The New Way of Working at the WTO: Looking at ECA

Abstract

The growing inability of the World Trade Organization (WTO) to conclude new multilateral agreements has prompted groups of like-minded Members to experiment with alternative forms of trade cooperation. The Agreement on Electronic Commerce (ECA), concluded in 2024 but not incorporated into Annex 4 of the Marrakesh Agreement, exemplifies this evolution. Negotiated by WTO Members, within the WTO, and with the stated objective of eventual integration into the multilateral framework, the ECA nevertheless currently operates as an autonomous treaty governed by public international law. This article argues that it represents a new way of working within the WTO and illustrates the emergence of what the authors describe as an “imperfect agreement”: a legally binding instrument that remains institutionally connected to the WTO while operating outside its formal legal architecture. The article first examines the legal status of the ECA and the institutional consequences of its implementation outside the WTO framework, including its interim governance arrangements and autonomous dispute settlement mechanism. It then analyses how the agreement manages its interaction with WTO law by preserving the rights of non-participating Members, limiting potential conflicts with existing WTO obligations, and maintaining a credible pathway towards future incorporation into the WTO legal system.

1. Introduction

The emergence of new issues such as digital trade, data governance, environmental sustainability, supply chain resilience, and the multi-crisis pattern that the world has been facing in the last 6 years has exposed the growing difficulties for governments to adopt new rules. Multilateral negotiations no longer allow to quickly respond to new needs and adapt to new realities. Nonetheless, the World Trade Organization (WTO) remains a forum of the future, but governments need to work differently (Marceau 2020).

Although the WTO was designed to be a permanent multilateral negotiating forum, in the face of these multiple crises, groups of like-minded Members are seeking alternative negotiating formats. They first sought to negotiate formal Annex 4 plurilateral agreements as part of Joint Statement initiatives, then moved on to informal (open or non-open) plurilateral agreements, mini-deals, targeted trade deals, and many other imperfect agreements through which flexible cooperation mechanisms are implemented for like-minded Members (Marceau 2026; Klausen 2021; Nakatomi 2013; Pagani 2006; Zampetti et al. 2022)1. On the one hand, these initiatives are always second-best and exploratory options to the more formal and traditional WTO approach. On the other hand, they allow regulatory innovation in areas where multilateral negotiations remain blocked (Hoekman and Sabel 2021), and develop state practices that could eventually lead to more formal amendments or codifications (Marceau and Marquet 2019).

Although groups of willing Members may negotiate new disciplines among themselves, the incorporation of the resulting agreements into the WTO framework generally remains subject to voting or consensus decision-making (Boklan and Bahri 2022). Recent developments illustrate the practical implications of this institutional constraint. At the 14th Ministerial Conference (MC), India alone, despite broad support among WTO Membership, opposed the incorporation of the Investment Facilitation for Development Agreement; the United States (US) resisted the adoption of ministerial declarations relating to agriculture; and Brazil blocked the renewal of the moratorium on customs duties on electronic transmissions. These examples demonstrate how the consensus rule allows a single Member to block the formalization of negotiated outcomes despite significant political support. More fundamentally, the function of consensus has evolved. It no longer necessarily reflects substantive opposition to the content of a proposed rule. Increasingly, Members use their ability to block decisions as a strategic instrument to secure concessions in unrelated areas of negotiation (Van den Bossche 2025). Members are therefore constantly seeking flexibility to overcome the obstacle of consensus2.

The recently concluded Agreement on Electronic Commerce (ECA) provides one of the most significant examples of this phenomenon. Originating from the Joint Statement Initiative launched at MC11 in Buenos Aires in 20173, with negotiations formally commencing in 2019, the process culminated in the conclusion of an agreement on 26 July 20244. However, participating Members were unable to secure their incorporation into the WTO legal framework through the addition of the agreement to Annex 4 of the Marrakesh Agreement due to a lack of consensus5. At MC14, participating Members decided to proceed with its implementation through an interim arrangement pending future incorporation into the WTO framework6.

Notwithstanding, using the ECA as a case study, this article suggests that, announced by Director-General (DG) Ngozi at MC14, the ECA reflects a new way of working for WTO Members and argues that such imperfect agreements constitute a viable transitory pathway for addressing emerging issues in international economic law. Nevertheless, this calls for a certain degree of caution in defining the legal status of the ECA and in determining the implications of its imperfect nature. The fact that the ECA remains “outside” the WTO legal framework does not preclude potential normative interactions, both with respect to non-participating Members and in relations among participating Members themselves, which must be given appropriate scrutiny. The article argues that imperfect agreements can provide a pragmatic response to negotiating deadlock and facilitate regulatory innovation. Their long-term viability, however, depends on their ability to preserve a credible pathway towards eventual integration within the multilateral trading system.

2. The Rise in Imperfect Agreements

The decision to proceed outside the traditional WTO framework has both legal and institutional implications. It is interesting to question the legal status of the ECA and whether it can be viewed (at least until its full integration in the WTO) as an autonomous treaty governed by international law. Further, questions remain on the “transition” from an imperfect (WTO) agreement to a fully integrated WTO Agreement. Attention must be given to the institutional arrangements established to ensure the functioning of the agreement and to the continuing efforts of participating Members to secure its eventual integration into the WTO framework.

2.1. The Legal and Institutional Imperfections of the ECA

After several years of negotiations, ECA participating Members announced the substantial conclusion of the Agreement on Electronic Commerce in July 2024. However, the conclusion of the agreement did not resolve the issue of its place within the WTO legal architecture (Marceau and Nandakumar 2025). In February 2025, participating Members sought to incorporate the agreement into Annex 4 of the Marrakesh Agreement, thereby transforming it into a plurilateral agreement formally integrated within the WTO framework. This proposal failed to obtain the consensus as required under Article X:9 of the WTO Agreement, reflecting persistent opposition by several Members who questioned both the procedural legitimacy of this pathway and the substantive implications of the ECA7. However, for participating Members, waiting for consensus among the entire WTO membership risked leaving a significant regulatory vacuum in one of the fastest-growing sectors of the global economy. Rather than abandoning this project, participating Members opted to proceed in what some considered to be “outside” the WTO legal framework. From the perspective of general international law, there is nothing to prevent countries from signing and implementing treaties. That said, the situation is obviously far from ideal. The ECA nevertheless occupies an unusual position within the international legal order. Unlike traditional WTO agreements, it is not formally fully part of the WTO acquis. Yet neither is it entirely detached from the WTO. ECA was negotiated exclusively among WTO Members, in the WTO building; it carries a WTO symbol and is, in this regard, an official WTO document; it designates the WTO DG as the depositary of the instruments of acceptance and seeks explicitly to become part of the WTO framework8.

Notwithstanding its uncertain relationship with the WTO legal framework, the ECA possesses all the characteristics of a treaty governed by international law. The ECA is open to acceptance by all WTO Members, including those that did not participate during the negotiations9, and provides that it shall be registered with the United Nations (UN) Secretariat in accordance with Article 102 of the UN Charter10. WTO Members wishing to become parties must complete their respective domestic constitutional procedures and deposit an instrument of acceptance11. The ECA remains a “hybrid agreement” in the sense that it currently stands somewhat “outside” the WTO’s traditional legal framework until its complete incorporation. It has to be distinguished from the provisional application of certain international agreements under Article 25 of the Vienna Convention on the Law of Treaties (VCLT)12. The ECA does not provide for any form of provisional implementation. Instead, it follows a conventional treaty model whereby it will enter into force thirty days after the deposit of the forty-fifth instrument of acceptance13. Once in force, the ECA creates binding legal obligations under international law for those Members that have accepted it. These obligations vary in nature. Some provisions contain precise and enforceable commitments, such as the prohibition of customs duties on electronic transmissions14, while others establish softer forms of cooperation15, encourage the adoption of best practices, or promote regulatory convergence in areas such as consumer protection16, personal data protection17, or unsolicited commercial electronic messages18. Yet the existence of such obligations in the international legal order does not necessarily imply that such provisions will produce corresponding effects within WTO law. The latter (i.e., WTO law) remains a distinct legal regime whose rights and obligations are determined by the covered agreements and enforced through the Dispute Settlement Understanding (DSU).

By opting for a formal treaty subject to ratification and entry into force requirements, participating Members chose one of the most institutionally demanding pathways available. Alternative approaches were possible. For example, willing Members could have proceeded through the implementation of coordinated domestic measures, informal cooperative arrangements, joint declarations, or non-binding guidelines aimed at addressing emerging trade issues without creating new treaty obligations (Marceau 2026). The ECA reflects a deliberate choice in favor of legal formalization, even in the absence of incorporation into the WTO framework. The ECA thus illustrates how willing Members may continue to develop new trade disciplines even in the absence of multilateral approval.

2.2. Interim Arrangements and the Continued Search for WTO Integration

The conclusion of the ECA outside the traditional WTO legal framework creates a series of institutional challenges that would not arise had the agreement been incorporated into Annex 4 of the Marrakesh Agreement. In this respect, the ECA occupies an intermediate position between a fully autonomous international treaty and an agreement integrated within the WTO system—it is a hybrid agreement.

The most immediate consequence of the agreement’s external status concerns dispute settlement. Because the ECA does not form part of the WTO “covered agreements”, it is not automatically subject to the DSU to resolve disputes arising under its provisions. WTO Members could have chosen to argue that ECA disputes could lead to a specific narrow mandate under Article 7 of the DSU, but they did not envisage such an approach. Instead, they decided that ECA would have its fully autonomous and ad hoc DSM19. To promote efficiency and predictability, this mechanism largely reproduces, mutatis mutandis, the procedures and principles developed under the WTO DSU20. The ECA also incorporates a temporary appeal arbitration mechanism inspired by the Multi-Party Interim Appeal Arbitration Arrangement (MPIA), thereby seeking to preserve the consistency and legitimacy of dispute settlement despite the continuing paralysis of the WTO Appellate Body (AB)21. The result is a system that remains heavily influenced by WTO practice while formally operating outside the formal WTO legal order.

More complex questions arise with respect to the institutional support that may be provided by the WTO Secretariat. While Members that are parties to an Annex 4 agreement may establish a specialized committee within the WTO and assign to it functions they deem appropriate22, the ECA establishes its own governance structures, including a Council on Electronic Commerce responsible for supervising the operation of the agreement23. Yet the extent to which WTO facilities and resources may be used to support these activities remains uncertain. WTO Members can reserve rooms in the WTO building so it seems clean that meetings of ECA bodies may be hosted on WTO premises at the request of WTO Members; the extent to which the Secretariat may provide administrative and logistical assistance, or whether WTO staff may contribute to the implementation and monitoring of the agreement, will depend on the WTO DG who is responsible for the management of the staff and their duties24. Similar uncertainties concern the provision of technical expertise. The WTO Secretariat possesses unique expertise in the collection of trade data, the preparation of economic analyses, the monitoring of trade measures, and the administration of dispute settlement proceedings. Participating Members may naturally wish to draw upon these resources in order to facilitate the operation of the ECA. Such involvement could raise institutional and political concerns if support for the ECA were perceived as affecting the allocation of resources available for multilateral activities. It seems clear that the DG is entitled to instruct WTO staff to assist WTO Members who also happen to be ECA members; the payment of fees of panelists working for the panel and MPIA process out of the regular budget may trigger resistance in the Committee on Budget, Finance and Administration. It may very well be that the panelists and MPIA members dealing with ECA disputes will have to be paid by the parties or the ECA members themselves.

Although the precise contours of WTO support remain uncertain, the DG has already demonstrated a willingness to assist participating Members. Further, the DG has agreed to act as depositary for instruments of acceptance and has facilitated the circulation of certified copies of the agreement to all WTO delegations. The new way of working and the need to adapt to the new multilateralism were also noted. More broadly, the DG has repeatedly emphasized the importance of the initiatives on e-commerce and the need to support Members seeking to develop rules for the digital economy25.

For this reason, participating Members continue to pursue incorporation of the ECA into Annex 4 of the Marrakesh Agreement. Such a development would clarify ECA’s legal status and facilitate the provision of institutional support by the WTO Secretariat. If this happens and once it takes effect, the institutional arrangements will cease to apply, and dispute settlement functions will be entrusted to the DSU to ensure the unity of the WTO’s legal framework. More broadly, it would reconnect ECA formally with the WTO Agreement from which it originated and would reinforce its long-term legitimacy.

3. Managing the Interactions Between the ECA and the WTO Legal Framework

The viability of an imperfect agreement ultimately depends on its ability to coexist with the WTO legal framework from which it emerged. In the case of the ECA, the agreement was negotiated by WTO Members, concerns issues closely connected to existing WTO disciplines and aspires to eventual integration into the WTO framework. Consequently, its success depends not only on the commitments made by participating Members but also on the careful management of its relationship with WTO law. The ECA addresses this challenge through a combination of substantive accommodation and institutional arrangements designed to preserve the rights of non-participating Members (A) and prevent conflicts with existing WTO obligations (B).

3.1. Preserving the Rights of Non-Participating Members

From the perspective of general international law, the starting point is Article 34 of the VCLT, which codifies the well-established principle pacta tertiis nec nocent nec prosunt. Under this principle, a treaty neither creates obligations nor confers rights upon third States without their consent. Applied to the ECA, this principle means that non-participating WTO Members cannot be bound by the disciplines established under the agreement and cannot be compelled to implement its commitments. The ECA expressly reaffirms this principle by providing that it creates neither rights nor obligations for WTO Members that are not parties to it26.

Additionally, Article 41 VCLT recognizes that parties to a multilateral treaty may agree to modify that treaty inter se. Such modifications are permissible where the treaty itself provides for that possibility or, in the absence of such provisions, where the modification is not prohibited by the treaty and does not affect the enjoyment by other parties of their rights under the treaty, impede the performance of their obligations, or undermine the object and purpose of the treaty as a whole. The relationship between Article 41 VCLT and WTO law remains subject to debate. In Peru—Agricultural Products, the AB observed that WTO law contains specific rules governing certain forms of inter se modifications, notably through the disciplines applicable to Regional Trade Agreements (RTAs), and as such Article 41 VCLT is of no relevance in WTO law27. However, one could argue that such WTO-specific rules do not exhaust all possibilities for modifying legal relations among subsets of WTO Members because the AB’s reasoning was limited to the cases of RTAs (Mathis 2016). The central question therefore remains whether such agreements affect the rights enjoyed by non-participating Members under the WTO Agreements. The ECA seeks to address this concern through a provision stipulating that “nothing in this Agreement shall be construed as diminishing a Party’s rights and obligations under the WTO Agreement” (see Footnote 26). Since ECA deals mostly with new disciplines on new markets, it will often be difficult for ECA non-participating members to demonstrate that ECA nullified benefits from old General Agreement on Tariffs and Trade (GATT) provisions. The Arbitration in Australia—India Domestic Regulation also showed how difficult it is for a WTO Member to demonstrate that additional General Agreement on Trade in Services (GATS) transparency and other commitments by another Member can nullify prior benefits and rights28.

Among the rights, the most frequently invoked is probably the most-favored-nation (MFN) principle. The relationship between the ECA and MFN obligations is, however, more complex than it may initially appear. The difficulty lies in identifying which ECA commitments fall within the scope of this principle. The ECA contains a wide range of disciplines, from market-access commitments and trade facilitation measures to regulatory cooperation. The most obvious situation relates to the commitment not to impose customs duties on electronic transmissions. To the extent that this commitment concerns market access conditions covered by Article I GATT or Article II GATS, participating Members seeking to ensure the compatibility of the ECA with their WTO obligations should implement it consistently with the MFN principle. A participating Member cannot, for example, exempt electronic books originating in another ECA party from customs duties while starting to impose such duties on identical electronic books originating in a non-participating WTO Member. In such a circumstance, the non-participating Member could challenge the measure before the WTO dispute settlement system on the ground that an “advantage” granted to one WTO Member had not been extended immediately and unconditionally to all other Members.

It should be noted that the opposite position was adopted by another group of thirty-three Members that announced their intention to maintain a moratorium on customs duties on electronic transmissions only among themselves29. In a way, this is reminiscent of Robert Hudec’s analysis of Section 30130, in which he argued that the necessary reforms would not be implemented without the driving force of justified disobedience on the part of the US31. While this Hudec statement may appear at first sight to raise MFN concerns, its compatibility with WTO law is not necessarily excluded. Three considerations are particularly relevant. The first one refers to a legal issue: some could argue that what is covered by the moratorium decision was never covered by prior GATT/GATS rules, so MFN does not necessarily find easy application. The second refers to a technological obstacle; it remains uncertain whether customs duties on electronic transmissions could be implemented in a technically feasible manner by Members, even if they are willing to do so32. The third reason is purely practical: the administrative burden and costs associated with maintaining a differentiated customs regime for non-participating Members would likely be considerable enough to discourage Members (Vanderbruggen 2026).

Similar considerations may apply to several ECA disciplines relating to trade facilitation, such as paperless trading, electronic contracts, electronic invoicing, or electronic authentication. Even if some of these measures could theoretically be implemented on a preferential basis, the administrative costs associated with maintaining distinct regulatory regimes constitute constraints or burdens which will significantly reduce the incentives for Members to depart from MFN treatment and create an incentive to extend the benefits of ECA to all trading partners. As a result, such provisions are likely to operate, in practice, on an MFN basis, regardless of whether WTO law strictly requires that outcome.

Some ECA commitments may certainly be implemented preferentially, particularly those concerning regulatory cooperation, information sharing, and equivalence arrangements. This issue is especially important for personal data protection. For instance, some Members permit cross-border data flows only to jurisdictions that provide an equivalent level of protection. As ECA parties undertake to negotiate mutual recognition of equivalence, it remains to be seen whether non-ECA parties can claim that it is also entitled to similar mutual recognition since their data protection regime is objectively equivalent—as no Member has ever challenged a mutual recognition arrangement as inconsistent with MFN. This uncertainty illustrates the broader difficulty of defining the limits of the MFN principle in an increasingly regulatory and digital trading environment.

Preserving the rights of non-participating Members addresses only one dimension of the relationship between the ECA and WTO law. Another challenge concerns the legal relations among participating Members themselves. While the ECA seeks to avoid prejudice to third parties, it must also coexist with a legal framework that already governs many aspects of electronic commerce and international trade more broadly.

3.2. Avoiding Conflicts with WTO Obligations Inter Partes

This issue is best understood through the broader phenomenon of fragmentation in international law (Koskenniemi and Leino 2002). The multiplication of specialized treaty regimes has increased the possibility that different legal instruments may regulate the same factual situation from distinct perspectives33. Such tensions are particularly significant in the WTO context because the WTO dispute settlement system occupies a unique institutional position (Marceau 1999, 2001, 2002). Article 23 of the DSU grants WTO Members exclusive recourse to the WTO dispute settlement mechanism for the preservation of rights and obligations under the covered agreements, while Articles 1.1, 3.2, 7, 11 and 19 of the DSU limit the jurisdiction of WTO adjudicators to those agreements. As a result, WTO adjudicating bodies cannot directly enforce obligations arising under agreements such as the ECA.

Against this background, two categories of normative tensions may arise. The first scenario concerns measures that are consistent with the ECA but incompatible with WTO law. Such situations would arise where the implementation of an ECA commitment results in conduct that violates obligations contained in the WTO Agreements. In that case, the existence of an ECA flexibility or right or even obligation could arguably constitute a defense or a justification to violate WTO law if it is connected to a WTO justification (Article XIX, XVIII, XX and XXI) (Marceau and Tomazos 2008). WTO adjudicating bodies have consistently maintained that Members cannot justify violations of WTO obligations by invoking commitments undertaken in another international agreement unless it can be considered a fact or an agreement to develop a WTO flexibility34. But most ECAs improve the transparency and flow of digital trade for the eventual benefits of all Members. The second scenario concerns measures that are compatible with WTO law but incompatible with the ECA. This situation may arise where the ECA establishes obligations that go beyond those contained in WTO Agreements. A Member could therefore comply fully with its WTO obligations while nevertheless violating commitments undertaken under the ECA. In such circumstances, WTO law would offer no remedy because the dispute would concern obligations falling outside the covered agreements. Resolution would therefore depend entirely on the DSM established under the ECA itself.

This raises a subsidiary issue which is whether a violation of ECA and noncompliance with ECA ruling can authorize retaliation or counter-measures in the form of WTO inconsistent measures. This issue arises with retaliation under regional trade agreements: can a party to a regional trade agreement (RTA) retaliate against another RTA party that does not comply with an RTA ruling in imposing non-MFN tariffs above those consolidated at the WTO? There is no clear response yet to this question. Could an ECA member simply invoke authorization under the ECA to suspend WTO obligations or impose trade restrictions that would otherwise violate WTO law? It is unlikely that the Dispute Settlement Body will issue a favorable ruling, since there is “no basis in the DSU for panels and the AB to adjudicate non-WTO disputes”35 and that agreements may not diminish the rights and obligations of Members arising from WTO agreements36. These situations illustrate that the effectiveness of the ECA dispute settlement mechanism ultimately depends on its ability to operate alongside the multilateral legal framework.

The solution is not without limits. Like any treaty, the ECA may eventually give rise to interpretative disagreements concerning the precise scope of its commitments or their interaction with WTO rules. Nevertheless, the risks of normative conflict are mitigated by Article 3.2 of the ECA, which governs the relations inter partes as well. In this respect, Article 3.2 operates as a rule of systemic coordination between the ECA and WTO law. Rather than presenting the agreement as a self-contained regime capable of displacing existing WTO obligations, the provision requires ECA commitments to be interpreted, to the greatest extent possible, in a manner consistent with the covered agreements. This interpretative presumption significantly reduces the likelihood of conflicts of norms by favoring constructions that preserve the continued application of WTO rights and obligations. As a result, where several interpretations of an ECA provision are available, preference should be given to the interpretation that maintains consistency with WTO law.

4. Conclusions

The conclusion of the ECA marks an important development in the evolution of international trade governance. Faced with the inability of the WTO membership to reach consensus on new rules governing the digital economy, a group of Members chose to pursue a different path. The resulting agreement illustrates the emergence of what this article has described as an imperfect agreement, an instrument negotiated by WTO Members, addressing issues closely connected to WTO law, yet operating outside the formal framework of the Organization. The ECA demonstrates that such agreements can offer a pragmatic response to negotiating deadlock. By allowing willing Members to move forward without waiting for multilateral consensus, they create opportunities to develop new disciplines in areas where regulatory needs evolve more rapidly than multilateral negotiations. In this respect, imperfect agreements may contribute to preserving the relevance of international trade law in the face of emerging challenges.

At the same time, the experience of the ECA highlights the limitations inherent in this approach. It has become apparent that several of the concerns we have raised remain abstract for the time being and will depend to a large extent on Members’ state practices, i.e., how the ECA is implemented and how it is received by Members who are not participating in it. The WTO has always been viewed as a Member-driven organization, where governments’ practices sometimes take precedence over strict legality. While such approaches may seem questionable or even raise concerns, they should not frighten us unduly. Legality in international law is the cause of a contemporary complex for us international lawyers. Out of fear that our discipline will be negated, we forget that it was originally built upon the practice of States, which still occupies a central place in treaty law and may even eventually form a subsequent customary rule. The agreement’s effectiveness depends not only on the commitments undertaken by its parties but also on careful management of its relationship with WTO law. Its provisions concerning third-party rights, the preservation of existing WTO obligations, and the establishment of interim institutional arrangements all reflect an awareness that regulatory innovation needs to be managed within the coherence of the multilateral trading system. Perhaps the most significant lesson is that imperfect agreements are not conceived as substitutes for nor alternatives to multilateralism. WTO remains the best forum for multilateral actions, but WTO Members need to accept working differently, as is done with ECA, which constitutes only one example of this “new way of working”.

The ECA therefore suggests that imperfect agreements may provide a means of keeping international trade regulation moving forward while preserving the possibility of future multilateral convergence in a period marked by persistent institutional deadlock. This is WTO reform in action.