he dispute settlement system of the World Trade Organization (WTO) — long held to be the crown jewel of the multilateral trading system — is in crisis, potentially endangering the future of the WTO. The United States has been blocking new appointments to the Appellate Body (AB), which plays a key role within the WTO’s compulsory third-party adjudication process, as the terms of sitting members expired. The United States’ official justification is its dissatisfaction with the AB’s performance. At least three AB members are needed to consider an appeal of a dispute panel report, but at the time of writing, the number of AB members stands at one, making the appeals function of the WTO impossible to use. As a result, there is no longer a multilateral forum to hear new appeals. For the moment, dispute panels continue to be established, suggesting that WTO members retain confidence in the dispute settlement system overall, even if it does not include a functioning AB. However, there is uncertainty about how these disputes will be resolved if the panel’s recommendations are appealed. Some WTO members are seeking to self-insure against this risk by developing alternative appeal mechanisms. A prominent example is an EU-Canada initiative to have panel reports heard by an ad hoc appellate process.1 Such initiatives are patchy solutions at best. They risk creating a multi-tier system across WTO members, as some will participate in an appellate process, and some will not. This is unlikely to result in an internally coherent jurisprudence, the raison d’être of any appellate process. While most WTO members oppose the US decision to block new appointments to the AB,2 a recent survey of WTO delegations and practitioners3 reveals that the United States is not alone in having concerns about the performance of WTO adjudicating bodies.4 While dispute settlement may be the crown jewel of the WTO, it has imperfections. This is neither surprising nor contested. The problem is that the WTO membership collectively has been unable — and unwilling — to make timely repairs, and thus allowed the jewel to crack.
Because the Uruguay Round’s WTO Dispute Settlement Understanding (DSU) was a major innovation for the trading system — creating an appellate function and removing losing parties’ ability to block the adoption of rulings — the negotiators built in a formal review of the DSU’s operation. “The DSU Review” was duly initiated in 1998. Over the years, many suggestions to improve the operation of WTO dispute settlement were made by WTO members. A core substantive concern of the United States regarding the operation of the AB (that the AB has sometimes overreached its mandate) was raised in the DSU Review almost two decades ago. In 2002, the United States and Chile put forward a proposal on “improving flexibility and member control in WTO dispute settlement,”5 which aimed to address US concerns regarding AB rulings on anti-dumping that targeted zeroing, a US practice designed to inflate dumping margins, and more generally to create “some form of additional guidance to WTO adjudicative bodies.”6 Whether engagement with this proposal would have helped avoid the current AB crisis cannot be known, but other WTO members rejected the suggestion. Other proposals made in the DSU Review (for example, the European Union’s suggestion that a permanent panel body be established — a true court of first instance, which could have reduced the need for appeal) might also have helped prevent the current AB crisis, if they’d been taken up.
Whether engagement with this proposal would have helped avoid the current AB crisis cannot be known, but other WTO members rejected the suggestion.
In fact, many reform proposals were made during the DSU Review, and not one of them was adopted. The WTO’s working practice, in particular its consensus-based decision making, is likely the reason. Waiting on consensus also permitted the United States to block AB appointments. Although some WTO members sought to discuss matters raised by the United States during the DSU Review, the need for consensus prevented a flexible response to changed circumstances and priorities and made the DSU Review an exercise in futility.
The specific issues raised by the United States regarding the functioning of the AB eventually became the focus of a separate process launched by the General Council in December 2018. Ambassador David Walker (New Zealand) was appointed as “facilitator,” with a mandate to explore resolution of a number of issues raised by the United States, which arguably should have been addressed well before the crisis erupted, during the DSU Review. The consultative process proved to be too little too late; by that time, the key protagonists were deeply entrenched in their positions. These consultations were forced to address the relatively insignificant procedural issues raised by the United States, diverting attention from the core issues falling under the broad heading of “WTO Member control” in the DSU Review discussions.
The quintessential US criticism concerns the AB’s alleged overstepping of its mandate, most notably exemplified in the haphazard treatment of the idiosyncratic standard of review embedded in the WTO Anti-Dumping Agreement (article 17.6).7 This provision, introduced at the insistence of the US delegation in the Uruguay Round, was meant to act as a deferential standard in favour of interpretations adopted by investigating authorities, if panels found that more than one permissible interpretation were possible in any given dispute. The US delegation’s understanding was that article 17.6 served as a green light for zeroing. The AB was required to give meaning to article 17.6, but the US critique is that the AB only paid it lip service, and frankly, this critique is well founded. Panels and the AB have routinely said that the article 17.6 standard of review is not at odds with the generic standard of review, and so they have not seriously engaged with article 17.6. Arguably, nothing would have changed with respect to zeroing case law had the AB approached the interpretative issue from the angle of article 17.6, and it is unfortunate that they did not do so.8
The AB was required to give meaning to article 17.6, but the US critique is that the AB only paid it lip service, and frankly, this critique is well founded.
Renegotiation of the zeroing issue is probably the wisest path forward, as case law continues to be erratic on this matter (in April 2019, the panel on US-Price Differential Methodology went head-on against 25 years of AB case law and found that zeroing can be WTO-consistent). Where they are not clear, rules should be clarified by the WTO membership. One way to encourage such clarification would be for the WTO membership to require the AB to send cases where the rules are unclear to the WTO bodies responsible for implementing the agreements invoked in a dispute.
If this could be agreed, it would still need to be recognized that panels and the AB unavoidably will have substantial discretion, as they must interpret one incomplete contract (the WTO) by using another incomplete contract (the Vienna Convention on the Law of Treaties, which does not assign specific weights to its various elements). If it were possible to write a more complete contract, that would have happened. Against this background, what is needed is to better select those entrusted with adjudication, and to pay more attention to the organizational aspects of adjudication.
Elements for Future Negotiation on Reforming WTO Dispute Settlement Procedures
A Roster of 15–20 Permanent Panellists
- Panellists should serve for one term of eight to 10 years.
- Depending on criteria to be defined (new issues, value of disputes and so forth), disputes should be heard by divisions of three panellists (for relatively less important cases), or divisions of seven panellists (for relatively more important ones).
- Decisions should be taken by majority.
- Dissenting opinions should be published.
An AB of Nine Members
- Each AB member should serve one term of eight to 10 years.
- The AB will decide cases in divisions of three members.
- The AB will decide by majority voting.
- Dissenting decisions will be published.
- The collegiality requirement will persist.
- WTO members will decide on the establishment of a commission of eminent experts well-versed in General Agreement on Tariffs and Trade/WTO dispute settlement who will be entrusted with the task to screen the proposed panellists, as well as the AB members put forward by the members of the WTO.
- WTO members should decide the experts for the commission by consensus agreement but should be allowed to decide on panellists and AB members with a qualified majority vote.
- The AB members, as well as the panellists, will have the right to appoint their clerks. The number of clerks serving each judge will be decided ex ante, and AB members may select only one clerk of their own nationality.
There is, of course, much more to think about when determining how to regulate the WTO adjudicating bodies more comprehensively. The above could serve as basic axes to help address some important dimensions, such as the quality of judges, the incentives of adjudicators to please their nominating party and the confusion of functions of the WTO secretariat. These suggestions complement the so-called “Walker Principles” put forward by Ambassador David Walker to address US concerns with the operation of the AB and to ensure that:
- appeals are completed within 90 days;
- AB members do not serve beyond their terms;
- precedent (case law) is not binding;
- facts cannot be the subject of appeals;
- the AB be prohibited from issuing advisory opinions; and
- the AB’s findings cannot add obligations or take away rights provided by the WTO agreements.9
These principles are fully consistent with — and indeed often echo — what is already in the DSU. For this reason, they should be amenable to all WTO members and serve as the basis for the substantive agreement needed to address the core US concern: credible measures to ensure the AB will stick to its mandate.
One lesson from recent events is that more interaction between WTO members and a reconstituted AB is needed.
Some type of advisory review body, as proposed by the US business community, with a mandate to assess and report on compliance by the AB with the Walker Principles may help provide greater assurance that matters relating to the performance of the AB can be given greater attention in the DSB. However, at the end of the day, if WTO members believe the AB is exceeding its mandate, they will have to address the problem by renegotiating the substantive provisions of specific agreements. As mentioned, this work could, in part, be facilitated by putting the burden on the shoulders of the membership as opposed to the adjudicative function by stipulating — as a procedural matter — that the AB should request the relevant WTO bodies to clarify the pertinent disciplines, if rulings hinge on the interpretation of the invoked provisions of a WTO agreement in instances where there are gaps or where rules are unclear.10
One lesson from recent events is that more interaction between WTO members and a reconstituted AB is needed. In doing so, it is useful to distinguish between substantive and procedural rules. Procedural changes in the implementation of DSU by the institution lie at the heart of any resolution of the AB crisis. Such changes require deliberations and decisions by the membership to implement specific reforms to improve the operation of the DSU. If necessary, such process-related changes should be subject to a vote, as envisaged by article IX of the Marrakesh Agreement.11 Doing so is not in the DNA of the organization, for good reason. We strongly support the principle of consensus-based decision making when it comes to substantive rules and negotiated rights and obligations. But we are also of the view that voting on procedural reforms that improve the operation of the institution without affecting the rights and obligations of WTO members should not give rise to concerns that this is a slippery slope. If procedural reform proposals are well prepared — informed by consultations and supported by the good offices of the director-general — voting may not be needed in any event. If the membership is bold enough to adopt proposals along the lines indicated here, we might start seeing some light at the end of the tunnel.
This essay draws on work in progress by the authors.12