Against the backdrop of the problematic AB findings discussed in Parts I and II, the question is what can be done to avoid that missteps of such magnitude continue becoming part of the fabric of WTO case law, forever.
In this analysis a threshold issue that needs be touched upon is how the AB got to have the enormous power that it wields in respect of creating WTO case law. The answer is, that is not clear. The AB went from asserting in Japan-Alcoholic Beverages II that adopted reports (including AB reports) “create legitimate expectations among WTO Members” to holding in US-Stainless Steel (Mexico) that “panels are not free to disregard the legal interpretations in adopted AB reports”, in spite the absence of a mandate from the WTO Dispute Settlement Body to proceed in such manner.
Crucially, and this is a subject on which we delve at length in our paper, the AB adopted an approach to creating case law that is “super hardline” as compared to how case law is created by appellate courts in national jurisdictions, including in common law countries. Put differently, the system of precedents that has been created by the AB is far stricter than the case law system of common law countries. Three points are worth noting in this respect:
First, in the civil law jurisdictions that rely on case law, case law status is not conferred upon a single legal interpretation by an appellate court. Quite the opposite, in Mexico, for instance, for a legal interpretation to gain the status of “jurisprudence” (and thus become binding on lower courts), five consecutive and uninterrupted interpretations in the same sense must be adopted by higher courts. Similar approaches are used in other civil law jurisdictions. Thus, the AB’s case law system clearly surpasses in strength the notion of jurisprudence as understood in civil law countries.
Two, in common law jurisdictions there are avenues for appellate courts to backtrack on case law. For instance, in the United States circuit courts can overturn their own precedents with “special justification”, which requires “the full court of appeals sitting ‘en banc’ in plenary session” considering the merits of casting aside a prior decision in light of a series of factors including the “workability” of that particular decision. In assessing the “workability” of a decision under review, an appellate court would examine whether such decision “tended to generate inconsistent applications, fostered unclarity and uncertainty, or proven difficult to manage in any kind of principled way”.
Three, in common law jurisdictions it is well-established that not all precedents have the same value. For example, in the United States cases “decided by the narrowest of margins, over spirited dissents challenging the basic underpinnings of decisions” can receive lesser weight.
Conversely, in the approach to case law self-designed by the AB, case law status is conferred at the “first strike” (whether the AB got things right or not, which is unlikely on a general basis given the utmost complexity of the issues that end on its lap), the AB has never overturned one of its own precedents, at least not explicitly, and it remains to be seen whether the AB is willing to show any tolerance towards panel reports that latch on to the few dissenting opinions that have surfaced in AB reports, including the recent splendid dissenting opinion in the AB report in the 21.5 proceeding in US-Countervailing Measures (China).
Importantly, in national jurisdictions there is a political fix to deeply flawed legal decisions via amending the statute concerned or even the Constitution, provided the requisite majorities are assembled. In the WTO, by contrast, this political track is unavailable because modifying the text of any WTO agreement can only done by consensus, but the Member that benefits from a deeply flawed dispute settlement decision will block that consensus
We submit that, for WTO dispute settlement to be salvaged, things have to change. Nothing in the DSU prevents the AB from overturning its own precedents or from allowing panels to give dissenting opinions as much deference as to majority opinions. But for the AB to take either step it would have to concede that it is not infallible, particularly in the area of trade remedies and subsidy disciplines, and we are not there yet. Another possibility would be to shift from a “first-strike” model to case law to a model that requires several findings in the same direction for case law to coalesce, as happens in the civil law jurisdictions that rely on case law. While this approach is undoubtedly a tall order (because it would require amending the DSU), we think it would be very healthy, since it would open the door to debating the merits of AB decisions -not yet cast in stone- in panel reports and academic literature, and such debates can only be enriching for subsequent AB reports. In any event, what is clear for us is that continuing to assess superficially the origins of the current crisis is the best recipe for WTO dispute settlement to stall.
 John M. Walker, Jr., The Role of Precedents in the United States: How Do Precedents Lose Their Binding Effect?, Stanford Law School China Guiding Cases Project, Feb. 29, 2016, page 4, available at https://cgc.law.stanford.edu/wp-content/uploads/sites/2/2016/02/CGCP-English-Commentary-15-Judge-Walker.pdf
 John M. Walker, Jr., op. cit., citing Michael Stokes Paulsen, “Does the Supreme Court’s Current Doctrine of Stare Decisis Require Adherence to the Supreme Court’s Current Doctrine of Stare Decisis?, 86 N. C.L. Rev, 1165, 1172 (2008).
 John M. Walker, Jr., op. cit., citing Payne v. Tennessee, 501 U.S. 828-829 (1991).
 Frieder Roessler argues that the AB has indeed changed its jurisprudence in respect of a number of issues, although not openly. See his paper “Changes in the jurisprudence of the WTO Appellate Body during the past twenty years”, Journal of International Trade Law and Policy, Vol. 14, No. 3 (2015).
* The opinions presented in this post do not represent in any way official views of King & Spalding or its clients. We thank Bárbara Medrado of King & Spalding for having directed us to the Stanford materials on comparative case law. This post is a summary of research we have conducted preparing a paper intended for publication in the not so distant future.