To date, proposals for the reform of WTO dispute settlement have focused on institutional improvements of the AB, as represented by the Walker proposal, or possible replacement of the two-tier adjudication with one-tier adjudication by a standing body of full-time panelists, as suggested by Hoekman and Marvroidis (“To AB or Not to AB: Dispute Settlement in WTO Reform” 23 JIEL 703 (2020)). But the successful use of Article 25 arbitration for appellate review in Turkey–Pharmaceutical Products has raised a third possibility: instead of serving merely as a stop-gap measure, Article 25 arbitration could be structured for WTO appellate review on a permanent basis. Significantly, this third possibility appears to have the explicit support of the United States. As reported in this Blog, the US said: “If a Member supports dispute settlement reform, then a bilateral arrangement presents a unique opportunity to explore alternative approaches.” This positive attitude of the US is in stark contrast with its general silence on other reform proposals.
Compared to the other two proposals, Article 25 arbitration has some potential advantages. First, it may bring WTO appellate review back closer to its original design. As is well known, the Uruguay Round negotiators established an appellate review mechanism to ensure that “fundamentally flawed” panel reports would not be adopted automatically under the negative consensus rule of the DSU. With this goal in mind, the negotiators did not expect that recourse to appellate review would be frequent. But subsequent WTO practice has proven otherwise. Because appeal is part of the compulsory adjudication, the losing party at the panel stage would be tempted to exhaust available remedies. The expectation that a panel decision would be subject to the scrutiny of the AB led to a tendency toward overlitigation, resulting in substantial delays in dispute settlement. In contrast with the compulsory adjudication by the AB, Article 25 arbitration is voluntary and consensual in nature. In a system of voluntary dispute settlement, parties respond to incentives different from those in a system of compulsory adjudication. They are more likely to settle differences through negotiation than litigating every issue before the tribunal. Consequently, under DSU Article 25, resort to appellate review may become less frequent and resolution of disputes more expeditious.
Second, Article 25 arbitration is inherently more flexible than a standing AB. Such institutional flexibility is needed to ensure that major legal errors be corrected rather than cemented into WTO jurisprudence. Judges are fallible humans and must be allowed to make mistakes. The key is for the system to have a mechanism to correct mistakes. Unfortunately, such a mechanism does not exist at the WTO. In theory, the Members may overrule the AB by adopting an authoritative interpretation of WTO provisions (per WTO Agreement IX:2). But in practice, this is unlikely ever to happen. As a result, the task of rectifying mistakes fell onto the AB itself. However, for whatever reason, the AB in its 25-year existence never acknowledged making any error, let alone providing self-correction. Instead, the AB insisted on treating its rulings as de facto precedent. Such insistence may have eroded the flexibility necessary to maintain a balance between law and political constituencies of the system. In contrast with the standing appellate body, Article 25 arbitration will be conducted by arbitrators on an ad hoc basis. The arbitrators will not have the same power and authority as the AB. Consequently, it would be possible for one panel of arbitrators to disagree with another on the same issue, or for a WTO panel in a subsequent case to disagree with the appeal arbitrators in a prior case. While such disagreement might create some inconsistencies in case law, over time the better reasoning could emerge and be accepted as the proper precedent.
Thirdly, compared to the other proposals, adopting Article 25 proceedings for appellate review would require fewer changes to the existing system. Substantively, all that would be required for amending the DSU would be to delete the single provision on the Appellate Body (Article 17), while the panel proceedings and the negative consensus rule on the adoption of panel reports would all be kept intact. Technically, it would be necessary also to delete all other references to the Appellate Body in the DSU, and to adjust Article 16.4 (the adoption of panel reports) so as to clarify that “appeal” would be available via Article 25. This adjustment could be made by adding a simple footnote to Article 16.4, so the provision would read as follows:
- Within 60 days after the date of circulation of a panel report to the Members, the report shall be adopted at a DSB meeting unless a party to the dispute formally notifies the DSB of its decision to appeal* or the DSB decides by consensus not to adopt the report. If a party has notified its decision to appeal*, the report by the panel shall not be considered for adoption by the DSB until after completion of the appeal*. This adoption procedure is without prejudice to the right of Members to express their views on a panel report.
* The appeal of a panel report may be made via arbitration under DSU Article 25.
With this amendment, it would not matter whether a particular Member, such as the United States, would choose ever to use Article 25 arbitration itself, since any panel report not submitted to appeal arbitration would be automatically adopted by the DSB. Given the minimum changes required to amend the DSU, it could be easier for the 164 Members to reach consensus, which is a legal requisite for any DSU amendment (per Article X:8 of the WTO Agreement).
On the whole, Article 25 arbitration seems a possible compromise between the reinstitution of the AB, which the US has resisted, and a new system of one-tier adjudication by permanent panelists, which may not satisfy many other Members. There are, however, major institutional issues to be addressed. One is the pool of potential arbitrators. In principle, the disputing parties have control over the selection of arbitrators. In Turkey-Pharmaceuticals, the parties decided that the three arbitrators should be picked through a random selection from a list of former AB members and ten MPIA arbitrators. If Article 25 appeal arbitration were to become “institutionalized”, how should the pool of arbitrators be determined? Another issue is the role of the WTO Secretariat. In Turkey–Pharmaceuticals, the arbitrators were apparently assisted by the Secretariat, but presumably not by the same staff that had assisted the Panel in this case. Recall that the MPIA participants requested that the WTO provide separate secretarial support to the MPIA arbitrators so as to ensure the quality and independence of the proceedings, but the US opposed that request as it perceived the MPIA to be an attempt to create a substitute AB. Again, if Article 25 appeal arbitration were to become “institutionalized”, how should the arbitrators be assisted by the WTO Secretariat? In this regard, the analysis and suggestions by Joost and his coauthor in their recent AJIL article deserve special attention.
