments of some kind).

52  G Kaufmann-Kohler, ‘Soft Law in International Arbitration: Codification and Normativity’ (2010) 1 JIDS 283.

53  SW Schill, ‘Sources of International Investment Law: Multilateralization, Arbitral Precedent, Comparativism, Soft Law’ in J d’Aspremont and S Besson (eds), The Oxford Handbook on the Sources of International Law (OUP, Oxford 2017) 1095.

54  For more on treaty bodies, see Chapter 17; for more on treaty interpretation, including the import of subsequent practice or subsequent agreements, see Chapters 19–20.

55  Paris Agreement (adopted 12 December 2015, entered into force 4 November 2016) UN Doc FCCC/CP/2015/L.9/Rev1, Art 16.4.

56  Draft Articles on Responsibility of States for Internationally Wrongful Acts, with commentaries [2001] YBILC, vol II(2), UN Doc A/56/10, as corrected (ASR) (‘[F]or responsibility to attach to the act of the State, the conduct must constitute a breach of an international legal obligation in force for that State at that time’).

57  Although the United States maintains that the JCPOA is non-binding, the existence of these mechanisms may be part of the reason that Iran argues that the JCPOA is, in fact, binding.

58  The OECD Guidelines for Multinational Enterprises: Decision of the Council (June 2000), available at <http://www.oecd.org/daf/inv/mne/theoecdguidelinesformultinationalenterprisesdecisionofthecouncil.htm>.

59  See Hollis, Preliminary Report (n 22).

60  See eg J d’Aspremont, Formalism and the Sources of International Law (OUP, Oxford 2011).

61  Case Concerning Maritime Delimitation and Territorial Questions between Qatar and Bahrain (Qatar v Bahrain) (Jurisdiction and Admissibility) [1994] ICJ Rep 112.

62  Case Concerning Pulp Mills on the River Uruguay (Argentina v Uruguay) (Judgment) [2010] ICJ Rep 128.

63  See Klabbers (n 51) 228–9.

64  In the Arbitration regarding the Iron Rhine (‘Ijzeren Rijn’) Railway between the Kingdom of Belgium and the Kingdom of the Netherlands (Award) (24 May 2005) RIAA XXVII 35, 92 (‘A key factor in distinguishing a “non-legally binding instrument” from a treaty is the intention of the parties. To ascertain this intention, the Tribunal will, first, review the circumstances that preceded the signature of the March 2000 MoU. It will then set out the content and determine the legal significance of this particular instrument. Finally, it will summarize the circumstances that followed the signature of the March 2000 MoU, and that ultimately led to the present arbitration between the Parties’); In the Matter of the South China Sea Arbitration (Award on Jurisdiction) (29 July 2015) PCA Case No 2013-19 [243].

65  AT Guzman, How International Law Works: A Rational Choice Theory (OUP, Oxford 2008) 142–6.

66  See generally Weil (n 6) 414–15; see also I Wuerth, ‘International Law in the Post-Human Rights Era’ (2018) 96 Tex L Rev 279.

67  Pauwelyn (n 15) 22–30.

68  C Lipson, ‘Why are Some International Agreements Informal?’ (1991) 45 IO 495, 500.

69  Ibid.

70  Ibid.

71  Ibid.

72  But see J Klabbers, ‘Informal Agreements in International Law: Towards a Theoretical Framework’ (1994) 5 Finnish Ybk Int’l L 267 (arguing that informal agreements should be presumed binding unless evidence to the contrary can be shown).

73  Cf Hollis, Preliminary Report (n 22) [43]–[46].

74  Ibid.

75  Ibid.

76  If the State consented in the soft law instrument to the withdrawal of the benefit in the event of non-conformance, the withdrawal of the benefit by another State would not itself be an internationally wrongful act.

77  See generally Raustiala (n 24).

78  AT Guzman, ‘A Compliance-Based Theory of International Law’ (2002) 90 Calif L Rev 1823, 1849–50.

79  JK Cogan, ‘Noncompliance and the International Rule of Law’ (2006) 31 YJIL 189, 205.

80  J Tankersly and K Bradsher, ‘Trump Hits China with Tariffs on $200 Billion in Goods, Escalating Trade War’ New York Times (17 September 2018).

81  T Meyer, ‘Shifting Sands: Power, Uncertainty and the Form of International Legal Cooperation’ (2016) 27 EJIL 161, 168.

82  Ibid.

83  Shaffer and Pollack (n 37) 743–8.

84  See D Caron, ‘The International Whaling Commission and the North Atlantic Marine Mammal Commission: The Institutional Risks of Coercion in Consensual Structures’ (1995) 89 AJIL 154, 159–65.

85  See eg North Atlantic Marine Mammal Commission (NAMMCO), Shooting Trials on Heads of Dead Pilot Whales: Guidelines to Test the Efficiency of Rifle Ammunition Used for Hunting and Euthanasia of Small Whales (March 2006) 14–16; NAMMCO, Report of the Twentieth Meeting of the Scientific Committee (13–16 November 2013) (discussing conservation and hunting efforts).

86  See eg J Galbraith and D Zaring, ‘Soft Law as Foreign Relations Law’ (2014) 99 Cornell Law Rev 735, 749.

87  US Constitution Art II, § 2[2] (‘[The President] shall have power, by and with the Advice and Consent of the Senate, to make Treaties provided two thirds of the Senators present concur’).

88  ET Swaine, ‘Unsigning’ (2003) 55 Stanford L Rev 2061, 2061–2. States cannot technically ‘unsign’ treaties, and the United States is still listed as a signatory to these treaties. The legal significance of the US action is to clarify that the United States no longer intended to ratify the treaties, thereby removing any obligation under Art 18 VCLT and customary international law not to defeat the object and purpose of the treaties.

89  Pauwelyn (n 15) 19–20.

90  See DB Hollis, Second Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, OEA/Ser. Q, CJI/doc.553/18 (6 February 2018) [24]–[28].

91  NAFTA (Canada–Mexico–US) (signed 8, 11, 14, and 17 December 1992, entered into force 1 January 1994) [1993] 32 ILM 296 and [1993] 32 ILM 605; Marrakesh Agreement establishing the WTO (adopted 15 April 1994, entered into force 1 January 1995) 1869 UNTS 299; Minamata Convention on Mercury (adopted 10 October 2013, entered into force 16 August 2017) [2016] 55 ILM 582.

92  Such agreements might be considered ‘informal’ because they do not go through the most rigorous available approval process. See S Voigt, ‘The Economics of Informal International Law: An Empirical Assessment’ in Pauwelyn and others (n 15) 81 (treating binding international agreements not submitted to the US Senate for advice and consent as ‘informal’).

93  Pauwelyn (n 15) 19–20.

94  Hollis, Second Report (n 90) [54]–[55]; DB Hollis, Binding and Non-Binding Agreements: Third Report, Inter-American Juridical Committee, OEA/Ser.Q/, CJI/Doc.563/18 (15 July 2018) [3].

95  See eg AT Guzman, ‘Against Consent’ (2011) 52 VJIL 747, 748–9.

96  Ibid 775.

97  For instance, the Montreal Protocol contains provisions on trade sanctions for non-members. See Montreal Protocol on Substances that Deplete the Ozone Layer, as adjusted and amended (adopted 16 September 1987, entered into force 1 January 1989) 1522 UNTS 3, Art 4.

98  M Gilligan, ‘Is There a Broader-Deeper Trade-off in International Multilateral Agreements?’ (2004) 58 IO 459, 459–60.

99  See Shaffer and Pollack (n 37) 726 (discussing the literature on soft law as complements).

100  D Trubek and others, ‘ “Soft Law” “Hard Law” and EU Integration’ in G de Burca and J Scott (eds), Law and New Governance in the EU and the US (Hart Publishing, Oxford 2006); DM Trubek and LG Trubek, ‘Hard and Soft Law in the Construction of Social Europe: The Role of the Open Method of Co-ordination’ (2005) 11 Euro LJ 343.

101  See eg D Shelton, ‘Introduction: Law, Non-Law and the Problem of “Soft Law” ’ in D Shelton (ed), Commitment and Compliance (OUP, Oxford 2000) 1, 13; CM Chinkin, ‘The Challenge of Soft Law: Development and Change in International Law’ (1989) 38 ICLQ 850, 866.

102  See JJ Kirton and MJ Trebilcock, ‘Introduction’ in JJ Kirton and MJ Trebilcock (eds), Hard Choices, Soft Law: Voluntary Standards in Global Trade, Environment, and Social Governance (Routledge, London 2004) 3.

103  F Sindico, ‘Soft Law and the Elusive Quest for Sustainable Global Governance’ (2006) 19 LJIL 829.

104  Raustiala (n 24) 398; Guzman (n 78) 1879–81.

105  See Abbot and Snidal (n 18) 423.

106  See International Convention for the Regulation of Whaling (done 2 December 1946, entered into force 10 November 1948) 161 UNTS 72, Art V.

107  Kyoto Protocol to the United Nations Framework Convention on Climate Change (adopted 10 December 1997, entered into force 16 February 2005) 2303 UNTS 162, Art 18 (describing the procedures for adopting an enforcement system and providing that if those procedures are binding, they shall be adopted in the form of an amendment).

108  D Bodansky, The Art and Craft of International Environmental Law (Harvard University Press, Cambridge 2010) 248.

109  T Miles, ‘U.S. Blocks WTO Judge Reappointment as Dispute Settlement Crisis Looms’ Reuters (27 August 2018) at <https://www.reuters.com/article/us-usa-trade-wto/u-s-blocks-wto-judge-reappointment-as-dispute-settlement-crisis-looms-idUSKCN1LC19O>.

110  See eg Abbott and Snidal (n 18) 422.

111  This may also explain the rising use of ‘tacit’ amendment procedures in various multilateral treaties. For more, see Chapter 14.

112  Meyer (n 81) 162.

113  Qatar v Bahrain (n 61).

114  Pulps Mills (n 62).

115  Iron Rhine (n 64)

116  Klabbers (n 51) 228–9 (suggesting that the ICJ’s jurisprudence equates the existence of a commitment with the existence of a legally binding commitment).

117  United Nations Framework Convention on Climate Change (adopted 9 May 1992, entered into force 16 November 1994) 1771 UNTS 107, Art 4.2 (‘… with the aim of returning individually or jointly to their 1990s levels these anthropogenic emissions of carbon dioxide and other greenhouse gases not controlled by the Montreal Protocol’).

118  Pauwelyn (n 15) 22–8.

119  R Grant and R Keohane, ‘Accountability and Abuses of Power in World Politics’ (2005) 99 Am Pol Sci Rev 1.

120  See JO McGinnis and I Somin, ‘Should International Law be Part of Our Law?’ (2007) 55 Stanford L Rev 1176; S Wheatley, ‘A Democratic Rule of International Law’ (2011) 22 EJIL 525.

121  On another view, this hands-tying aspect is a feature, rather than a bug. See eg A Moravcsik, ‘Is there a “Democratic Deficit” in World Politics? A Framework for Analysis’ (2004) 39 Gov’t and Opposition 336, 344–6

122  ‘Letter from Senate Republicans to the Leaders of Iran’, New York Times (9 March 2015) at <https://www.nytimes.com/interactive/2015/03/09/world/middleeast/document-the-letter-senate-republicans-addressed-to-the-leaders-of-iran.html>

123  See generally Hollis and Newcomer (n 12) 575 et seq (discussing the circumstances in which the US Congress should be involved in the creation of non-binding agreements); see also MD Ramsey, ‘Executive Agreements and the (Non)treaty Power’ (1998) 77 NC L Rev 133, 143.

124  See J Nzelibe, ‘The Fable of the Nationalist President and the Parochial Congress’ (2006) UCLA L Rev 1217; T Meyer and G Sitaraman, ‘Trade and the Separation of Powers’ (2020) 107 Calif L Rev 100.

125  See Hollis, Second Report (n 90) [53]–[56].

126  The UK Parliament’s (multiple) rejections of the May government’s Brexit Withdrawal Agreement is a notable exception.

127  See eg Clean Diamond Trade Act, Pub. L. 108–19, 117 Stat. 631 (2003).

128  PH Verdier, ‘U.S. Implementation of Basel II: Lessons for Informal Lawmaking’, in Pauwelyn and others (n 15) 437.

129  Pub. L. No. 114–17, 129 Stat. 201 (2015).

130  See DB Hollis, Fourth Report on Binding and Non-Binding Agreements, Inter-American Juridical Committee, Annex I, OEA/Ser. Q, CJI/doc.580/19 (11 February 2019), Draft Guideline 4.

131  Ibid.

132  Ibid Draft Guideline 6.

133  Hollis, Third Report (n 94), Draft Guideline 3. For more on these two approaches, see Chapter 1, at 27.

134  Ibid Draft Guideline 3.2 and accompanying commentary.

135  Ibid Draft Guideline 3.3.

136  A statement of intent still might not conclusively resolve an instrument’s status in all circumstances. During the International Law Commission’s discussions leading to the VCLT, for instance, Ago argued that international law might treat agreements among States on certain subject matters, such as territory or borders, as treaties regardless of States’ intentions. [1962] YBILC, vol I, 52 [19].

137  Such a disagreement took place regarding the status of the JCPOA, with the United States and Iran taking opposite sides as to the agreement’s binding status. Similarly, the United States and the United Kingdom differed in the 1980s over the legality of a number of MOUs related to defence cooperation. See McNeill (n 31).

138  Disputes about the significance of the word ‘will’ spring to the mind of foreign ministry lawyers past and present.

139  Klabbers (n 51) 231.

140  See ibid (discussing examples).