As Susan Esserman and I pointed out in Foreign Affairs some years back, the WTO dispute settlement system is among the most advanced adjudicative mechanisms yet devised for international law. One of its innovative features is a means by which disagreements over compliance are themselves subject to resolution by the judges. The risk is that this process, stipulated in Article 21.5 of the Dispute Settlement Understanding, may be abused to undermine the finality and certainty of rulings, extending or even expanding the dispute instead of resolving it.
An egregious example of such abuse has been the way in which Mexico has chosen to attack the US dolphin-friendly voluntary labeling scheme in WTO litigation. Mexico’s bad faith conduct began with its very filing of the dispute, in flagrant violation of a choice of forum clause in the NAFTA, which specified that disputes of this nature, involving environmental or scientific facts, must be brought before the NAFTA forum not the WTO (NAFTA Article 2005.4 (a) and (b))
When the dispute finally got the WTO Appellate Body, Mexico lost on almost all fronts, except one narrow issue. The United States changed its scheme to address that one issue, concerning even-handedness of regulatory burden being calibrated to the risk to dolphins in Mexico's Eastern Tropical Pacific (ETP) fishery and other fisheries. Then Mexico decided to abuse the 21.5 process to open up many issues already decided by the AB, and indeed other issues never litigated, basically manipulating the 21.5 process to re-open and reframe the entire dispute. Incredibly, the panel bought it; and even more incredibly, although with some important pushback, the AB has rewarded or condoned such behavior. Now, a sophisticated WTO litigant that doesn’t like the outcome of the original litigation has an option to constantly re-litigate under the guise of the 21.5 process.
One of the key advantages of the WTO adjudicative system, to produce some kind of finality in controversial trade matters, can easily be eroded. Especially by WTO Members who have lots of experience in trade disputes and lots of trade lawyers. At the limit, and fortunately it was the United States on the other end of this dispute and not a capacity-limited Member, the respondent could be forced into settlement in order to obtain some certainty, regardless of the legal merits of its claim. The Appellate Body in its 21.5 ruling found the United States in continuing violation not because of its changes to the original measure, nor because it had failed to amend the original measure in the ways the AB had originally suggested were necessary, but on the basis of features of the original measure unaltered from before the original dispute and un-impugned in the original AB ruling. Almost by good fortune, the finding of violation is so obscure and imprecise that there will certainly have to be a further round of 21.5 proceedings. These should give the AB an opportunity to reconsider the system-threatening features of the decision just rendered. The decision did close some doors to the kind of abuse exemplified by Mexico’s conduct in this dispute; next time round the AB must shut the others. My sense is that the AB was divided in this case, perhaps quite sharply. It would not be difficult to reverse engineer as it were (pardon the slight misuse of that phrase) the AB 21.5 ruling into different opinions, including on the issue of the need to shut the door to abusive relitigation of the original dispute in 21.5 proceedings. (Apparently, the head of the WTO Appellate Body secretariat, Werner Zdouc has a horror of separate opinions-one that seems to outweigh the horror of incoherent unanimous reports) The outcome of the AB 21.5 ruling, a finding of violation that is rather fuzzy and based on abstract comparisons of specific wording in different clauses, without a real world grasp on the meaning for even-handedness as far as effects go, may reflect a very strenuously bargained overlapping consensus, a constructed and arguably artificial compromise.
The Original AB Ruling
Let’s start with what the AB decided in its original ruling. In its report in US-Tuna the Appellate Body 1) reversed findings of the panel that led it to the conclusion that the original measure violated TBT Articles 2.2 and 2.4); 2) reversed findings of the panel that led it to the conclusion that the original measure did not violate TBT Article 2.1; 3) completed the analysis under TBT Article 2.1 and in so doing found that the United States was in violation of Article 2.1; 4) found that the panel’s approach to judicial economy, in not considering Mexico’s GATT claims, violated Article 11 of the DSU, but did not complete the analysis under GATT, because Mexico stated in response to questioning by the division of the Appellate Body that it did not wish the Appellate Body to complete the analysis under GATT if it found a violation of TBT Article 2.1. In sum, the sole inconsistency with the covered agreements found by the Appellate Body was a violation of 2.1, upon completion of the analysis. The nature of this inconsistency was articulated by the Appellate Body in paragraphs 284ff of its Report. The Appellate Body first noted that “The aspect of the measure that causes the detrimental impact on Mexican tuna products is thus the difference in labelling conditions for tuna products containing tuna caught by setting on dolphins in the ETP, on the one hand, and for tuna products containing tuna caught by other fishing methods outside the ETP, on the other hand. The question before us is thus whether the United States has demonstrated that this difference in labelling conditions is a legitimate regulatory distinction, and hence whether the detrimental impact of the measure stems exclusively from such a distinction rather than reflecting discrimination.”(Paragraph 284) The legitimate regulatory distinction invoked by the United States was that the regulatory costs, and costs imposed on the industry, by the labelling conditions in the case of the ETP were not justified in the case of non-ETP tuna because of the considerably lower risk to dolphin mortality of tuna fishing outside the ETP. The Appellate Body found the United States had not discharged its burden of proof with respect to legitimate regulatory distinction, since it had not shown that only through the same kind of costly measures that the US used to address dolphin mortality in the ETP could it tackle the lesser risk of dolphin mortality outside the ETP. In particular, the AB raised the possibility that the US scheme could be “calibrated” to this lesser risk through certification by the ship’s captain, avoiding the expense of independent observers, etc. This notion was not vetted by the AB merely as a random hypothetical: as the AB pointed out, certification by the captain in the case of fisheries other than the ETP was an option already contemplated by the relevant US statute. It is not surprising that the AB came to the conclusion that the United States had not discharged its burden of proof, since the US authorities had not even attempted to use the means already available to them by statute (namely, captain’s certification) to address the outside the ETP risks, means that did not present the kinds of regulatory costs and burdens associated with independent observers in the case of the high-risk ETP tuna fishery.
In sum, it was evident to the Appellate Body that these high regulatory costs and burdens could not persuasively constitute a legitimate regulatory distinction, if there were clear, available alternative means for the United States to calibrate its labeling conditions based upon the relative risks to dolphin mortality posed by different fisheries. Through its 2013 Final Rule the United States adopted these alternative means.
The 21.5 Compliance Panel
The question for the 21.5 panel should have been two-fold: 1) first and foremost, whether the regulatory changes the US made to bring itself into compliance had the result of sufficiently narrowing the difference in labeling conditions between ETP and non-ETP fisheries such that this difference could be considered now to derive exclusively from a legitimate regulatory distinction; 2) whether the original measure was altered such as to give rise to some new, different violation of a provision of a covered agreement. The Appellate Body observed in Softwood Lumber VI (Article 21.5-Canada): “the scope of Article 2.15 proceedings logically must be narrower than the scope of original panel proceedings.” (Paragraph 1708) Further in Canada-Aircraft (Article 21.5-Brazil) the Appellate Body held: “Proceedings under 21.5 do not concern just any measure of a Member of the WTO; rather, Article 21.5 proceedings are limited to those ‘measures taken to comply with the recommendations and rulings of the DSB….in principle, there would be two separate measures: the original measure which gave rise to the recommendations and rulings of the DSB and the ‘measures taken to comply which are-or should be-adopted to implement those recommendations and rulings.”(Paragraphs 1720-1721;emphasis in original). In EC-Bed Linen (Article 21.5 India), the Appellate Body held: “If a claim challenges a measure which is not a ‘measure taken to comply’, that claim cannot properly be raised in Article 21.5 proceedings.” (Paragraph 78: emphasis in original).” Further, in US-Shrimp (Article 2.15 Malaysia), the Appellate Body held that a 21.5 panel may not re-visit the WTO consistency of any aspect of a new measure taken for compliance that a) remains unchanged from the original measure and b) was not found to violate a provision of a covered agreement in the original proceedings. (Paragraphs 89-106) In particular, the Appellate Body emphasized the importance of the finality of the original findings and recommendations adopted by the DSB as a binding settlement of the dispute between the parties, in accordance with Article 17.14 of the DSU. (Paragraph 97)
The 21.5 panel, however, in its report made set of comparisons de novo between situations that were not compared by the Appellate Body in its completion of the analysis under TBT Article 2.1, and made findings of violation of 2.1 based upon these comparisons. That would have been acceptable, if this set of comparisons resulted either from changes to the various categories in the original measure and/or to the requirements applicable to each category, which raised new issues of WTO compliance. The paragraphs of its report where the panel engages in this exercise of de novo comparisons are confusing and at times almost incomprehensible. (Paragraphs 7.249-7.263). Part of the reason for this murkiness could that this adventure of de novo comparisons was undertaken by the panel on its own motion not based on any prior pleadings of Mexico, and indeed was raised for the first time by the panel with the parties in its questions to them.
The AB 21.5 Report
As already noted, there are parts of the AB report that rightly push back on Mexico’s attempt to relitigate the whole dispute under the guise of 21.5 proceedings. At least some AB Members had serious concern that Mexico’s conduct might be seen as an abuse of 21.5. Thus, in paragraph 5.8 the AB underlined: “5.8. In reviewing the WTO-consistency of a measure "taken to comply", compliance panels should be mindful of the principle of prompt settlement of disputes embodied in Article 3.3 of the DSU. Accordingly, compliance proceedings cannot be used "to 're-open' issues decided in substance in the original proceedings." Further, the AB questioned whether the panel’s method of operating in finding new violations concerning the original measure was consistent with its justification that it had jurisdiction under 21.5 due to interlinkages between the original measure and the steps taken to comply: “The Panel had, in discussing its jurisdiction under Article 21.5 of the DSU, emphasized the interlinkages between elements of the amended tuna measure, it subsequently conducted a segmented analysis that isolated consideration of each element of the measure without accounting for the manner in which the elements are interrelated, and without aggregating or synthesizing its analyses or findings relating to those elements before reaching its ultimate conclusions as to the consistency or inconsistency of the amended tuna measure.”(7.2) At least, here the AB recognized that the panel wasn’t limiting itself to examining aspects of the original measure in their interaction with the steps that the US had taken to comply but rather was using this as a pretext to allow Mexico a second kick at the can, simply re-arguing the original measure. A similar recognition is reflected also in the following observations of the AB: It is also surprising that, in this part of its reasoning, the Panel made little reference to the original Appellate Body report in this dispute and did not acknowledge or discuss the concept of "calibration" advanced by the United States, and used by the Appellate Body as a means of testing the even-handedness of the original tuna measure” (Paragraph 7.101).
While saying all this, however, the AB found a way of permitting a second kick at the can, a relitigation of the original measure: aspects of the original measure could be reviewed, even if they were not challenged or decided in the original proceeding, where these aspects of the original measure constitute "an integral part of the measure taken to comply".(Paragraph 5.8) Of course, unless the original measure is completely replaced with a new measure, it is not surprising that aspects of the original measure will interact with the new measure. The real issue for preserving the proper scope of 21.5 proceedings is: does the interaction lead to the aspects of the original measure being in non-conformity with the WTO law in a manner different or more severe than before that interaction? If not then this is a second kick of the can. But here the AB was dealing with the same question of non-conformity of the original aspects of a measure that could and should have been dealt with in the original proceedings.
The AB Finding of Violation
The AB finding of violation (upon completion of the analysis) entirely centers on the issue of the circumstances in which heightened risk to dolphins can trigger the highest level of monitoring and verification in the case of non purse-seine and non-ETP tuna, including the requirement of an observer on every boat. As noted, the lack of calibration that the AB had found in the original report, the sole basis for a violation, was the between the treatment of ETP purse seine fishery tuna and all non-ETP tuna. The reasoning was that Mexico had established there is some dolphin mortality risk in the latter case and therefore one could not justify on the basis of a legitimate regulatory distinction that there would be no real monitoring of dolphin mortality as a requirement for using the dolphin-safe label in the case of non-ETP tuna generally.
In addressing the situations where non-purse seine and non-ETP tuna could trigger the highest level of monitoring and verification, because of some specific identified heightened risk to dolphins, the AB was in fact examining an aspect of the original measure unaffected by the US modification, which introduced an intermediate level of monitoring and verification (captain’s certification) in the case of non-ETP tuna generally. Thus, even if not disingenuous, the AB justification for 21.5 jurisdiction here on the basis of supposed interaction between the modification of the original measure and aspects that remain unchanged is certainly strained. Basically, the new requirement of captain certification of non-ETP generally would in no way change the operation of the provisions where, in the original measure, because of heightened risk to tuna in a specific fishery the highest level of monitoring and inspection could be required. It is not as if, for example, captain’s certification introduced a lower level of monitoring and inspection for non-ETP generally but sacrificing the possibility of the highest level where a specific fishery was shown to present a special heightened risk to dolphins.
One would think that pre-existing provisions that allow the imposition of the highest level of monitoring and verification (the level applied to the ETP purse seine fishery) in other cases where comparable high risks to dolphins exist in specific fisheries would reinforce the evenhandedness of the modified measure not undermine it (there is even some language in the AB report to this effect, suggesting some considerable cleavage between AB Members in this dispute). So it is truly difficult to fathom the AB’s reasoning. But let’s try. It has something to do with two turns of phrase that are used in the original US legislative scheme to articulate situations where heightened risk to dolphins may trigger the highest level of scrutiny even in the case of non-ETP and/or non-purse seine fisheries.
In the case of purse seine fishing outside the ETP, the determination of “regular and significant association between dolphins and tuna” is what triggers the requirement of independent observer certification. This is B)(i) as set out in the original measure. This provision is based on the recognition that, even though the dolphin-tuna association is already well-established and pervasive in the ETP, such an association could occur elsewhere and thus should trigger comparable intensive monitoring in order to ensure the integrity of the “dolphin-safe” label, given the well-established dangers of using purse seine nets. Now if we turn to the second “determination” provision, which concerns “other fisheries”, this is the one set out in in (D) of the original measure. Here the trigger for independent observer certification is indeed different: “regular and significant mortality or serious injury to dolphins.”
According to the AB, “The determination provision applicable to the non-ETP purse-seine fishery allows for the addition of a requirement for observer certification if there is a determination of "regular and significant association", but not for a determination of "regular and significant mortality or serious injury". In our view, this is difficult to reconcile with the fact that such an observer certification is required in the ETP large purse-seine fishery, and that such a determination, on the basis of "regular and significant mortality or serious injury", can be made pursuant to the other relevant determination provision, which is applicable to "all other fisheries." (Paragraph 7.258)
Now it will immediately observed that the conditions of competition affected by the differences in wording here, if there is any effect, would be between non-ETP purse seine tuna and other non- ETP tuna. In other words, the relative treatment at issue does not affect the treatment of Mexican purse seine ETP tuna.
Why the different language? In the case of non-ETP purse seine tuna, regular and significant association with dolphins, combined with the use of purse seine technology, raises a robust presumption that there will be significant dolphin mortality. In other words, the presence of these two risk factors together leads to a common sense judgment that regular and significant mortality in likely to occur and thus the highest level of monitoring and verification is necessary.
The AB seems to think that there could be situations in the non-ETP purse seine fishery where there could be regular and significant mortality or serious injury to dolphins other than those that result from regular and significant association. Now this is of course entirely hypothetical on the AB theory. On that theory, there would only be a violation in situations where in the case of non-ETP purse seine tuna the highest level of monitoring and verification was not imposed regardless of there being regular and significant mortality to dolphins, i.e. merely because the “association” requirement had not been met. Even if one abstracts from the error of seeing this finding as within 21.5 jurisdiction (it concerns comparisons in the original measure unaffected by the US compliance steps), and also ignores the error of regarding the difference as affecting the conditions of competition with the Mexican purse seine industry while it could only affect the conditions of competition between non-ETP purse seine and other non-ETP tuna, it would be hard to find an “as such” violation here because there is no necessity based on the provisions themselves that certification distinctions would ever result in any modification of the conditions of competition. This would only happen if the regulator, in the case of non-ETP purse seine tuna, exercised its discretion not to impose the highest level of monitoring and verification in a situation where the existence of regular and significant mortality to dolphins could nevertheless be shown—i.e. simply because the “association” requirement was not met. It is unclear empirically that purse seine tuna fishing outside the ETP would ever pose a regular and significant risk to dolphins in instances where there is a lack of tuna-dolphin association. So, in other words, it is speculative whether the difference in the provisions could ever modify conditions of competition (even assuming the misguided view of the AB as to what conditions of competition we are dealing with); this would depend on the exercise of discretion against the highest level of monitoring and verification in a situation that might never arise-where there was serious and regular dolphin mortality in a non-ETP purse seine fishery without tuna-dolphin "association" existing . In sum, there is no “as such” violation because the mere wording of the provisions would not compel or create a serious risk of discretion being exercised so as to allow laxer monitoring and enforcement in a situation where regular and significant dolphin mortality or injury actually exists.
The above is an attempt to engage with what the AB was saying. But the reasoning is so confused and twisted that one cannot be sure of what they actually had in their heads in thinking of the difference in operation of the “association” and “mortality” conditions. For example, they say at 7.260: “we do not find persuasive the United States' argument that there is no basis for imposing an observer requirement in the absence of evidence that a purse-seine fishery exists where there is regular and significant mortality without tuna-dolphin association also being present.” (emphasis added) Yet, as just noted it is precisely in the case of “other” e.g. non-purse seine non-ETP fisheries, that there is a requirement of “mortality” but without that of “association”. The AB seems to have thought, contrary to what the law says, that the highest level of monitoring and verification was being imposed in the case of non-purse seine, non-ETP tuna only where both "association" and "mortality" were established.
What Is To Be Done?
What recourse is there against a ruling that is so confused and wrong-headed, and with negative systemic implications, opening the door (seemingly narrowly but actually pretty widely when you see how loosely the AB uses the concept of interlinkage) to misuse of the 21.5 process to relitigate the original dispute? One option, invoking the ghost of Hudec, is “civil disobedience”-simply not implement the decision given its gross illegitimacy, and accept the possibility of retaliation by Mexico. This might make sense in the circumstances, as the entire premise of the WTO dispute settlement authorities’ jurisdiction was Mexico’s internationally wrongful act of violating the forum selection clause in NAFTA. In turn, the US could continue to pursue a NAFTA claim against Mexico for its violation of NAFTA. This would result in US retaliation against Mexico that would fully match WTO countermeasures by Mexico. There is a possibility, Simon Lester raised this in a recent exchange of ours, that Mexico is blocking a panel to hear that claim. If that is the case then perhaps the US would be justified in taking countermeasures against Mexico under general international law for its violation of NAFTA and lack of good faith in implementing NAFTA’s dispute settlement procedures. It will be recalled that in the Soft Drinks dispute Mexico claimed that its own countermeasures against the US for blocking a NAFTA panel could be justified under GATT Article XX (d): that claim failed. But the AB in that case never addressed whether the GATT/DSU is a lex specialis that completely displaces the operation of the ILC Articles of State Responsibility with respect to countermeasures, especially to the extent that they reflect custom. Arguably, there are no provisions of the GATT or the DSU that purport to contract out explicitly of the right to take non-forcible proportional countermeasures against an un-remedied international wrongful act. Thus, again arguably, WTO Members have reserved the right to take countermeasures in the form of trade restrictions that, but for this right under custom, would be internationally wrongful as violations of WTO rules.
But there is a less unconventional course of action, which might permit the AB to reconsider some of its misadventures in this ruling. There could be a rule-making change or indeed merely an interpretative statement, that the regulator shall not in exercising discretion in the application of the determination provisions, make any determination that would result in a situation where the monitoring and verification requirements are uncalibrated to the level of risk of dolphin mortality and injury, as objectively determined. Further it is understood that the requirement of an independent observer where there is tuna-dolphin association in the case of non-ETP purse seine fisheries is intended, in its effect, to ensure that there is an independent observer in situations of heightened risks of dolphin mortality and injury of the kind that are known to exist in the case of the ETP fishery. The discretion of the regulator shall be exercised accordingly.
This approach to implementation will remind some readers of this blog of the solution in the Section 301 dispute. There was concern that the US law on its face would lead to serious risk of the exercise of discretion in a manner that violates WTO rules. The panel accepted a statement by the US in the litigation process that the law would not be interpreted and applied so as to lead to a violation of the WTO norms at issue. This satisfied the panel that there was no “as such” violation. It will almost certainly to Mexico bringing further 21.5 proceedings, and thus to a possible self-correction by the AB of its most ill-considered ruling.