Part III - Trade, Finance and Investment

8 Trade’s Enforcement Conundrum

Kathleen Claussen

8.1 Introduction

International law scholars have spilled much ink on questions of institutional design surrounding dispute settlement. Commentators over the last forty years have praised the concept of third-party dispute settlement as a great achievement in our sovereigntist discipline.¹ Yet, reviews of “dispute settlement mechanisms” tend to concentrate on courts on the one hand and on tribunals on the other. These are typically State-to-State mechanisms, although not exclusively so. Thus, when we consider “compliance” in international law, most questions of design concentrate on these institutions in which one State maintains that another has violated the latter’s commitments.

Today, however, the targets of international legal obligations are changing and with them, the concept of compliance, especially in trade law on which this chapter concentrates. By focusing on dispute settlement as the primary means by which to achieve compliance, the literature tends to constrain the sources it consults and the range of devices that may be appropriate to achieve its predetermined goals. It is overly limited to these now widely shared ideals about dispute resolution, operating within a closed set of options.

That our ideas of compliance are limited to settlement of disputes is the consequence of another limitation in thinking: that of the purpose and aims of international law more generally. We examine State behavior because one of the primary goals of international law is to shape that behavior. That is true likewise in trade: commitments made in trade agreements seek to install reciprocal behavioral constraints. The primary mechanism for holding States accountable in trade is, like elsewhere, third-party dispute settlement. The culmination of that exercise is license to a State to impose economic penalties (often in the form of suspension of concessions) on the party that has acted in breach of the agreement.

The “compliance” story is complicated by the terminology that practitioners and scholars apply. Trade officials and other stakeholders regularly call for greater “enforcement.” By that, they mean holding other States accountable when those other States act in a way believed to breach a trade agreement.² But recent innovations in trade agreements have concentrated on how to enhance compliance with agreement norms by both State actors and especially non-State actors. In these new iterations, “compliance” appears to encompass ideas beyond individual State action in dispute settlement, unlike “enforcement,” although they are not used precisely. Few have studied these innovations in detail and those that have tend to concentrate on the substance of the commitment or the procedures behind the tool.³ This chapter, in contrast, queries how these evolving concepts of trade “enforcement” might enhance “compliance” by diverse stakeholders – some of whom may not be readily identifiable by the governments that negotiate these agreements in the first place. It peels back the layers of the trade agreement compliance apparatus.

Recently, and as will be discussed further, some of these innovative mechanisms for compliance have gained considerable public attention. They have garnered notice for their innovativeness as well as for their considerable reach and emphasis on “trade-plus” as seemingly more important than “ordinary trade” matters. They have altered the conversation on the meaning of “compliance” and “enforcement.” These recent mechanisms have implications for how we think about trade agreements as instruments and the power of States to precipitate change behind the border.

The chapter proceeds in three parts. First, it analyzes the trade enforcement/compliance conundrum through examples arising under these innovative mechanisms in recent trade agreements. Second, it turns to an assessment of trade non-compliance mechanisms (NCMs) and argues that they both exhibit significant potential for an expansive reach and also suffer from shortcomings. Finally, the chapter closes by mapping these normative evaluations onto conventional compliance theories to draw conclusions about those theories’ resilience and flexibility before making recommendations both for trade law and international law more generally.

8.2 Trade’s Enforcement/Compliance Conundrum: The Mechanisms

Apart from international criminal law that seeks to hold individuals accountable for their actions in contravention of international law, international law focuses largely on State behavior. Central to that enterprise is the law on State responsibility which sets out generalized norms for appropriate responses to breaches of international law by States. To supplement those amorphous norms, treaties and other forms of agreements often develop their own closed set of responses to breaches of an individual agreement. Those mechanisms are most robust in international economic law where both individuals and other States have means by which to hold States accountable to their commitments under trade agreements and investment treaties. In these contexts, where the emphasis has been on securing State compliance with obligations made by States in those agreements, “enforcement” and “compliance” were often viewed as synonymous, and those terms were in turn synonymous with “dispute settlement,” even though in common parlance they are undoubtedly not synonyms. States could invoke the dispute settlement chapter of a trade agreement as a means by which to ensure that other States were complying – in effect, to enforce the agreement.

Traditionally, trade agreements have relied on adjudication for enforcement, where they have had any enforcement mechanism at all. The trade “enforcement” system is generally one of State-to-State dispute resolution carried out by a neutral panel which makes a recommendation that then must be adopted by a larger community of States or by the parties, at risk of economic penalty. In recent years, trade agreements have expanded their scope both in substance⁴ and in terms of their enforcement mechanisms.⁵ The reach of traditional adjudicatory mechanisms has been extended from, originally, economic measures to what some have called “trade-plus” measures: areas with an impact on and part of trade broadly defined such as intellectual property, labor, environment, sustainable development, and more. Consequently, trade agreement NCMs have begun to shift the compliance landscape not just on trade but also in trade-adjacent areas as will be described. Central to this thesis is that new moves in the trade law and policy space are changing how we think about compliance, and the central actors behind the agreements.

While trade negotiations have always reflected a cacophony of views and interests, the issues surrounding enforcement of trade-plus matters are even more muddled than those generally complex debates. Actors from various political backgrounds have advocated for different institutional structures to address complications in supply chains, migration, the perceived effects of globalization, and domestic economic policies. Many States still do not subject the bulk of those “trade-plus” measures to ordinary dispute settlement, making their enforceability seem elusive to some advocates.⁶ In other instances, however, States have gone beyond ordinary State-to-State dispute settlement in their trade agreements on these “trade-plus” matters.

8.2.1 Innovations in Trade-Plus Enforcement

New mechanisms seek to shift the enforcement effort toward other actors including but not limited to the State. Most of these come from the US experience where the voices of advocates for greater domestic penetration by trade agreements are loudest. The most recent and most celebrated example of this move comes from the 2020 economic agreement signed by the North American countries known as the United States–Mexico–Canada Agreement (USMCA) which launched a new tool: the “Facility-Specific Rapid Response Labor Mechanism” (RRM).⁷ This tool was the product of an intense deliberation between the Democrats in the US Congress and the Trump Administration in which the former sought greater labor protections in the USMCA text.

The RRM is a novel compliance tool and a supplement to the State-to-State dispute settlement system. Operating in addition to the traditional adjudicatory mechanism for disputes about the interpretation and application of the labor chapter, the RRM is a unilateral means for any of the parties to seek to force individual worksites in the territory of another party to comply with domestic law.⁸ It has a relatively straightforward aim: to ensure remediation of a denial of collective bargaining rights. Contrast this with one of the primary aims of most other labor chapters in US trade agreements to date: to ensure that each party does not fail “to effectively enforce labor laws through a sustained or recurring course of action.” Rather than hold governments accountable for the administration of their domestic law, the RRM is a way to handle a specific denial of the right of free association and collective bargaining by a private entity at a particular worksite. It still turns on engagement between the relevant two governments and the use of an arbitral panel, but it is far more incident-specific than the traditional adjudicatory labor compliance mechanisms.

Another feature of the USMCA trade-plus compliance machinery is similar to that used in other US trade agreements: public engagement. As in the context of other US trade agreements, each of the three governments in the USMCA has devised a system for receiving information about possible denials of rights at worksites, although at the time of writing, the governments had not committed to making publicly available information they received. This mechanism expands the reach of each of the governments, allowing individual actors and civil society to bring to the attention of the authorities problems of which they are aware.

Following a multi-step investigation process, governments can prompt the constitution of a neutral panel of labor experts to make a determination on whether workers are being denied their rights at the factory in question.⁹ Where the panel identifies that the factory is denying workers’ rights, the responding government may choose to hold consultations with the complaining government before the complainant then “impose[s] remedies.”¹⁰ Importantly, those “remedies” or penalties are not against the respondent but rather against the worksite company itself. The complainant can choose a remedy which “may include” suspending preferential tariff treatment for the goods made at that worksite or other “penalties” on the goods.¹¹ At the least, the agreement text specifically notes that the complainant may deny entry to goods from the company in question in instances of repeat offenses. The complainant is limited only by some imprecise principles of proportionality. The “remedy” continues to be applied until the denial of rights has been “remediated.”¹²

But relatively broad language in the agreement suggests the complaining government is not limited in its choice of remedies other than in their magnitude. That is, the agreement provides a non-exhaustive list of possible “remedies” against a worksite or the home government and only requires that the complaining government’s selection be proportional. As of the time of writing, there had been no test of the panel or remedies aspect of the RRM.

The United States activated the RRM tool twice during the summer of 2021 and once more in May 2022. In the first instance, it reached agreement with Mexico on how to remediate the situation. In the second, the United States reached out directly to the company and its US parent company to develop a means to address the worksite issues. The third remains ongoing at the time of writing.

A second innovative and recent model for trade-related compliance may be found in the United States–Peru Trade Promotion Agreement (PTPA) which entered into force in 2009.¹³ The PTPA contains a unique Environment Chapter and Annex on Forest Sector Governance, which includes a requirement for Peru to conduct audits of particular timber producers and exporters, and upon request from the United States, perform verifications of shipments of wood products from Peru. The United States may then take action directly against the shipment that is subject to the verification if it is not satisfied with the results. The purpose of this Annex is to protect against illegal logging that threatens to deplete forests and exacerbate climate change. The Annex also creates an Interagency Committee on Trade in Timber Products from Peru to monitor Peru and its companies’ actions in this respect.

The PTPA Forest Annex was and remains the first of its kind, although some of the aspects of the USMCA RRM resemble its features. The PTPA offers an illustrative list of actions the United States may take with respect to the shipment or enterprise that is the subject of the verification. For example, the United States may deny entry to certain products for up to three years, or until the Timber Committee determines that the company in question has complied with all applicable laws, regulations, and other measures of Peru governing the harvest of and trade in timber products, whichever is shorter.

The company-specific aspect of the Annex was tested first in October 2017 when the Trump Administration denied entry of timber products and exports by a Peruvian company. Peru was unable to verify that the shipment complied with all applicable Peruvian laws and regulations. Again, in July 2019, the US Trade Representative directed US Customs and Border Protection to block future timber imports from a different Peruvian exporter based on illegally harvested timber found in its supply chain. The Trump Administration characterized this move as a way to “ensur[e] that [US] trading partners live up to their [trade agreement] obligations” by not allowing illegally harvested timber to be exported to the United States.¹⁴

8.2.2 Drivers of Change

What has motivated these shifts in the US trade agreement compliance context? To be sure, examining the wide variety of motivations from multiple sectors and among multiple actors exceeds the scope of this chapter. But when we ask where these moves come from and why they have emerged, there appear to be two important features: first is a frustration with trade liberalization and second is recognition that existing or traditional means to countervail what many see as the negative effects of liberalization have failed to do so.¹⁵ Thus, the obligations have shifted and the NCMs have followed. It is a two-step design story. The agreement provides for the home State to ensure the State maintains certain domestic commitments, often modelled after international standards, such as those found in the International Labor Organization Conventions. Where the State cannot deliver on those, the new mechanisms allow the other State to exact demands on private actors within the territory of the other State. While States still enter into reciprocal commitments in these trade-plus areas, they do not rely on State action to ensure those commitments are upheld.

One can again trace this two-step move in the debates surrounding the mechanisms for enforcing labor commitments in the USMCA leading to the RRM. Some lawmakers insisted that the US Government devise a unilateral enforcement mechanism such as that which emerged; others argued for an inspection system through which US officials would visit Mexican factories; some sought third-party dispute settlement with enhancements of various types; and still others preferred the status quo.¹⁶ It was one of the most exploratory and expansive debates on institutional design for labor since the development of the original North American Free Trade Agreement (NAFTA) labor protections in the early 1990s.

The final product of the RRM and likewise the United States–Peru logging arrangement seek to replicate and supplement the domestic enforcement chain by outsourcing the ultimate compliance pressure to the complaining State. These mechanisms that reach beyond the border into the territory of a trading partner extend the traditional trade law institutional design questions from actions taken by that trading partner to actions taken by private actors on the ground. They empower the receiving or complaining State to penalize firms that are violating domestic law. By shifting the roles and responsibilities through these NCMs, these developments have upended the traditional discussions on enforcement institutional design in trade law and called into question whether the field has a reliable and consistent theory of compliance at all.

8.3 Assessing these New Tools: Costs, Benefits, and the Space Between

Trade NCMs like public complaint processes regarding labor rights violations or environmental abuse are not a new concept, but their latest experimental forms advance their promise. While these advances have benefits, they also have shortcomings. This section briefly assesses the behind-the-border direct compliance obligations along two dimensions. First, it will evaluate the practical consequences of such a scheme. Second, it will turn to the transformative power of this approach.

8.3.1 Practical Problems

There are at least seven issues that behind-the-border compliance schemes raise and that some critics have noted as creating potential problems for the legitimacy of such tools: transparency, predictability, selection bias, irony or hypocrisy, aggrandizement of authority, extraterritoriality, and judicial review. I will take up each in turn.

First is transparency. The present iterations of these tools are largely internal guiding schemes, but they do little to require governments to make publicly available the documentation or reasoning behind the steps taken. Transparency is not a problem unique to behind-the-border schemes in trade law¹⁷ but the problem is surprisingly acute in these new tools – perhaps as a result of their newness.

Second is predictability. These tools do not create any degree of consistency in outcome. While there are some guidelines and the commitments are articulated to a limited degree, some private companies have noted that there is not enough information available to them to be able to comply effectively with the substantive and procedural requirements involved. They have raised questions of due process and coordination which they claim mechanisms lack.

Third is selection bias and power differentials. Although the tools are reciprocal,¹⁸ the expectation, and so far, the practice, has been one-sided. That is, in the case of the RRM, only the United States has taken action against companies operating in Mexico, rather than the other way around, consistent with the expectation of many. The reasons for this lopsided situation include (1) the United States has limited the scope of actions that Mexico could take toward US companies to a very limited set;¹⁹ (2) the power dynamics of the unions in the United States; and (3) power dynamics between the countries. While these are still instruments to which States agree, the bargains are not always reciprocal. The result is often advanced and wealthy economies pursuing compliance by not just developing States but now entities in developing States, including multinationals from the wealthy State. These are tools through which the United States and other countries can exert force directly on companies operating in places like Mexico and Peru and not because of any characteristic of the product that those companies seek to import, but rather based on their behavior behind the border in those other States.

Fourth is irony, or what some would call hypocrisy. Concern about the operation of the RRM is not just limited to the selection of what might be subject to the RRM. It is also a bias in result – what some would call hypocrisy or irony. This is where considerable criticism has been raised about the application of the RRM to date. The mechanism permits US Government actors to demand of companies operating in Mexico labor practices that it cannot demand from US companies operating just a few kilometers across the border.²⁰ A related point is the selection bias in the types of societal problems these US-designed tools are intended to solve. One does not see – yet – similar tools focused on private action to address carbon emissions, for example.

Fifth is aggrandizement of authority. The experience of the United States with its second RRM situation at a company called Tridonex in Mexico has led to some criticism about the possible aggrandizement of authority on the part of US authorities. In that context, rather than reach agreement with Mexico about a course of remediation as provided by the agreement, the US Trade Representative entered into an agreement with the company, in which the company promised to take certain actions in the interests of the workers at the identified worksite. Nothing about the RRM language or the United States implementing legislation provides authority to the US Trade Representative to enter into agreements with companies. Thus, this agreement, apart from raising questions of legality, has also suggested to some outsiders that these compliance moves are means through which US executive branch agencies may grow their authority.

Sixth is extraterritoriality. In the State-to-State context, some governments have traveled to the other State to carry out in-person investigation of the measure or practice, or they have relied on their government officials already on the ground in those places. Indeed, as noted, for some trade-plus areas, public submission or consultation opportunities have allowed foreign civil society actors to provide information from inside the breaching State to the complaining government to expand their reach. But the RRM and the Peru mechanism go still further in two respects. One way is by targeting individual companies not operating in the United States. Second is by taking action against companies for their violations of domestic labor law in Mexico (and likewise for their violation of domestic law in Peru).

And seventh, finally, is judicial review. Although one of the aforementioned mechanisms (the RRM) involves the possibility for review by a neutral panel, the actions undertaken by the US agencies involved in raising these accusations and taking such actions are not subject to review. Although their actions are regulatory, much like anti-dumping and countervailing duty actions taken by US Customs and Border Protection or the US Department of Commerce, companies affected do not have means to challenge their targeting. They cannot bring claims to a judicial body to raise concerns about the process through which they have been prosecuted.

8.3.2 Achievements

Despite these criticisms, advocates celebrate the new compliance tools for making positive change in the lives of many workers, as well as in combatting climate change in the case of the Peru logging situation. Further, we can respond to each of the criticisms above to defend the deployment of these tools. Some of the criticisms are about the application of the tools to date rather than the existence of the tools themselves, for example. To take one, the hypocrisy critique may be seen as short-sighted, given that these tools may assist with norm-building at the international level, even if US institutions will not permit those norms to be entrenched at home. Like elsewhere, we may be able to achieve internationally what we cannot achieve domestically due to political constraints. Similarly, the extraterritoriality claims are somewhat tempered by the fact that these do not involve US authorities arriving at Mexican companies or at Peruvian logging sites and arresting people. The action that the US Government can take is limited to that which it can exert at its borders. In theory, companies that come into focus for compliance may direct their goods elsewhere.

In sum, while these new mechanisms create some problems, they also may make positive changes toward broader goals. It is too early to draw conclusions in evaluating these tools. Further, measuring success in these trade-plus NCMs is just as difficult as it is in measuring success in other areas. A wide range of metrics may be appropriate. One metric that is often raised is increased use: the more we use an NCM, the more successful it is, according to some lawmakers. But for tools that are intended to create penalties for bad behavior, the better metric may be that there are no instances of bad behavior and therefore the NCM has effectively created a deterrent effect. Likewise, an application of the tools is hardly a sign of success. Counting up their uses does not begin to take into account the possible ancillary effects they may have on communities or within workplaces and in other respects. In the case of the RRM, some commentators have raised concern about how the US engagement with the targeted worksites may exacerbate other local problems, whether related to violence or otherwise. There are not simply winners and losers in these stories; the impact is far more textured than meets the eye.

8.3.3 In Between

These tools have also precipitated ideological shifts and normative changes that are difficult to square as criticism or achievement, as cost or benefit. Those sorts of assessments shift easily with one’s perspective. For example, some proponents of free trade have seen these tools as advancing protectionism, and have suggested that in so doing, they create avoidable costs for all. These critics argue that the creators and users of these tools aim to discourage US investment in Mexico and encourage US and multinational companies to invest instead in the United States. Through a combination of tariff tools and these behind-the-border mechanisms, the United States is seeking to ensure that US and other companies will not choose Mexico over the United States.²¹ Others would say this is not just about keeping US businesses at home. To these advocates, helping Mexican workers is paramount and these sorts of tools permit the US Government to help the Mexican Government achieve that goal, particularly in areas where the Mexican Government is constrained from making progress. In both camps are those for whom the best way to promote US economic competitiveness is by making sure competitors in Mexico are held to the same standards.

At the least, these tools recognize that changing State behavior is challenging and that States’ comparative advantage, or at least that of the United States, may be in changing company behavior. That is a typical domestic law function. These tools position States to achieve these goals in new ways, to decide that certain types of trade are good and other types are bad, and to put a thumb on that scale. Thus, we can evaluate these tools as a matter of policy, and as a matter of practicality, but we might also consider whether one can judge the achievements of these new institutions merely by their concrete outcomes. We can likewise examine these tools based on principles about how laws are designed to operate, where the power lies, or the value of this particular legal practice.

The operation of these new compliance tools has multifaceted effects that space does not permit me to explore here, but we might ask the following questions as a means of further evaluation from either a trade perspective or a compliance perspective: Are they redistributive or wealth-producing? Do they produce better governance? Are they norm-enhancing? Are they gap-filling or capacity-building? Are they hand-tying (self-regarding or other-regarding)? These questions and others like them require greater exploration before we can draw conclusions about trade NCMs.

8.3.4 Lessons for Compliance Theories in International Law

The new trade NCMs with their behind-the-border compliance focus also offer lessons for compliance theories more generally in international law. I highlight just a handful here for further exploration in later work.

First, behind-the-border NCMs adjust the levers in the conventional compliance stories that have dominated international law scholarship.²² In addition to focusing largely on State compliance with agreements or with dispute settlement reports, prior scholars have sought to explain how international law serves as a constraint on behavior. The trade NCM experience of recent years aligns the features of our international law compliance story in different ways. The ordinary touchpoints of the realist or rational functionalist literature are not salient here. For example, reputation – a key input in the functionalist compliance theory – has a diminished role in the operation of trade NCMs. Rather, the trade NCM developments purport to align more with normative theories and identify normative aims of the States involved. But that analogy is likewise strained. The shifting targets are part of the categorization issue: there is no common view on the problem these tools are intending to solve, making the compliance discussion much harder to formulate.

Second, rather than look at trade NCMs as dispute resolution or “enforcement” tools for States, we might characterize this activity as “regulation.” Then we might ask: What are the options, values, and scholarship that we might seek to integrate in examining such moves? What are the implications of characterizing them as one or the other? In the existing literature on compliance, some scholars have considered the role of sanctions and incentives when comparing an “enforcement model” with a “management model.” Here, we are somewhere between those poles. Again, this comparison is constrained by the differences in levels of actors involved in this story and the regulatory nature of some of them. While the literature to date has looked at first and second-order compliance, these trade NCMs might be properly characterized as third-order, given bureaucratic attention to private actors. In third-order compliance, the bureaucracy deepens as trade agencies move from negotiation and litigation to regulatory monitors.²³

Third, these NCMs often try to replicate the domestic enforcement chain through new institutions. They do not rely on domestic actors as in a complementary system, but they include multiple steps of review by bureaucratic actors. Putting bureaucrats at the center of these mechanisms may enhance their legitimacy in the eyes of a domestic public. This approach also allows that government to retain control over the process and its outcome, as compared to using international adjudicators. In these ways, these mechanisms go beyond what other scholars have tracked in international law, blurring the lines between domestic and international authority.

Fourth, trade theories usually instead speak in terms of remedies rather than sanctions and they do so with little emphasis on deterrence given that “dispute settlement” language conveys a sense of misunderstanding rather than bad behavior. This distinction is more than a matter of semantics. Rather, the new compliance models operate within this trend and foundational understanding of how States ought to interact with one another. It becomes impossible to isolate the impact of these new tools and norms from the broader backdrop on remedies rather than the typical punitive action. There are, in fact, multiple layers of compliance at work in ways not seen before.

Fifth, the trade NCM experience demonstrates how control has shrunk in importance while cooperation is more prominent. Query then, whether cooperation can constitute compliance. In each of the aforementioned tools, the States must cooperate, and ultimately the companies likewise. Even if they are not aimed at correcting State conduct that diverges from the text of international obligations, can they still qualify as compliance mechanisms? When the US Government works with companies to fix problems abroad, is that a matter of compliance? These trade NCMs do not rely on either dispute settlement, or domestic adjudication, or entities that use force as we can see in other areas of government and international law. Rather, the trade NCMs reconsider the dynamics among the parties and expand the meaning of “party” broadly, again making traditional compliance theories more difficult to apply here.

This review of the contributions made by trade NCMs to compliance theory suggests that these may merit a “new category” of transnational compliance ideas or theory that forces a reevaluation of the structure of the compliance problem. If that is so, then it is worth complicating the analysis somewhat further by examining whether the parties involved in shaping this mechanism have a shared understanding of its goals and objectives.

8.4 Conclusion

The conundrum of trade-plus enforcement is the absence of a shared understanding of what enforcement is for and what institutional designs serve those likely diverse ends. Rather, the system has marched toward making commitments enforceable through third-party dispute settlement across a wide array of issue areas. For nearly twenty-five years, as free trade agreements have proliferated, most trade-plus advocates and members of civil society have advocated for third-party dispute settlement and the availability of the same remedies for trade-plus disputes, as for conventional commercial disputes as a means to serve their ends.²⁴ Opposition to this now-establishment view has manifested not based on effectiveness of the third-party mechanism, but rather based on concerns that the concrete remedies obtained through these types of mechanisms could create additional barriers to trade – direct and indirect. The result has been a narrow two-sided conversation about the range of possibilities for trade-plus enforcement.

This binary representation of trade-plus enforcement has diminished the relevance and prominence of a more fulsome conversation about institutional design. Today, the growing prevalence of trade NCMs pushes the boundaries of that conversation and opens a field for research in compliance theory that this chapter has sought to begin.

9 How the World Bank’s Dispute Resolution Services Should Benefit Affected Persons and Borrower States

Jonathan Brosseau

9.1 Introduction

Members of the Kawaala community in Kampala, Uganda, report that the Kampala Capital City Authority and armed guards woke them in the early hours of 4 December 2020, and began destroying their homes and farmlands.¹ Kawaala community members allege that their eviction and the ensuing destruction paved the way for the World Bank-funded Lubigi drainage channel project. They claim that the project began without proper consultation or plans for compensation and resettlement, and compromises their livelihoods and well-being, in violation of the Bank’s Environmental and Social Framework (Framework) policies.² In turn, Bank management claimed that the project-level grievance mechanism should handle most of the Kawaala community’s concerns about resettlement and that it is working with Kampala Capital City Authority to strengthen the resettlement plan related to the channel project.³ Dissatisfied with the World Bank’s (Bank) solution to their concerns, Kawaala community members filed a complaint to the Inspection Panel (Panel) in June 2021 with the support of local and international civil society organisations. The Panel subsequently recommended to the Bank’s executive directors that the complaint be investigated.⁴

Upon the executive directors’ approval of the Panel’s recommendation for inspection,⁵ the Kawaala community and Uganda were offered the opportunity to pursue dispute resolution rather than to go forward with the compliance review conducted by the Panel, a first in the Bank’s history. Indeed, the executive directors had only approved the updated Inspection Panel Resolution⁶ and the Accountability Mechanism Resolution,⁷ which established the new Dispute Resolution Services (DRS) in September 2020.⁸ Under the DRS, those affected by Bank-funded projects (affected persons, also referred to as requesters once they submit a request for inspection to the Panel) and the borrower State can now resolve a complaint before the Inspection Panel through joint fact-finding, mediation, and other similar approaches, where both agree to this process. In the autumn of 2021, the DRS was staffed and published its Interim Operating Procedures to implement dispute resolution processes. The Kawaala community and Uganda agreed to pursue dispute resolution shortly thereafter.⁹ According to a civil society organisation supporting the Kawaala community, the dispute resolution process would provide an ‘appropriate forum’ for the community to raise their demands, which include ‘a new, proper land survey and identification of project affected persons, provision of adequate compensation, and adequate time to resettle.’¹⁰ At the time of writing, two more cases have also begun dispute resolution process.¹¹

In July 2022, the accountability mechanism published a second version of the DRS’ operating procedures, on which it sought comments from any interested individuals and organisations.¹² An earlier form of the present chapter was therefore sent to the accountability mechanism, recommending amendments to the procedures similar to the ones set out below. Accountability Counsel and fifty-six other civil society organisations also submitted substantial joint comments on the procedures.¹³ In December 2022, the accountability mechanism published the third and final version of the DRS’ operating procedures: the 2022 Accountability Mechanism Operating Procedures. In contrast to other multilateral development banks, the World Bank did not make public the comments it received, simply noting they were from civil society organisations, other accountability mechanisms, former Panel members, and scholars. As will be explained, the revision in the accountability mechanism operating procedures that strengthens requesters’ protection the most is that the requirement that additional advisers may be engaged by one Party only subject to the other Party’s consent, which was present in the first two versions of the procedures, was removed. As also explained below, the revision that weakens requesters’ protection the most is that, while the first version required that dispute resolution agreements be consistent with Bank policies and domestic or international law, the final accountability mechanism operating procedures requires that the Accountability Mechanism Secretary requests the Parties ‘to make appropriate modifications’ if she believes there are inconsistencies with domestic or international law only.

Given the novelty of the DRS, this chapter examines how the dispute resolution process offered by the Bank should benefit affected persons and borrower States (together, the Parties). The chapter also critically evaluates the dispute resolution process in light of the mandates of the Inspection Panel and the DRS, as well as best practices concerning the right of access to a remedy under international law. In particular, it considers the strengths and weaknesses of the current dispute resolution process and suggests how the Bank should improve this process when it revises the Accountability Mechanism Resolution in late 2023¹⁴ or the accountability mechanism operating procedures in the future. By noting how the DRS procedures have been reformed through the three versions, the chapter also records the evolution of the DRS, which includes identifying whether civil society organisation-suggested revisions have been incorporated into the procedures and whether the revisions implemented are more or less protective of affected persons. As such, the chapter seeks to contribute to the global administrative law scholarship.

The task of evaluating the DRS as part of the Bank’s accountability system matters for several reasons. First, it will enable the Parties to understand better whether the DRS is indeed the ‘appropriate forum’ for resolving the complaint in each case, and whether they should therefore consent to it or opt instead for the compliance review process undertaken by the Panel. Further, the Bank, its executive directors, and its accountability mechanism could consider this evaluation when revising the DRS. The nearly twenty other multilateral development banks, which have been influenced by the Bank’s practices on accountability in the past,¹⁵ could also consider the parts that apply to them when revising their respective dispute resolution processes.

Overall, the chapter suggests that the DRS, as it is now designed, has the potential to enhance the right to access a remedy of affected persons. By emphasising party-led dispute resolution, it provides affected persons with the assistance of an independent third party, the opportunity to participate in the determination of remedial measures, and the chance to receive effective remedies beyond those envisaged by Bank policies. The heavy reliance of the DRS on the consent of both Parties may also serve to preserve a significant role for the Inspection Panel, because the Parties may not come to an agreement through dispute resolution in most cases. At the same time, this reliance on consent also enables affected persons to agree to remedies to some degree inferior to those envisaged by Bank policies. This is a real possibility, particularly where affected persons continue to bear the adverse material effects of violations of Bank policies while the dispute resolution process runs its course, and are typically vulnerable populations with fewer resources and less expertise on Bank projects than borrower States. Moreover, as Bank management is involved in the dispute resolution process only if the Parties consent to it and as a technical observer, management cannot ensure that affected persons obtain a meaningful remedy. Given that the DRS does not fully address these concerns at the moment, the question arises as to whether it will actually enhance the access to a remedy of affected persons, and whether it may instead prejudice the Panel’s mandate to provide access to a remedy to these persons. Against this backdrop, the Bank should consider altering the dispute resolution process to better address the inequality of power between affected persons and borrower States, such as by entrenching certain minimal safeguards for affected persons, to ensure that they make informed decisions on remedies.

The chapter proceeds in two sections. Section 9.2 examines the mandates of the World Bank’s three avenues for a remedy as they pertain to the right of access of affected persons. These avenues are the Inspection Panel, the Grievance Redress Service, and the DRS. This section shows that the Panel and Grievance Redress Service have succeeded in performing their respective functions of providing independent compliance review and management-led solutions. However, because they did not offer independent dispute resolution, a gap nevertheless remained in the Bank’s accountability system. The establishment of the DRS filled that gap by offering affected persons access to independent dispute resolution processes. It did so while enhancing the three pillars of the Panel, which can be identified as: effectiveness, accessibility, and independence.

Section 9.3 then proposes three areas of improvement to the DRS that the Bank should consider to comply with the mandate of the Inspection Panel and best practices related to accountability mechanisms. To enhance accessibility, the Bank should consider strengthening the procedural protections and opportunities for participation afforded to affected persons in the dispute resolution process, notably by providing a minimum standard of access to project-related materials. To enhance effectiveness, it should consider clarifying the minimum threshold for remedies that affected persons can accept and provide for the mandatory verification of the implementation of the Parties’ agreement. Finally, to enhance independence, the Bank should consider offering affected persons more options regarding the sequencing of compliance review and dispute resolution to protect the mandate of the Inspection Panel, and offer them funding to get support from professionals during the dispute resolution process.

9.2 The Mandates of the Bank’s Three Avenues for a Remedy

Section 9.2 provides an overview of the three avenues for a remedy within the World Bank that enables the Bank to meet its moral and legal obligations to provide affected persons the right of access to a remedy, given its immunity from suit in national courts.¹⁶ This serves to identify the legal and policy standards against which the avenues for a remedy provided by the Bank should be evaluated.¹⁷

9.2.1 Inspection Panel: Panel-Led Compliance Review

The Inspection Panel was established in 1993 as the first independent accountability mechanism at an international financial institution. The Panel’s mandate is to determine whether the Bank complies with its operational policies and procedures in any particular case. While the Panel thus has a compliance function and adopts a fault-finding approach, it also provides affected persons with a basic right of access to a remedy.¹⁸ It is also a quasi-judicial body: as described below, the executive directors of the Bank cannot change its findings, but they retain the power to decide on the outcome of requests at key stages of the process. For example, the Panel cannot issue binding orders, whether interim or final, as courts can.

The right of access to a remedy provided by the Inspection Panel can be distilled into three pillars, which are at once procedural and substantive.¹⁹ The first pillar is effectiveness, which is limited in practice by the Panel’s mandate. Once it receives a complaint, the Panel first issues its recommendation to the executive directors on whether a full investigation should be carried out.²⁰ If the executive directors approve an investigation, then the Panel submits its findings of facts regarding the Bank’s compliance with its operational policies and makes any related findings of harm.²¹ Although these findings are non-binding, they enable Bank management to propose remedial actions to prevent any non-compliance and harm from continuing. The Inspection Panel itself does not recommend remedial actions.

The second pillar is accessibility. The Panel has broad eligibility criteria, according to which any two or more affected persons may submit a request.²² The opportunity for procedural participation afforded to affected persons is also relatively broad, as they can provide information about the facts underlying the complaints during the investigation.²³ They are also ‘consulted’ on the plan of action agreed between the Bank and the borrower State on remedial efforts, but do not have decision-making power.²⁴

The third pillar of the right is independence and impartiality.²⁵ The Panel must be independent not only from Bank management, but also from the borrower States and requesters. The Panel must also be impartial to the merits of the complaints, meaning it should deal thoroughly and fairly with the requests brought to it. On this basis, the Panel is required to give reasons based on the evidence and facts supporting its recommendations and findings.²⁶

Best practices have developed in the three decades since the establishment of the Panel, suggesting today that the right of access to a remedy provided by international organisations, including multilateral development banks, should include access not only to a compliance review process but also to a dispute resolution process. Notably, the 2011 UN Guiding Principles on Business and Human Rights (UNGPs) identify best practices regarding access to a remedy, and in particular to dispute resolution functions. While most relevant for States, the UNGPs also set, by analogy, a benchmark to assess how multilateral development banks should provide access to a remedy.²⁷

Two of the dispute resolution mechanisms envisaged by the UNGPs are particularly relevant to multilateral development banks like the Bank. The first is ‘effective operational-level grievance mechanisms’, which should remedy complaints early and directly.²⁸ The second is ‘effective and appropriate non-judicial grievance mechanisms’, which must be part of a comprehensive system to address complaints.²⁹ These two types of mechanisms complement, but do not substitute, each other.³⁰ In terms of an effective remedy, both mechanisms must ‘ensur[e] that outcomes and remedies accord with internationally recognised human rights’.³¹ This criterion, among others, has been endorsed by a Bank publication on the evaluation of grievance mechanisms.³²

In response, in part, to the development of best practices concerning the right of access to a remedy, twenty multilateral development banks have established accountability mechanisms similar to the Inspection Panel to provide access to remedies through a compliance review process. Many of these banks have also included dispute resolution processes to increase the effectiveness of access.³³ All multilateral development banks similar to the Bank in terms of size and function, for instance, provide access to dispute resolution processes today.³⁴

The Inspection Panel has successfully exercised its mandate, even if it has not provided an effective remedy to affected persons through dispute resolution in line with best practices. The Bank receives complaints yearly on about 3 per cent of its 250 ongoing projects, and of that 3 per cent of projects, the Inspection Panel investigates about a third.³⁵ Most complaints concern environmental assessment, investment project financing, consultation/disclosure, and involuntary resettlement.³⁶ In terms of its fault-finding approach, the Panel has generally been very successful in holding the Bank accountable and promoting institutional learning.³⁷ It has also been moderately successful in preventing future harm.³⁸ For example, on the Uganda Transport Development Project, then-World Bank Group President Jim Yong Kim explained that ‘[t]he Inspection Panel’s investigation into the … Project identified multiple failures, including cases of gender-based violence’, which played ‘an important role in the Bank canceling the project’.³⁹

However, the Panel has been less successful in remedying the harm already suffered by affected persons.⁴⁰ While the result of the Panel’s investigation is to bring the project back into compliance, it does not guarantee compensation for affected persons in relation to the harm that occurred.⁴¹ Moreover, according to one study, compliance investigations at the Bank take on average fifteen months.⁴² Such a delay is significant for many affected persons, especially when investigations concern allegations of serious harm. Finally, as noted above, affected persons have no decision-making power on the remedial efforts agreed between the Bank and the borrower State. Management and the executive directors may also ignore – and in some cases have ignored – the findings of non-compliance by accountability mechanisms like the Panel.⁴³ In short, the Inspection Panel does not offer affected persons the same access to a remedy through a problem-solving approach as a dispute resolution process would.

9.2.2 Grievance Redress Service: Management-Led Solution

To bring its accountability system further in line with best practices concerning accountability mechanisms, the Bank established the Grievance Redress Service and two mechanisms related to the Inspection Panel.⁴⁴ First, the Grievance Redress Service is a complaint-handling mechanism that helps project teams broker solutions at the corporate level.⁴⁵ Established in 2015, it reports to senior Bank management. The mandate of the Grievance Redress Service is to address complaints directly and effectively with the project teams, with the purpose of ‘[closing] the gap between project-level grievance redress mechanisms … and the Inspection Panel in the Bank’s accountability structure’.⁴⁶ Seeking resolution first through one of the recourses offered by the Bank, such as the Grievance Redress Service, is one of the preconditions for submitting a complaint to the Inspection Panel.⁴⁷

The growing number of cases that the Grievance Redress Service receives each year demonstrates that it has effectively provided affected persons with access to certain remedies.⁴⁸ In 2020, the Grievance Redress Service worked on 211 admissible cases at different processing stages concerning various project-related issues.⁴⁹ It has also regularly implemented changes that have enabled it to perform its mandate better. For instance, the recent addition of an ‘escalation clause’ in its Directives allows the Grievance Redress Service to bring high-risk complaints to senior management’s attention quickly.⁵⁰ Given its features, the Grievance Redress Service, like project-level grievance mechanisms, fulfil the function of ‘operational-level’ grievance mechanisms envisaged by Principle 29 of the UNGPs. It has strengthened the governing framework of the Bank’s accountability mechanisms in a way that has complemented the mandate of the Inspection Panel.

But while the Grievance Redress Service has been successful at resolving relatively simple disputes concerning operational issues, it has been less successful at resolving disputes concerning more complex or controversial issues. This is in part because the Grievance Redress Service does not report to the top level of the Bank and has a junior status in the Bank hierarchy, which hampers its operation for those disputes.⁵¹ Its efficiency in resolving complex issues is also limited by its (real or perceived) lack of independence from management.⁵²

The limitations of the Grievance Redress Service have questioned whether the Bank was meeting best practices on the right of access to a remedy, given that the Grievance Redress Service was the only dispute resolution mechanism offered by the Bank itself for a long time. Neither the Grievance Redress Service, nor other avenues for a remedy, then fulfilled the function of ‘non-judicial’ grievance mechanisms envisaged by Principle 29 of the UNGPs. This has had implications for the credibility and reputation of the Bank, since all other multilateral development banks have been offering dispute resolution at the top level of the institution.⁵³

As mentioned, the Bank also introduced a second set of options to settle the complaints of affected persons. This came in the form of two mechanisms related to, but formally outside of, the Inspection Panel’s process. The first mechanism was a 2013 pilot project in which the Inspection Panel was empowered to postpone its decision on registration of a request, and thereby delay triggering the twenty-one-business day period for management to provide its response.⁵⁴ The second mechanism was based on the Inspection Panel’s 2014 Operating Procedures and entailed that the Panel delayed making a recommendation on investigation for a stipulated period.⁵⁵ Both mechanisms aimed to provide affected persons and management with more time to develop early solutions to complaints without a formal investigation by the Inspection Panel, to improve the ‘effectiveness’ of the access to a remedy of affected persons, while simultaneously adhering to the mandate of the Inspection Panel.⁵⁶

Despite the objective of these dispute resolution mechanisms, their success in practice was doubtful. The mechanisms were only employed in a few cases, which meant neither was subject to a systematic review of its effectiveness. A first-hand account of the only two cases that went through the first mechanism – the postponement of registration – suggests that one case was reasonably successful and the other was unsuccessful.⁵⁷

More significantly, concerns arise as to whether the mechanisms complied with the mandate of the Inspection Panel, let alone with best practices on access to a remedy. By seeking to improve the first pillar of the Panel (i.e., effectiveness), the mechanisms may well have compromised the other two (i.e., accessibility as well as independence and impartiality). As to independence, ‘[t]hese mechanisms blur[red] the clear distinction between the [Inspection Panel]’s responsibilities as an independent and objective fact finder and management’s role in the [Inspection Panel] process.’⁵⁸ For instance, the mechanisms lacked a neutral mediator that would oversee the problem-solving process.⁵⁹ As to accessibility, the mechanisms did not offer affected persons a meaningful opportunity to participate in the design and implementation of measures to address their complaints, and lacked procedural safeguards to counteract the inherent power imbalance between them and Bank management.⁶⁰

In summary, the Bank’s introduction of the Grievance Redress Service, the Pilot Project, and the Operating Procedure footnote to offer affected persons with options for dispute resolution can be interpreted as an acknowledgment of the dispute resolution gaps in the Bank’s accountability system. But because these mechanisms did not adequately fill the gap of an independent dispute resolution process, the Bank introduced a third avenue for a remedy: the DRS.

9.2.3 Dispute Resolution Services: Party-Led Dispute Resolution

The DRS was established in 2020 to increase the access to a remedy of affected persons through dispute resolution processes in addition to, but not as a substitute for, compliance review processes under the auspices of the Panel. This development was precipitated by the approval of the Bank’s revised operational policies and procedures, the 2016 Framework. The Framework, among other things, aligned with the concept of due diligence promoted by the UNGPs,⁶¹ and included the requirement that every Bank-funded project has a project-level grievance redress mechanism.⁶²

Following an external review and the recommendation of Bank management, the executive directors agreed to establish the DRS along the following lines. First, the requesters must meet the eligibility criteria for submission of requests to the Inspection Panel, and the executive directors must approve an Inspection Panel recommendation to investigate the project. Then, should both the requesters and the borrower State voluntarily agree, they would have the opportunity to resolve their disputes through dialogue, information sharing, joint fact-finding, mediation, and conciliation. In this case, the Panel will hold its compliance process in abeyance until the dispute resolution process concludes.

While the staff of the DRS will ‘administer’ the proceedings, an external neutral third party will help the Parties reach an agreement. With the agreement of the Parties, Bank management may be an observer in the DRS process, although the role of management remains only technical.⁶³ At the end of the dispute resolution process, the DRS will issue a report to the executive directors through the Accountability Mechanism Secretary, informing them of the outcome of the process. If the Parties cannot arrive at a settlement within a year and a half, then the complaint is brought back before the Inspection Panel. Like the Panel, the DRS, which facilitates the dispute resolution process, honours requests for confidentiality from the requesters.

Given its features, the DRS offers a true problem-solving approach to the Parties. It provides affected persons a greater opportunity to have alleged harm remedied than the Bank’s Inspection Panel process. Affected persons also benefit from having an additional avenue of remedy through which their concerns can be heard and addressed by borrower States. The DRS therefore fulfils the function of the non-judicial grievance mechanism envisaged by Principle 29 of the UNGPs.

At the same time, the DRS should not limit the access to a remedy of affected persons through the Inspection Panel, and it is not a substitute for the compliance review process. Indeed, the executive directors have endorsed the view that the mandate of the DRS is to ‘enhance the effectiveness of the World Bank’s accountability system’, while being accessible and independent, as is the Panel.⁶⁴ The Accountability Mechanism Resolution and the Accountability Mechanism Operating Procedures should therefore ensure that the results of problem-solving are no less protective of requesters than the one offered by the Inspection Panel.⁶⁵

9.3 The Compliance of the DRS with the Panel’s Mandate and Best Practices

Section 9.2 identified the legal and policy standards against which the DRS must be evaluated – respectively, the 1993 mandate of the Inspection Panel and the 2020 mandate of the DRS, and best practices concerning the right of access to a remedy to be provided by accountability mechanisms. Section 9.3 proceeds to determine the compliance of the DRS with these mandates and best practices, and suggests three areas of improvement.

9.3.1 Accessibility: Eligibility Criterion, Choice of Representatives, and Access to Information

The first area of improvement relates to accessibility. As mentioned above, to access the DRS, requesters must meet all the eligibility criteria of the Inspection Panel.⁶⁶ Arguably, some criteria should apply to requests before both the Panel and the DRS, such as the requirement that a request must concern a Bank-funded project.

But others, such as the requirement that the harm has been caused by the Bank’s violation and not the borrower State’s, appear less relevant and may well reduce the accessibility of a remedy, as compared to the original mandate of the Inspection Panel. This is because one of the Panel’s eligibility criteria – i.e., showing a plausible causal link between the alleged harm and the project⁶⁷ – may become more challenging under the Bank’s new Framework, given that the responsibilities of the Bank are set out more clearly and narrowly therein than previously.⁶⁸ In addition, this eligibility criterion is coupled with a new feature in the eligibility determination phase, whereby Bank management can submit evidence of actual compliance or intent to comply, and requesters cannot access or respond to this evidence.⁶⁹ This lack of opportunity for procedural participation afforded to affected persons therefore also reduces their accessibility to a remedy.⁷⁰

In the context of the Panel, it makes sense to have as one of the eligibility criteria that the harm is caused by the Bank’s violation, because a compliance review investigation will be focussed on this issue. In the context of the DRS, however, this criterion appears unwarranted, because the goal of dispute resolution processes is problem-solving with borrower States. Whether the requesters suffered harm caused by non-compliance with Bank policies and procedures is typically a secondary consideration.⁷¹

While the criterion adopted by the Bank on the eligibility of complaints to the DRS is consistent with those of most (but not all) other multilateral development banks,⁷² questions arise as to whether it complies with the Bank’s commitment to increase the access to a remedy with the DRS.⁷³ In comparison, an approach that would increase the accessibility of the DRS would be to allow the Parties to proceed with dispute resolution if they both agreed to it, without requiring requesters to meet all the Panel’s eligibility criteria additionally.⁷⁴ In such a case, the consent of borrower States would act as a sufficient barrier to prevent a potential flood of complaints to the DRS and preserve the Panel’s central role in the Bank’s accountability system.

For these reasons, the Bank should consider abolishing the eligibility criterion of the DRS that requires that the harm must be caused by the Bank’s failure to comply with its policies. Given that this improvement concerns the Accountability Mechanism Resolution and Inspection Panel Resolution, and not the Accountability Mechanism Operating Procedures, they should be re-evaluated as part of the three-year review of the DRS.

Another proposed improvement regarding accessibility concerns the Parties’ choice of representatives and advisers. Paragraph 21.2 of the Accountability Mechanism Operating Procedures requires that the appointment or change of appointment of representatives be made in ‘consultation with the DRS’. Paragraph 21.2 states that the Parties can engage additional advisers but removes the requirement that this is only when ‘subject to no objection of the other Party’, as was present in the first two versions of the procedures. By initially requiring the Parties to agree on each other’s additional advisers, the procedures risked exacerbating power imbalances that already exist between the Parties.⁷⁵ For instance, borrower States could object to requesters retaining the services of certain civil society organisations as additional advisers, because these civil society organisations would have been critical of their human rights record in the past. This could pressure requesters to ‘bend’ to the demands of borrower States regarding additional advisers to avoid objections concerning their choice of additional advisers.⁷⁶

Against this backdrop, a study on the Compliance Advisor Ombud-sman (CAO), the accountability mechanism of the International Finance Corporation and the Multilateral Investment Guarantee Agency, has found that when non-governmental organisations assisted complainants with dispute resolution processes, the complaints were more likely to receive a remedy or to get to compliance review.⁷⁷ It also observed that ‘CAO’s decision to limit the participation of civil society organisations and legal representatives during negotiation and mediation engendered distrust among complainants and in some cases prompted their decision to withdraw from the [dispute resolution] process.’⁷⁸ This is sensible, because dispute resolution may not result in fair outcomes where there is inequality of power and resources between the Parties, who are often, on the one hand, local communities in developing countries, and on the other hand, State entities.⁷⁹ Therefore, removing from the final Accountability Mechanism Operating Procedures the consent of the other Party as a requirement for engaging additional advisers further protects requesters and is more in line with the DRS’ mandate. Despite this positive change, the Bank should also consider revising the Accountability Mechanism Operating Procedures so that they specify what type of advice the DRS staff can give to the Parties on the choice of their representatives, and specifies that either Party can request that its representatives and advisers be copied on all communications sent to it and be present in any discussion on the complaint.

The last improvement regarding accessibility concerns the access to project information. Paragraph 12 of the Accountability Mechanism Operating Procedures does not specify the powers the neutral third party would have regarding access to materials, documents, and testimonies related to the project, leaving this issue entirely to the Parties’ consent. Furthermore, paragraph 16 of the Accountability Mechanism Resolution provides that only the ‘Accountability Mechanism [will] have full access to project-related information in carrying out [its] functions.’ In contrast, the Inspection Panel receives all available project documentation from Bank management.⁸⁰ The result of these provisions is that the Parties engaged in the dispute resolution process could, in principle, agree that the requesters may access an amount of information that is (much) lower than that provided to the Panel. It is not an unlikely outcome, because often in practice, the concerns about a project lead to a complaint before Panel based precisely on a breakdown in the sharing of information or adequate consultation by the borrower States. Yet this situation would be problematic, because requesters can only access limited information on the project via the World Bank Policy on Access to Information⁸¹ to assert their rights and interests,⁸² and most of the project information is typically in the hands of borrower States. In this case, the opportunity for requesters to obtain meaningful remedies would be hampered by their lack of access to project-related materials, especially early in the dispute resolution process when requesters need the relevant project information to assess their position.⁸³

Against this backdrop, the Bank should consider including in the Accountability Mechanism Operating Procedures a minimum standard of information that must be shared with the requesters, or at least a commitment from the borrower State to share in good faith information necessary to ensure the orderly conduct of the dispute resolution process. This improvement would regulate the Parties’ agreement on access to information, by ensuring that access is at the very least not significantly lower in the dispute resolution process than it would be in the compliance review process. It would be in line with best practices, which opine that ‘[m]ember States have a legal duty to cooperate with [the] duly established [accountability] mechanisms.’⁸⁴ This improvement would also balance the concerns about protecting the effective access to a remedy of requesters with the potential encroachment of such measures on the sovereignty of the borrower States.

In conclusion, the DRS may enhance the accessibility of remedy by providing affected persons with an alternative to the Inspection Panel, in which they play a central role in designing remedial measures that address the harm caused to them by a Bank project. At the same time, revising the aforementioned eligibility criterion, most notably, would better empower the Bank to achieve this goal.

9.3.2 Effectiveness: Types of Complaint, Content of Agreements, and Verification of Implementation

The second area of improvement relates to effectiveness of the right to access a remedy. Under the Accountability Mechanism Resolution and the Accountability Mechanism Operating Procedures, complaints concerning serious human rights violations, such as those related to torture, may be brought to the dispute resolution process. Yet, the violation of some of these human rights, such as the prohibition of torture, have jus cogens status.⁸⁵ This means that they are fundamental principles of international law that must be upheld in all circumstances, and no one may ever derogate from them. International organisations like the Bank are bound by these prohibitions, as they themselves acknowledge.⁸⁶ The Bank has a responsibility under international law to end any violation of a jus cogens norm that it may enable. When complaints at the Bank concern violations of jus cogens norms, it is therefore doubtful whether continuing a Bank project according to its original terms, scope, and specifications for up to a year and a half while the dispute resolution process is underway complies with internationally recognised human rights.

In comparison to the DRS, at the Compliance Advisor Ombudsman, a case can be transferred to compliance appraisal in response to an internal request from the Compliance Advisor Ombudsman director general (i.e., the equivalent to the Accountability Mechanism Secretary), the president, the board, or management.⁸⁷ This request may be made when ‘concerns exist regarding particularly severe harm’.⁸⁸ However, such a possibility for internal requests does not exist at the Bank.⁸⁹ In fact, the DRS cuts the dialogic function with Bank management and the executive directors. According to paragraph 22.1 of the Accountability Mechanism Operating Procedures, management can only be an observer in the dispute resolution process with the Parties’ agreement, and is constrained to a technical role.⁹⁰ Yet, the practice shows that Bank management engagement has proven critical to resolving disputes effectively.⁹¹ Given its obligation to uphold jus cogens norms, the Bank should revise the Accountability Mechanism Operating Procedures and Accountability Mechanism Resolution to ensure that allegations of violation of these norms are investigated promptly by the Panel instead of moving forward with a dispute resolution process.

Another improvement concerns the content of dispute resolution agreements. According to paragraph 16 of the first version of the procedures, ‘Dispute Resolution Agreements should be consistent with World Bank policies and relevant domestic and international law.’⁹² This provision is in line with that of other international accountability mechanisms.⁹³ It only requires ascertaining whether agreements are ‘consistent’ (and not fully ‘compliant’) with Bank policies, and therefore does not call for conducting a process similar to a compliance review in parallel to the dispute resolution process. In fact, the Parties can even voluntarily agree to some deviations from the policies under this provision. As Professor Bradlow noted in his external review,

[t]his could happen, for example, if the complainants decide to accept less compensation than they may be entitled to under the policies because they believe that it is more useful to obtain certain compensation now rather than the possibility of more compensation in the future or they could agree to accept less compensation than the policies stipulate in return for access to other project benefits.⁹⁴

In contrast, the revised version of the provision, paragraph 23.1 of the Accountability Mechanism Operating Procedures, provides that, ‘[i]f the DRS has reason to believe that the Parties intend to include anything in a Dispute Resolution Agreement that is inconsistent with relevant domestic or international law, the [Accountability Mechanism] Secretary will request the Parties to make appropriate modifications.’ This provision makes two significant changes as compared to its previous iteration. First, it removes the requirement of consistency of dispute resolution agreements with Bank policies. Yet, under the Articles of Agreement⁹⁵ and the Inspection Panel Resolution,⁹⁶ the executive directors have an institutional responsibility to ensure the Bank’s observance of its operational policies and procedures, an international legal obligation the Bank has no power to modify unilaterally, let alone relinquish. It is therefore doubtful that the Bank would comply with its international obligation should any agreement reached through the dispute resolution process it established be inconsistent with the policies. Second, the provision shifts from an objective requirement of consistency of dispute resolution agreements with domestic and international law, to a subjective requirement that the DRS doubts such consistency. It therefore waters down an obligation of result into an obligation of means, without imposing any burden of investigation on the DRS to absolve itself of this obligation. This change significantly weakens the protection of affected persons.

More broadly, given the inequality of power and resources between the Parties, the procedural protections afforded – or rather, not afforded – to the requesters that were examined in the previous subsection 9.3.1 are all the more important to ensure that requesters do not feel pressured to agree to a remedy that is substantially less than the one to which they are entitled under Bank policies and that would normally be assessed by the Inspection Panel. As a United Nations report noted, ‘in many situations, complainants may legitimately feel that partial redress is their only feasible option.’⁹⁷ According to best practices, the DRS must ensure at least that its ‘outcomes and remedies accord with internationally recognised human rights’.⁹⁸ Therefore, the Bank should revert to a provision similar to paragraph 16 of the Interim Operating Procedures, which requires consistency with Bank policies.

The last improvement regarding effectiveness concerns the verification of the implementation of the Parties’ agreement. While the Accountability Mechanism Resolution states that the Parties should agree on a ‘time-bound implementation schedule for agreed actions’,⁹⁹ it is silent on how compliance with this implementation is monitored. Paragraph 24.1 of the Accountability Mechanism Operating Procedures adds that the DRS will monitor implementation subject to the Parties’ agreement. It is therefore allowed for the Parties to agree to a relatively weak provision on implementation, whereby compliance with the agreement and the agreed remedial actions are not effectively monitored.

This provision is compatible with the 1993 mandate of the Inspection Panel, because the Panel was not originally granted monitoring powers. In the three decades following the inception of the Panel, intense public scrutiny was often helpful for the actual implementation of an agreement to occur.¹⁰⁰ Best practices have since evolved in parallel to the point where it has become widely accepted that the effectiveness of a dispute resolution process depends on the implementation of agreed remedial actions being monitored.¹⁰¹ This is because monitoring dispute resolution agreements has proven to be a key factor in ensuring that affected persons achieve a remedy. For instance, as part of a complaint before the accountability mechanism of the Inter-American Development Bank, the monitoring of the dispute resolution agreement signed between the Haitian government and local farmers has shown that implementation remained partial and has proposed solutions to live up to the commitment made of restoring the livelihoods of displaced farmers.¹⁰² Against this background, the executive directors have allowed the Inspection Panel in recent years to monitor compliance on a case-by-case basis.¹⁰³ The Inspection Panel Resolution went a step further: it required verification by management, and in some specific cases by the Inspection Panel and the Bank Audit Unit, of the action plan’s implementation.¹⁰⁴

Given that the mandate of the Inspection Panel is to provide affected persons with basic access to a remedy, while the mandate of the DRS is to provide them an additional path of access, it is unclear why management (and the Panel) now have monitoring authority with regards to compliance review, but the DRS does not have such authority with regards to dispute resolution. In fact, all international accountability mechanisms currently have monitoring authority regarding dispute resolution, except for the DRS.¹⁰⁵ For example, the Compliance Advisor Ombudsman will monitor the implementation of the Parties’ agreement, and a complaint will be transferred to the compliance review process if the Parties fail to implement this agreement.¹⁰⁶ As noted in the external review, failing to ensure that agreements are implemented may have ‘adverse reputational consequences’ for the Bank.¹⁰⁷ The Bank should therefore consider revising the DRS to require monitoring of implementation. Since this change is significant, it may be best addressed through the three-year review of the DRS in the Accountability Mechanism Resolution.

In short, whether the DRS strengthens or weakens the effectiveness of the right to access a remedy depends on whether the Parties agree to a remedy that is superior, equal, or inferior to the one mandated by Bank policies. In a few cases, affected persons and borrower States may arrive at a win-win agreement, where their respective interests align, and no compromise is needed. But it seems unlikely that in all complaints the borrower State will agree to a remedy that significantly advantages affected persons,¹⁰⁸ given that affected persons will only have brought their complaint before the Panel after their efforts to resolve it with the borrower State and management have already failed.¹⁰⁹ Moreover, the ‘worst-case scenario’ of a failed dispute resolution process for the borrower State is that the complaint will move forward with the compliance review process, whereby the remedy provided would be no more and no less than the one prescribed by Bank policies. In these cases, the only disadvantage for the borrower State is that it will have to go through a lengthy and public investigation.

Affected persons, on the other hand, continue to experience the harm caused by the Bank project while the dispute resolution process and the compliance review process are ongoing, and therefore are incentivised to agree to some form of remedy quickly. In this context, it is all the more important that significant procedural protections ensure that affected persons do not feel pressured to agree to a remedy substantially less than the one to which they are entitled under Bank policies.

9.3.3 Independence: Panel Mandate, Staff Involvement, and Party Funding

The third area of improvement relates to independence and impartiality. As mentioned above, the DRS is independent of Bank management and the Inspection Panel. The Panel ‘will not opine on policy compliance in dispute resolution or the outcome of the dispute resolution process’.¹¹⁰ This firewall between the structure of the two mechanisms is warranted to avoid conflicts of interest, ensure that each mechanism performs its functions independently, and enable the Parties to fully engage in the dispute resolution process without fearing that the information divulged as part of it can be used in the compliance review process.

The DRS is intended to complement, not substitute, the compliance review process. In the Inspection Panel Resolution, ‘[t]he Executive Directors reaffirm[ed] the importance of the Panel’s function, its independence and integrity’.¹¹¹ In practice, however, the mandate of the DRS may infringe on the mandate of the Inspection Panel. To take one example noted by commentators, a Party agreement reached through the dispute resolution process would ‘forestall any Inspection Panel review or investigation of the matter and prevent any members of the affected community, who otherwise feel that their concerns were not addressed in the process … to request a new investigation’.¹¹² This is because the complaint on that project will be considered closed by the Panel, unless there is new evidence or circumstances unknown at the time of the request.¹¹³ Thus, the result of the dispute resolution process will prevent the Inspection Panel from carrying out its role of investigating compliance with Bank policies.

To address the potentially conflicting mandates of the dispute resolution and compliance review processes generally, scholars and civil society organisations have advocated that multilateral development banks like the Bank should provide more options for sequencing these processes. At most banks today, requesters typically have two options: they can either resort to dispute resolution first and then move on to compliance review if they are dissatisfied with the former, or go straight ahead with compliance review but thereby relinquish the possibility of dispute resolution.¹¹⁴ Scholars and civil society organisations suggest that affected persons should be able to choose which process to undertake first and to change to the other one once, or to pursue both processes simultaneously.¹¹⁵ They argue that compliance review can provide information and analysis to affected persons to which they might not otherwise have access in dispute resolution given their power imbalance vis-à-vis the borrower States; conversely, dispute resolution can highlight systemic issues relevant to compliance review that might not have become apparent without dialogue between the Parties.¹¹⁶ At the United Nations Development Programme (UNDP) for example, affected persons can choose to go through compliance review¹¹⁷ and dispute resolution¹¹⁸ simultaneously. This shows that the concern, according to which allowing Parties to use compliance review irrespective of the outcome of the dispute resolution process would disincentivise borrower States from fully participating in the dispute resolution process,¹¹⁹ may be overblown.

Another improvement regarding independence and impartiality concerns the relationship between the DRS staff and the Parties. Paragraph 14.1 of the Accountability Mechanism Operating Procedures states that ‘[t]he DRS is impartial as between Parties and as to the merits of the dispute.’ However, the Accountability Mechanism Secretary and the DRS staff are also significantly involved in the dispute resolution process. This involvement raises the question of whether they are perceived as independent of the Parties. As mentioned, paragraph 21.2 of the Accountability Mechanism Operating Procedures requires that the Parties ‘consult’ with the DRS staff regarding the choice of their representatives, which must be voluntary. These provisions imply that the DRS staff must determine whether this choice is in reality ‘voluntary’. Meanwhile, neither the Accountability Mechanism Resolution nor the Accountability Mechanism Operating Procedures set limits on the content and means of communication to the Parties, which raises questions as to the extent of the DRS staff’s influence in the Parties’ decisions. For example, would the DRS staff give its opinion to the requesters on the quality of representation that different civil society organisations may offer them? Would it advise on the relation that the requesters could have with their representatives regarding the management of their complaint? The Accountability Mechanism Resolution and Accountability Mechanism Operating Procedures are silent on these issues. Rather, the Accountability Mechanism Operating Procedures should only require that the Parties ‘inform’ the DRS staff about their choice of representation.

The DRS staff is also involved in the very decision of the Parties to pursue the dispute resolution process. Paragraph 11.3 of the Accountability Mechanism Operating Procedures puts forward that ‘[i]f either of the Parties indicate, or the DRS assesses, a need for capacity building to allow them to better make an informed decision on whether to participate in a dispute resolution process, this may be offered by DRS within the resources and time frame available.’ Under paragraph 21.4 of the Accountability Mechanism Operating Procedures, the Parties must bear the costs of their representation and advice during the dispute resolution process. Since requesters have fewer resources than borrower States, they are more likely to ask for, or be assessed as needing, this advice and capacity building. Although the requesters may benefit from this opportunity, the concern is that by treating them differently than it does borrower States, the DRS may be perceived as lacking independence and impartiality.¹²⁰

The Bank should therefore consider addressing, through institutional changes, the general tension between accessibility and independence at the DRS. The World Trade Organization (WTO) is an example of how an international organisation successfully managed this tension. On the one hand, the secretariat, as the WTO administering institution, ‘assist[s] panels, especially on the legal, historical and procedural aspects of the matters dealt with, and … provide[s] secretarial and technical support’.¹²¹ In parallel, the WTO Advisory Centre is a separate and independent institution that offers free advice and training on WTO dispute settlement proceedings to developing countries.¹²² Because the secretariat cannot provide such assistance to less well-off States without risking its independence, this separate entity was established.¹²³ In contrast, the DRS plays the role of both the administering institution and advisory/training institution. This dual role in turn may jeopardise the perceived independence of the DRS. Further, by confirming that the choice of representatives is voluntary, or by offering guidance on disagreement as to the scope of the dispute resolution process between the Parties,¹²⁴ the DRS staff may also play a role typically reserved for third-party neutrals.

To address these concerns at the DRS, and more broadly increase the accessibility and effectiveness of access to a remedy, a pragmatic approach may be for the Bank to provide funding to affected persons to get support from professionals during the dispute resolution process. A recent UN report has suggested a range of funding mechanisms that international accountability mechanisms could set up to do so, which includes stand-alone remedy funds, escrow accounts, trust funds, insurance schemes, guarantees, and letters of credit.¹²⁵ Scholars and civil society organisations have long called for establishing such funds at the World Bank and other multilateral development banks, because civil society organisations currently supporting requesters in dispute resolution processes, free of charge, do not have the budget to assist most of them.¹²⁶ The argument is that, as part of the development mandate of multilateral development banks, they should reserve part of the project budget to fund potential complaints launched by affected persons, who are typically vulnerable populations. Canada, for instance, sets aside a small portion of the total project budget of its large infrastructural projects to help minorities in areas covered by these projects to express their concerns about them.¹²⁷

In sum, with its mandate of independence and impartiality, the DRS offers affected persons access to a neutral third party to access a remedy. As designed, however, it may infringe on the mandate of the Inspection Panel, and the involvement of the DRS staff in dispute resolution processes raises some concerns about its appearance of independence.

9.4 Conclusion

The Kawaala community and Uganda are now attempting to resolve the complaint concerning the Lubigi channel project amicably through the dispute resolution process offered by one of the Bank’s avenues for a remedy, the DRS. At the time of writing, the Parties had asked and were granted the additional six months to pursue the dispute resolution process.¹²⁸ Only time will tell whether this new avenue will enhance the right of access to a remedy of the Kawaala community and all other requesters participating in the dispute resolution process, as the Bank sought to do by establishing the DRS.

This chapter has shown that, in the meantime, different aspects of the DRS raise the concerns of whether dispute resolution will actually enhance the right of access to a remedy, and whether it may instead prejudice the Inspection Panel’s mandate to provide this right of access. Given these concerns, the chapter has set out three areas of improvement that the Bank could consider which, if adopted, would empower the DRS to better realise its mandate.

10 IMF Surveillance as a Non-Compliance Mechanism

Ambroise Fahrner

The Fund shall oversee … the compliance of each member with its obligations¹

10.1 Introduction

This chapter evaluates the emergence and development of “surveillance” as the preferred non-compliance mechanism within the IMF architecture. This study highlights the specific role of international law within the field of international monetary relations, as well as illustrating how international monetary relations provide international law with original new tools and concepts.

John M. Keynes famously stated with respect to the creation of an international institution in charge of handling the international monetary system – what became the International Monetary Fund – that “(t)he most difficult question to determine is how much to decide by rule, and how much to leave to discretion”.² For lawyers, this often-quoted sentence evokes the systemic need to combine rules of conduct, aiming at ordering States’ behaviors, with adequate legal structures, aiming at ordering these rules of conduct and guaranteeing their efficiency, such as mechanisms of adjudication.³ This chapter starts from this premise that there is a specific and systemic need to efficiently link “command” to “compliance” in the field of global monetary governance. This is of a particular and concrete relevance as it is often held that the defining features of international monetary law are that it is soft by nature, that its institutional framework is loose by design, and its dispute settlement system mainly informal and political.⁴ The chapter will therefore investigate the types of non-compliance options available in international monetary law which aim at the constitution of the international monetary system.

Surveillance is an original mechanism designed to ensure the conformity of countries’ behavior to the obligations set under the IMF legal regime, concerning exchange rates policies, broadly speaking.⁵ Through surveillance, the IMF monitors the international monetary system, world economic health, and also the economic policies of its member countries (to the extent that they influence international monetary conditions). IMF surveillance enables the international monetary system to achieve its purposes of sustaining monetary and financial stability, as well as promoting sound economic growth by facilitating the exchange of goods, services, and capital among countries, including by monitoring compliance with exchange rate obligations.⁶ This is achieved at two complementary levels: “bilateral surveillance,” bringing together, on a regular basis, the IMF and a given country; and “multilateral surveillance,” which provides an annual analysis of the international monetary system and global economic forecasts to the international community.⁷ The Fund also advises countries about the necessary policy adjustments to be made to prevent potential crises.

A paradox lies nonetheless in Article IV of the IMF Articles of Agreement which enables surveillance: “The Fund shall oversee … the compliance of each member with its obligations.” On one hand, obligations undertaken by States under the IMF Articles of Agreement should be obeyed, as they are conventional obligations.⁸ Reference to “compliance” in Article IV should therefore come as no surprise. On the other hand, the IMF must only “oversee” such “compliance” and it is usually assumed that this is the reason why the track record with respect to the non-compliance of States with their IMF legal obligations is said to be unsatisfactory.⁹ And each past financial crisis has prompted heated debate among both the public and policy experts about how the international monetary system could strengthen existing mechanisms to monitor and forecast global economic and monetary developments and enforce States’ obligations in this respect.¹⁰ That surveillance is the preferred compliance mechanism in the international monetary system emphasizes the lack of actual jurisdictional venue in international monetary relations.

This chapter engages with the hypothesis that the more broadly a legal interest is shared among States, the less desirable it is that a compliance procedure should bring about a particular result; more relevant is some ownership of the process.¹¹ Global monetary and economic stability is, by definition, a broad objective. At the same time as constituting a direct interest for every State, it also constitutes a community interest.¹² This contribution addresses both the extent to which the surveillance mechanism set up by the IMF ensures the compliance of its members with its code of conduct and the extent to which the hypothesis above is verified in the monetary field. The success of IMF surveillance can be explained according to this hypothesis, because the IMF’s surveillance involves broad flexibility in a way international adjudication does not. That IMF surveillance by essence is in the realm of flexibility does not mean, however, that the process is without rules. The fact that surveillance is meant to allow discretion for States and the IMF to achieve relevant objectives does not mean that it is a process that is unlegalized, more political or floats in a vacuum. It has developed procedural rules of its own. We will focus hereafter on its procedural characteristics. The fact that the IMF process is sustained universally and with regularity, that it is often exposed to the changes in the economic landscape, and that it is regularly reviewed, means that it is in a constant process of refining these procedural rules.

Nonetheless, “surveillance” remains a strange word in the realm of legal notions. It sounds familiar to the lawyer’s ears as it conveys a sense of discipline. However, it also sounds odd as it does not clearly express how it differs from functions, such as adjudication by international courts or political decision-making processes used to settle a disagreement. As a sui generis concept under IMF law, it was never explicitly defined, and evolved constantly. Interestingly, the word is also used in the World Trade Organization (WTO) or Organisation for Economic Co-operation (OECD) legal regimes.¹³ IMF surveillance has never been the object of major doctrinal interest, as is the case also for international monetary law generally.¹⁴

The success of IMF surveillance will be examined in three ways. Firstly, it will be shown that surveillance appears to be the most successful mechanism to enforce international monetary obligations thanks to its broad flexibility and original mechanism. The nature and scope of surveillance, as well as the factors explaining its success, will be assessed. IMF surveillance contrasts positively with alternatives. International courts outside the IMF or political dispute settlement systems inside the IMF indeed offer limited options to settle States’ disagreements with respect to their monetary obligations under the IMF Articles of Agreement. Finally, the chapter will underline how the legal dynamics of surveillance have provided States and the IMF with a dynamic and complete set of procedural rules addressing the process of surveillance as a transparent, rule-of-law inspired, and sophisticated procedure.

10.2 The Success of Surveillance as the Primary Compliance Mechanism in International Monetary Law

Explaining the relative success of IMF surveillance as the main non-compliance mechanism in global monetary governance¹⁵ requires analyzing its essential features and how they constitute assets for the task of assessing international obligations regarding global monetary governance.

10.2.1 Nature of IMF Surveillance

10.2.1.1 Legal Basis of IMF Surveillance

Today, the legal basis for IMF surveillance is primarily rooted in Article IV, section 3(a) and (b) of the IMF Articles of Agreement and complemented by three Executive Board Decisions from 1977,¹⁶ 2007,¹⁷ and 2012.¹⁸ It also draws inspiration from the 2015 Guidance Note for Surveillance under Article IV and its 2021 Supplement. With respect to its function, surveillance requires members to provide relevant and accurate information about the conduct of their policies, not only on the basis of Article IV, section 3(b), which directly addresses surveillance, but also of Article VIII, section 5. Surveillance at a basic level has three faces: bilateral surveillance – which is led on the basis of Article IV consultations; multilateral surveillance, published twice a year in two reports: the “World Economic Outlook Report” and the “Global Financial Stability Report”; and regional surveillance, when the IMF considers, for instance, the EU or the Euro area.

Surveillance’s legal basis is also derived from the purposes of the IMF objectives as stated in Article I of the Articles of Agreement, and from a broader obligation to cooperate.¹⁹ The objectives of the IMF are expressly stated in Article I of the Articles of Agreement, in terms almost unchanged since its adoption. The first of its objectives stated in Article I is “to promote international monetary cooperation through a permanent institution which provides the machinery for consultation and collaboration on international monetary problems”.²⁰ Surveillance must be interpreted as contributing to fulfilling this objective. Historically, it was after the collapse of the Bretton Woods system of fixed (and adjustable) exchange rates in 1971 that the surveillance function was incorporated into the Articles of Agreement through the Second Amendment.

As previously stated, the IMF enjoys a range of options to “remind” its members of their obligations as stated in its Articles of Agreement and made precise in various Executive Board Decisions.²¹ Furthermore, it is expected that the nature of the IMF as an international organization informs us about the contours of this duty. In this perspective, it must be highlighted that the IMF Articles of Agreement truly constitute the world monetary constitution and, as such, create obligations of a more far-reaching kind than any other conventional monetary regime. They provide a “code of conduct” (the articles of the IMF Articles of Agreement, which refer to members’ obligations to follow a given policy) concerned with exchange rates, international transfers, and liquidity assistance to its members.

In addition, IMF members also are under a general obligation to collaborate with the Fund to manage a stable system of exchange rates and exchange arrangements.²² This duty is extended through a specific duty to consult with the Fund when requested and to provide relevant information. Against this backdrop, it is noteworthy that these obligations of collaboration, consultation, and then provision of information apply to all three core IMF functions: financial assistance,²³ technical assistance,²⁴ and surveillance²⁵ – surveillance being the monitoring of the members’ compliance with their obligations under the IMF Articles of Agreement. The IMF’s general functioning is thus inspired by these various duties to transparently collaborate, consult, and provide information.

10.2.1.2 Scope of Surveillance

Article IV not only sets out procedural rules as to the conduct of surveillance but also covers substantial obligations concerning exchange rate arrangements and the policies of IMF members. It is complemented by and must be read in conjunction with a set of Executive Board Decisions establishing principles that guide members.²⁶ Yet the exact material scope of surveillance requires a careful and dynamic analysis. Article IV’s history is illuminating in this respect. Whereas it was initially dedicated to the management of a system of fixed exchange rates, the demise of this system in the 1970s led to a more diffuse set of rules of conduct, allowing all types of exchange rate arrangements, including floating arrangements. Therefore, the ratione materiae jurisdiction of the field of surveillance must now be understood in a more dynamic way. There is an expectation that members will pursue policies favoring a “stable system of exchange rates” and “avoid manipulating exchange rates”.²⁷ The material scope of surveillance relates not only to exchange rates, but also to other domestic or external economic and monetary policy “to the extent that they significantly influence present or prospective external stability”.²⁸

Obligations related to exchange rates are expressed in stronger language than those related to other fields of domestic or external policy, reflecting the dominant post-war consensus over the importance at that time of fostering effective collaboration in the field of exchange rates, while preserving the domestic policy space of members of an unprecedented international organization.²⁹ However, IMF surveillance focussed in the past on a wide range of sectors. For instance, it delivered prescriptions over members’ financial sector policies despite the lack of a formal mandate relating to the international financial system.³⁰ Surveillance was also used to assess the capital flow policy of its members, despite the right of members embodied in the IMF Articles of Agreement to regulate this field according to their preference and without IMF involvement.³¹ More surprisingly, as it does not seem to be directly linked to IMF goals, surveillance also addressed non-economic policies such as environmental, labor, military, or institutional policies to the extent that they influence external stability.

The IMF considers obligations related to exchange rates to constitute obligations of result as opposed to the obligations of conduct, that prevails with respect to domestic or external policy. This dichotomy should not be interpreted as distinguishing between optional and mandatory obligations, or between political and fully legalized norms. Indeed, the asymmetry rather lies in the distinction between the obligation to attempt to achieve a given result, and the obligation to achieve the desired outcome. In the first case, it goes without saying that a breach of obligation can be found. At the same time, contrary to what is usually assumed, nothing stands in the way of finding a breach of an obligation of conduct. However, the breach will not be directly concerned with the unachieved goal, but with a failure in the means employed by the member State. The fact that the nature of members’ obligations varies according to the material domain of the policy at stake constitutes a challenge for efficient IMF leadership.

10.2.1.3 Procedural Stages

Article IV consultations typically start with the annual IMF mission visit to each member State. They aim to gather relevant data for the purpose of updating analyses of members’ current economic situation. They can include meetings with State officials, but also more surprisingly “other stakeholders such as parliamentarians, and representatives of business, labor unions and civil society”.³² This is followed by an “assessment” by the mission of the member’s economic situation. The mission then engages in a further round of discussions to address the actual or prospective efficiency of the implemented policies. The mission closes its intervention by sending a report containing its preliminary findings, and then a final report to the Executive Board. The Executive Board discusses the findings of the final report sent by the mission and agrees with the State on specific conclusions. The Chairman of the Board provides then a Summing Up of the discussion which is formally addressed to the member’s authority. This terminates the consultation phase. After consultations close, the IMF offers to publish the report. This only occurs if the State consents to it. Interestingly, whereas States’ consent is typically verified at the jurisdictional phase in front of international courts, surveillance as a process is mandatory for both the IMF and States. Consent is expressed at the end of the process and relates only to publication of the result.

10.2.2 Factors Explaining IMF Surveillance Process: Scope, Normativity, and Authority

The dominant narrative identifies “surveillance” as one of the core functions of the IMF in the international monetary system, possessing characteristics that can be contrasted with formal dispute settlement mechanisms.³³ Its contours are blurred and its influence mainly political. As such, it is said to be fitted for the needs of international monetary governance. As a result, surveillance has mainly been analyzed so far through an economics lens, stressing the importance of transparency, peer pressure, or the relevant sectors to be monitored as factors explaining its relative success as an enforcement mechanism.³⁴ Legal considerations have been underemphasized so that its precise legal contours remain in the shadow of institutional practice. What are the legal features that explain the use of surveillance as a preferred method to achieve members’ compliance with their obligations under IMF law?

Surveillance appears to be a practical procedure to monitor the global economy from a holistic perspective. This all-encompassing approach is increasingly needed in today’s economy, in which international monetary and economic spillovers play a significant role. The initial Bretton Woods institutional set-up indeed conceived of global economic governance as best organized through a three-pronged scheme, dividing monetary, trade-related, and development-related issues, and allocating their management to three corresponding universal international organizations. Nonetheless, today’s globalized economy differs greatly from the post-war era and the legal framework must mirror these paradigmatic shifts, including the substantial inter-linkages between the monetary, trade, and development fields. Global economic governance also increasingly requires not only a public international law understanding, but also awareness of key international economic issues and international political dynamics. Surveillance has been able to adapt to these key economic changes thanks to an evolution of its ratione materiae jurisdiction. IMF surveillance’s ratione materiae jurisdiction is built around a dynamic principle: “Other policies will be examined in the context of surveillance only to the extent that they significantly influence present or prospective external stability.”³⁵ Though this has been possible only at the cost of an asymmetry in terms of bindingness as between exchange rate-related obligations and obligations related to “other policies”, this evolution was key to maintaining the IMF’s relevance in global economic governance.

Whereas financial and technical assistance offered by the IMF are voluntary in nature, surveillance is mandatory and universal. As previously explained, the IMF is legally obliged to conduct surveillance proceedings and the member is legally obliged to participate and to conduct itself in certain ways during the proceedings. Nonetheless, surveillance does not aim at producing mandatory decisions of the nature an international court would. It also does not necessitate the existence of a “dispute”. The political cost of engaging in a review of a member’s obligation is therefore significantly lower than in an international court’s typical proceedings. Regularity of the process also contributes to assuring a review of obligations, without triggering the political or economic costs a “dispute” would.

Though non-binding, IMF surveillance nevertheless offers an authoritative assessment of a member’s compliance. The IMF cannot oblige members to implement assessments resulting from surveillance procedures. It can only persuade them. Persuasion can be direct and result from the intrinsic quality of the IMF assessment. A report’s authority can also flow less directly from the reputational effects it creates, triggering peer-pressure mechanisms or market pressure. Markets or peers are given access to the IMF reports and will immediately reflect the IMF analysis in their own economic analysis. Non-compliance with IMF reports will lead to higher borrowing prices, lower foreign direct investment, and, generally speaking, lower trust in the country’s word and ability to successfully manage its economy. In a nutshell, the procedure does not encompass rules obliging members to comply with the IMF assessment. This does not, however, mean that it does not possess any effective authority. Reputational effects fulfill a non-compliance sanction role.

The IMF surveillance procedure navigates between an initial confidential stage, and then a process of making the result of the procedure transparent. During the consultations, the procedure requires confidentiality and informality. It allows for more space to build trust between the stakeholders. Importantly, it avoids sending adverse signals to markets, which could trigger by anticipation precisely the effects that the process aims at preventing. Some information might also be confidential. The publication of the IMF view is then critical to obtain the peer pressure and market signal effects on which compliance is based. While consent remains the rule, publication might also be decided upon under certain specific circumstances which require immediate action.³⁶ The balance between other members’ interest in a stable international monetary environment and a member’s right to confidentiality in the management of its balance-of-payments depends upon the circumstances and the adverse effects that publication might cause. And the determination of such circumstances will be made according to a political process ending with a voting procedure.

10.3 Weaknesses of Non-Compliance Alternatives to IMF Surveillance

We now turn to the existing alternatives to the IMF in order to contrast their respective features. We will focus on both internal and external alternatives.

Firstly, international courts have not played a significant role in global monetary governance. Matters relating to compliance with IMF obligations have not been brought to general international courts, or even international arbitration proceedings, in which specialists on monetary issues could have been appointed. This does not appear to be a matter of the skills of the judicial bodies with respect to specialized issues. The existence of an all-encompassing and conventionalized sub-regime explains the situation more satisfactorily. Secondly, within the IMF, options implying a withdrawal, a loss of rights by IMF members, an interpretation of the Articles of Agreement, or an amendment have appeared too political.

10.3.1 The Limited Recourse to International Courts in Global Monetary Governance

The International Court of Justice (ICJ) (or its predecessor, the PCIJ) has, in the past, seized the opportunity to clarify significant general international monetary law issues, such as the customary international law principle of monetary sovereignty.³⁷ The Court has also shown its ability to deal with monetary issues in a case involving a discriminatory system of license control in respect of imports not involving an official allocation of currency.³⁸ However, the fact is that the ICJ has not played a regular role in solving disputes between States in relation to international monetary relations. It is generally assumed that the ICJ, as a non-specialized court, is not the most appropriate venue to deal with international economic law issues and that such issues should be scrutinized by specialized arbitral bodies like arbitral investment tribunals or the Dispute Settlement Body of the WTO. Yet, after a closer look, it appears that the explanation is not to be found in the often-alleged distinction between specialized and non-specialized issues, and the resulting ability of the respective judicial body to handle the issue at stake.³⁹

A more explanatory distinction is the one between topics extensively covered by international conventions on one side, and residual, general, and systemic issues on the other side. Monetary rules are generally embodied in complex treaties, sometimes forming entire sub-regimes, and having their own dispute settlement systems, be it political, institutional, or to a certain extent judicial. Recourse has mainly been had to the ICJ to clarify architectural issues of international law external to or underpinning these treaties and regimes, but not specialized and conventionalized matters. The recent involvement of the ICJ in matters pertaining to monetary flows is only incidental and is a by-product of the embargo put in place by the United States against Iran in a broader geopolitical context.⁴⁰

Neither has the IMF had recourse to ICJ Advisory Opinions. The IMF Articles of Agreement entered into force on December 27, 1945, two months after the UN Charter, and the IMF and the UN entered into a relationship agreement on November 15, 1947. From this perspective, it will be of no surprise that the “Agreement between the United Nations and the International Monetary Fund” offers the IMF the possibility to “request advisory opinions of the International Court of Justice on any legal question arising within the scope of the Fund’s activities”. It was expected that this option could prove useful where general international law became relevant in the activities of the IMF, such as State succession. In practice, the IMF has never seized the opportunity to request an Advisory Opinion from the ICJ and has handled matters of general international law independently.⁴¹ Two points could be deduced. Firstly, the lack of expertise of the IMF in general international law has not prompted the need to resort to general international courts. Secondly, the fact that ICJ Advisory Opinions avoid the confrontational aspect of judgments has not led the IMF to resort to such opinions.

In the search for a dispute settlement alternative fitted for international monetary issues, it appears that the WTO dispute settlement panels or specialized arbitral tribunals could play such a role, as they incorporate technical expertise that the ICJ allegedly lacks in the field of economic law. In practice, the WTO and international investment tribunals have determined a certain number of disputes with a monetary component so far.⁴² This apparent success comes at a cost. Monetary issues are only litigated insofar as they pose trade or investment-related issues.⁴³ As with the ICJ, these specialized bodies have proven capable of addressing global monetary issues, albeit from a generalist economic legal perspective rather than a specialist monetary and financial perspective. However, once again, it is not the alleged lack of capability that explains non-recourse to international investment tribunals or the WTO dispute settlement system, but more particularly the fact that monetary issues are specific and submitted to a comprehensive conventional regime.

10.3.2 The Limitations of Recourses to IMF Dispute Settlement Options other than Surveillance

The IMF possesses a range of compliance mechanisms, though many of these are political in nature and seldom explicitly identified as compliance mechanisms. Disputes as to monetary issues can firstly be solved via a political or institutional change; and where a solution is difficult to reach, there are sanctions available, such as the forced withdrawal of a State from the IMF, or the suspension of a member’s benefits. Although these options are not meant to be used on a regular basis⁴⁴ to solve members’ disagreements, they offer ultimate institutional and political mechanisms to settle a dispute between members. As will be discussed, IMF members have alternatively been known to amend the IMF’s constitutional charter itself, the IMF Articles of Agreement, in order to settle a dispute. There is also available a sophisticated authoritative interpretation mechanism, though this is in practice no longer used. The IMF has mainly resorted to non-authoritative interpretation.

10.3.2.1 Options Involving Forced Withdrawal or Suspension of Rights

The IMF Articles of Agreement provide for a procedure to force a member to withdraw as per Article XXVI, section 2(c). The only case to date is that of Czechoslovakia in 1954.⁴⁵ It is also noteworthy that Article XXIX of the IMF Articles of Agreement provides for the constitution of an “Arbitral Tribunal” to assist with the settlement of potential disputes arising from the withdrawal of a member. It goes without saying that this mechanism is only designed as a last-resort option.

A softer version of this approach consists of suspending a member’s financial⁴⁶ or political⁴⁷ rights. These approaches are not, however, perfectly fitted for the compliance task as non-compliant States are precisely those requiring financial and political rights to address financial needs or to build political legitimacy in the context of complex reforms. Depriving members of the resources designed for them when they need them the most might not always be the best dispute settlement or compliance option.

10.3.2.2 Options Involving an Amendment of the Articles of Agreement of the IMF

The IMF Articles of Agreement also provide for an Amendment procedure embodied in Article XXVIII. The Second Amendment of 1978⁴⁸ testifies in a spectacular way that the inability of a member like the United States to comply with core IMF obligations concerning exchange rates can be settled through this means, providing a workaround so that US policy is not qualified as a breach of IMF obligations. The United States announced on August 15, 1971, that it would suspend the convertibility of the dollar into gold.⁴⁹ The collapse of the par-value system that resulted was only legalized thereafter by the Second Amendment.

Amendments have also from time to time contributed to strengthening rules aimed at creating discipline in the international monetary system. The First Amendment (1969)⁵⁰ improved the rules relating to an authoritative procedure for interpretation by the introduction of a Committee for Interpretation. The Third Amendment (1992)⁵¹ introduced stricter rules for the suspension of voting rights of members who failed to repay the IMF. However, the amendment procedure requires a very constraining process and is therefore not fitted to settle most disagreements between States. Indeed, and it is quite unique in the world of international organizations, the procedure requires a majority of at least three-fifths of the members holding 85 per cent of IMF voting rights.⁵² Compared with such radical and constraining dispute settlement options, “interpretation” offers a much more nuanced approach.

10.3.2.3 Options Involving Interpretation of the Articles of Agreement of the IMF

The IMF is the first international organization to be vested with the jurisdictional power to interpret its constitutive act, to formally “hear” and discuss complaints of its members through this means. Article XXIX(a) of the IMF Articles of Agreement provides that “any question of interpretation of the provisions of this Agreement between any member and the Fund or between members of the Fund shall be submitted to the Executive Board for its decision”. This mechanism additionally sets out an appeal to the Board of Governors, which is assisted by a Committee on Interpretation of the Board of Governors. Whereas the first phase of the procedure employs the usual weighted voting procedure, the second phase of the procedure, the appeal, is conducted following its own voting system, according to which each member has one equal vote.⁵³ Article XXIX(b) of the IMF Articles of Agreement refers to the Decision of the Board of Governors as being “final” with respect to questions of interpretation. The mechanism under Article XXIX has been used ten times with only one appeal. The most recent use of the procedure ended in 1959.⁵⁴

Interpretation could therefore have constituted a preferred, if unorthodox, means of dispute settlement within the IMF framework. It has indeed the advantage of bringing a potential solution to a situation without directly highlighting a member’s misconduct. This mechanism significantly evolved over time. Formal interpretations were useful in the first years of the life of the IMF, to clarify elements of such a new legal regime in a world not yet accustomed to international organizations of this kind. However, whereas in the first years of the IMF its formal interpretation mechanism was used on a regular basis, it has since disappeared to the benefit of regular interpretations made outside the formal IMF framework. Informal interpretations are adopted, leaving open the possibility of authoritative interpretation for settling any future or ongoing difficulty with respect to a previous interpretation. Both regular and authoritative interpretations are binding as they reflect the decision-making process of the Fund and members are obliged to comply by virtue of the obligation to collaborate with the Fund. But regular interpretations are not final because the possibility of a subsequent authoritative interpretation remains open. This two-tiered approach caters to IMF reluctance to issue formal interpretations that would inevitably tie its hands for the future and considerably reduce its space for discretionary measures.

10.4 The Dynamics of Procedural Rules Related to the Process of Surveillance

“Surveillance” has evolved over time, under the pressure of crises and internal reviews, or in order to adapt to joint surveillance exercises with other international institutions. One defining theoretical issue is the extent to which the IMF can draw from general principles used by international courts to develop its framework while preserving the unique and defining features that constitute IMF surveillance.

10.4.1 Dynamics of Procedural Rules: Crises, Internal Reviews, and “Joint Surveillance” with Other Institutions

Surveillance has become the main IMF instrument to oversee IMF members’ compliance with their obligations. As opposed to other IMF or non-IMF existing dispute settlement methods, surveillance offers a non-compliance mechanism in large part based upon flexibility. Despite the flexibility that characterizes both the substantial and procedural aspects of its enforcement, surveillance seems to achieve compliance in an indirect way. That flexibility is the key to this mechanism does not mean, however, that it has not evolved over time or been provided with procedural rules.

Two types of factor have shaped the development of surveillance. Firstly, structure has followed substance as surveillance has had to adapt to changes in the global economy. The most spectacular case is that of financial crises. As the etymology suggests, “crises” differ from simple “difficulties” in that their nature and gravity require a change of a systemic nature. More gradual changes in the structure of the global economy have also prompted substantial changes in the IMF surveillance process. Secondly, surveillance procedures have also evolved as a result of scrutiny, through the IMF internal schemes of evaluation, or through confrontation with other international institutional fora.

In 1998, the IMF undertook an ex-post analysis of the causes of the Asian economic crisis and of its management by the IMF services. Previously, the Mexican crisis had also left its footprint on surveillance’s procedures. The inability of the IMF to anticipate the Mexican crisis of 1993–1994 caused the IMF to review surveillance modalities. As a result, IMF surveillance procedures now encompass a stronger focus on sensitive matters, better internal coordination (information of the Executive Board), a broader scope of analysis (inclusion of data non-formally provided by the member), and more regular contact with officials of the member countries.⁵⁵ The 2008 global economic crisis prompted a paradigmatic shift as it introduced a more intensive recourse to joint surveillance with other international fora, such as the G20 through the Mutual Assessment Process (MAP)⁵⁶ or the Financial Stability Board through the IMF-FSB Early Warning Exercise.⁵⁷

The IMF periodically reviews its activities to better adapt to changes in the global economy, which includes surveillance in all its aspects. This reviewing process has been driven by the need to face the weaknesses unveiled by various crises and the ambition to anticipate future developments in the world economy. The 2012 Integrated Surveillance Decision, the 2014 Triennial Surveillance Review, the 2018 Interim Surveillance Review and the ongoing Comprehensive Surveillance Review (CSR)⁵⁸ have identified substantial weaknesses in the monitoring of the global economy, but also improved the modalities of the conduct of the surveillance process. The IMF also has an Independent Evaluation Office (IEO), established by the Executive Board in 2001. It is functionally independent of the IMF and establishes its own agenda. It has access to all relevant data. The IEO regularly issues reports containing surveillance-related advice.⁵⁹

There has been some debate arising from the fact that the World Bank and the IMF’s respective jurisdiction overlap from time to time with respect to States’ borrowing. Overlapping conditions attached to external aid have been described as “cross-conditionality”. Similar issues can be highlighted, which we will call “cross-surveillance” issues by analogy since existing terms do not fully describe the phenomenon. The Financial Sector Assessment Program (FSAP), for instance, is a joint World Bank–IMF program.⁶⁰ It functions very much like typical IMF surveillance activities, but its legal status is that of technical assistance. Previously operating on a voluntary basis, it is now mandatory for twenty-nine jurisdictions selected because of their systemically important financial sector. The G20 Mutual Assessment Process (MAP) involves the IMF and similar jurisdictions in a similar exercise. The legal basis is that of technical assistance, in conjunction with the legal framework of the IMF’s ability to engage in joint activities. It is assumed that such collaborations have prompted discussion as to the methods and procedures employed by the IMF and other institutions. On this basis we can say that this joint exercise of surveillance has also contributed to an evolution in IMF procedures in exposing surveillance to the test of efficient collaboration.

10.4.2 Dynamics of Procedural Principles: Borrowing from General Principles?

Surveillance is a very demanding process: it is very broad in scope, it operates on a regular basis, and it is universal. This requires a sound and solid procedural framework to assure its legitimacy and efficiency. Its sui generis nature, the evolution of the world economy and the recurrence of crises have prompted regular review processes and resulted in densification of the rules relating to the procedural aspects of surveillance. It is submitted that the development of procedural rules fitted for surveillance can be viewed as reconciling two contradictory dynamics. Firstly, the rules draw inspiration from the general procedural principles used by international courts. Secondly, however, they do not properly borrow from this vocabulary, as they seek to underline the specificity of surveillance as a specific process. For instance, IMF surveillance must be conducted as a “dialogue,”⁶¹ but it is not referred to as an “adversarial process.” The procedure should be held in all “candor,”⁶² “frankness”⁶³ and “openness,”⁶⁴ but the IMF does not precisely refer to the general principle of “good faith” as such. The IMF assessment must be “persuasive”⁶⁵ and “clear,”⁶⁶ but the wording of “reasoning” is not used.⁶⁷

Yet, overall, IMF surveillance procedural rules are expressed in terms that evoke the general procedural principles of international litigation. This could imply that surveillance is considered as the exercise of a “quasi-international jurisdiction” to which general procedural principles of international litigation apply. At the same time the IMF’s emphasis on using a different set of terms to those employed in respect of proceedings in international courts and tribunals must be emphasized. The matter may become important in the future development of the IMF legal framework.