For part I of this article, click here.

I. Allegation Regarding Creation of Law: Lust?

Now, to tackle head on the main argument of the authors: that the AB creates law when it has no business to do so.  But, we ask: who is to blame?

The WTO agreements as a whole (not just with respect to the ADA as mentioned above) constitute, what is referred by those familiar to the issue, as “constructive ambiguity“.  There are vast swathes of the law which are in dire need of meaning.  Who is to provide this meaning? The Marrakesh Agreement, agreed upon by all Members, put the “exclusive authority to adopt interpretationsin the hands of the Members.[1]  

But the Membership has failed and rarely, if ever, have such interpretations been adopted by the Membership.[2]  Indeed, then, in such case, the onus falls onto the AB vide  DSU Art. 17.6, which requires the AB to look into “…issues of law…and legal interpretation…”.[3]

The AB has simply been doing its assigned task, filling in for the Members, per its duties under the Marrakesh Agreement, so that everyone, not just the disputing parties, have a clearer understanding of what WTO law says.[4]  And then, as mentioned before, the lines between clarification, interpretation, gap-filling and law-making, are thin; sometimes these lines get blurred, but even that blurring (or the extent thereof) is a matter of perception.[5]  And, in any case, what then is the preferred counterfactual: no more independent third-party adjudication, compounded with a return to power-diplomacy?  Interpretative mistakes are part and parcel of the game, when a “law” is composed of vague texts of full of compromises,[6] but ‘course correction’ as foreseen in the founding Marrakesh text should be exercised before deciding to burn down the house.

II. “Ultra-hardline stare decisis“: Pride?

Finally, with respect to the allegation of super (or ultra)stare decisis, we can concede a point but not a sin.  We need not elaborate on the basis for such “de facto stare decisis“: suffice to highlight one of the highest goals of the WTO dispute settlement system: to ensure “security and predictability” in the trading system,[7] which, we contend, can only be achieved in a system that prizes uniformalised legal interpretation and jurisprudence.

To this end we agree with the authors that the “first strike” model of the AB, with regard to the creation of “jurisprudence“, is not ideal, and should probably be modified.  We wonder, however, if the comparator chosen by the authors (national courts) is the most apt when talking of international dispute settlement.  However, were something on the lines of the Mexican model to be adopted,[8] it would of course have to be tempered according to WTO reality: so while in Mexico a legal interpretation acquires the status of “jurisprudence” after five consecutive and uninterrupted iterations, the number required for the WTO, given a smaller number of disputes vis-à-vis Mexican courts, should also be lower.

III. Exceeding 90 days: Sloth?

A popular allegation that is frequently doing the rounds (and one not mentioned by the authors) is that the AB is too slow.  In fact, this is one of the US’ most significant repeat complaints – to the extent that some time ago it claimed that it would refuse to consider anything after the 90-day deadline as an “AB report” compounded with recent statements (adding insult to injury) that it considers that AB members should not even get paid for work done beyond their term.

But that contention of sloth is ridiculous.  Apart from the fact that no high-quality Court renders incessantly renders high quality verdicts within 90 days,[9] the 90-day deadline (established in the ’90s) reflected the understandable unfamiliarity with how successful the Dispute Settlement system could and would develop itself, especially when the only closely resembling experience that Members had to go by were the times of GATT and its Panel Reports.

Moreover, several things have changed.  Studies have shown how more recent cases are much more technically complex than their predecessors.[10]  Further, the number of complainants and third parties, have vastly increased compared to earlier cases; the AB is duty-bound to give consideration to all of their representations, and this takes time, not to mention that submissions made by these parties (often due to highly specialized pay-by-hour law firms) run into several thousands of pages.

Furthermore, this time-frame includes even some ancillary steps not in the hands of the AB, for example translation into the other two official languages, cutting 90 days down with, probably, a number of weeks.  The “problem” of ‘slow’ reports is moreover compounded by the party who complains the loudest – the US, for it is the US’ bad-faith-blocking[11] of AB appointments that has left us with ever-fewer adjudicators to deal with the case load.  And in any case, why should verdicts be rendered so fast and in such a hurry?  Speedy justice should not come at the cost of well-done decisions. This is particularly the case if one desires “good” rulings (as the US indeed does).  Speed is never really a good advisor and, if anything, it would actually be better if the suffocating 90-days deadline could be significantly loosened.  Slow, in this regard, is fast.


In conclusion, we would like to make some brief remarks about how to look at the issue of the AB.

First, and quite simply: everyone makes mistakes.  Yet, we are somewhat surprised at the extravagant attention that the handful of erring(s) of the AB receive.  Any adjudicatory body is liable to make mistakes.  Even the centuries-old US Supreme Court is wrong at times; by comparison, the AB is only 24 years old.  While it has crossed adolescence and glaring mistakes cannot be excused, it has also not fully matured, and so some patience, from the Membership, practitioners, and from scholars, should rightfully be expected.  This is particularly so when the AB, whether we admit it or not, performs a fine act of balancing law, economics and geopolitics.  Let us let the AB breathe.  And in this regard it is important, as we have mentioned above, to highlight and appreciate the places where the AB did, in fact, get it right.

Second, we agree with the authors (and also suggest) that the AB should admit mistakes and backtrack wherever needed and possible.  The authors have admitted that this has happened before.[12]  The authors are also correct in highlighting the work of Friedl Roessler, an old GATT/WTO hand, who argues that the AB has in fact made changes to its jurisprudence (which has reduced predictability in the system), but that the predictability suffers even more when the changes are made “in disguise” because “panels and Members … receive confused or conflicting normative signals“.[13]  Given that a recent panel has, for the first time, explicitly disagreed with a previous AB ruling (using the ‘cogent reasons‘ test),[14] the time is ripe for the AB to feel self-confident enough to admit previous shortcomings.

Third, we would like to spring-board off something mentioned by the authors, and stress on its importance – dissenting opinions in AB reports.  One scholar, at least, has hinted at the possibility of AB Members being pressurized by the Secretariat into delivering unanimous decisions.[15]  Such latter practice, if true, must be kept in check.  In our view, dissents would actually add legitimacy to the system and further soften the “hardline stare decisis” image of the AB.

Finally, what we all as engaged spectators, scholars and practitioners should be focusing on, are reforms.  The WTO dispute settlement system as we knew it, for all practical purposes, is dead.  It is time to think of what we want the future to look like, and there will be a future, no doubt: beyond the US-China trade war, trade will continue, which means trade disputes will continue, and after some amount of introspective soul-searching the global trade community will rebuild the multilateral dispute model.

It is time to think of that new model.  Important aspects need to be added to the AB’s (or its successor’s) arsenal; for example, the power of remand.[16]  Other cost-effective alternatives (compared to the complete shutdown of the dispute settlement system envisaged by the US) as indicated above, such as the use of interpretations by the Members should be made more easy so as to permit course correction of the AB, in order to deal with important law that has worldwide repercussions.  Regrettably, the US has approached the problems in a reckless and short-sighted way, and this, we submit, is actually what is “not helping“.


Views expressed by the authors are in personal capacity.


[1] See Marrakesh Agreement IX:2.

[2] The Ministerial Conference and General Council have not yet explicitly used the possibility of authoritative interpretation, as established in Article IX:2 of the WTO Agreement. Some provisions of the WTO Agreements have been interpreted in the Doha Ministerial Decision in Implementation-Related Issues and Concerns but no reference was made to Article XI:2 and this was adopted without an interpretation from the relevant Council. In this regard, the AB ruled in US – Clove Cigarettes that a recommendation from the relevant Council is an essential element of Article IX:2 (paras. 254-255).

[3] DSU Art. 17.6.

[4] Years ago, Ehlermann and Ehring have already (repeatedly) argued for a stronger role of the Membership, see, for example, Decision-Making in the World Trade Organization: Is the Consensus Practice of the World Trade Organization Adequate for Making, Revising and Implementing Rules on International Trade? Journal of International Economic Law, Volume 8, Issue 1, March 2005, Pages 51–75,

[5] See Akhil Raina, Meditations in an Emergency: The Appellate Body Deadlock – What It Is, Why It Is a Problem, and What to Do About It, 13(9) Global Trade and Customs Journal (2018).

[6] Notably the AD and SCM Agreements are arguably the worst sort of legal texts in terms of clarity, as a result of all the compromises that were agreed on.

[7] DSU Art. 3.2.

[8] See Miranda & Sánchez-Miranda, Part III.

[9] The ECJ sometimes takes years to issue a verdict.

[10] See Joost Pauwelyn & Weiwei Zhang, Busier than Ever? A Data-Driven Assessment and Forecast of WTO Caseload, 21(3) Journal of International Economic Law (2018); Louise Johannesson & Petros C. Mavroidis, The WTO Dispute Settlement System 1995-2016: A Data Set and Its Descriptive Statistics, 51(3) Journal of World Trade (2017).

[11] To repeat-block new AB Members from being appointed so that appeals cannot be performed, while simultaneously appealing panel reports so as to obtain an Appellate Body report (or to refrain a Panel report from getting adopted), and then continue to complain that the few remaining AB Members do not do work fast enough and should not work and should not get paid is manifestly and intrinsically contradictory.

[12] See Miranda & Sánchez-Miranda, Part II.

[13] See Friedl Roessler, Change in the Jurisprudence of the WTO Appellate Body during the past Twenty Years, 14(3) Journal of International Trade Law and Policy (2015).

[14] See DS534, US – Differential Pricing Methodology (also mentioned above in the context of ‘zeroing’).

[15] See: and

[16] See David Palmeter, The WTO Appellate Body Needs Remand Authority, 32 (1) Journal of World Trade (1998).


To make sure you do not miss out on regular updates from the Kluwer Regulating for Globalization Blog, please subscribe here.

This page as PDF

Leave a Reply

Your email address will not be published. Required fields are marked *