United States—Anti-dumping Measures Applying Differential Pricing Methodology to Softwood Lumber from Canada. At https://docs.wto.org/dol2fe/Pages/SS/directdoc.aspx?filename=q:/WT/DS/534R.pdf&Open=True. World Trade Organization Panel, April 9, 2019 (unadopted).

This dispute, brought by Canada against the United States, constitutes another chapter in three separate sagas: the enduring softwood lumber dispute between the two North American nations; the debate over the acceptability of the practice of “zeroing”; and the fight over the value and role of World Trade Organization (WTO) Appellate Body precedent. Notably, the panel departed from established Appellate Body decisions finding, inter alia, that zeroing was permissible under a weighted average-to-transaction (W-T) methodology. This departure is remarkable, not just because it runs counter to prior jurisprudence, but also for the reasoning supporting it and the circumstances in which it occurred. Indeed, the Panel Report was issued in the midst of a crisis of the WTO dispute settlement system arising from the United States’ decision to block the reappointment of Appellate Body members.Footnote ¹ The United States justified this action, which eventually resulted in the Appellate Body losing its quorum to hear new appeals on December 10, 2019, on the basis of complaints, among others, that the Appellate Body had championed an approach to precedent that the United States found incompatible with the intended role of dispute settlement within the WTO.Footnote ² While members worked feverishly to formulate a compromise that might respond to the United States’ criticisms and soften the effect of the Appellate Body's approach,Footnote ³ the Panel suggested its own. Thus, it found room to depart from prior precedent (which the United States argued had been wrongly decided) while paying lip service to the Appellate Body.

The present dispute focuses on two controversial approaches—the use of W-T methodology and zeroing in the calculation of anti-dumping duties that an importing country may impose on products that are sold there at a lower price than in their home market (their normal value). The first is a method for the calculation of the margin of dumping—the difference between the export price of a product and its normal value. This determination is ordinarily carried out through the comparison of the weighted average normal value to the weighted average of all comparable export prices (W-W), or comparison of normal value and export price on a transaction-to-transaction basis (T-T). A different approach is allowed when an investigating authority finds a situation of “targeted dumping”—a term of art, not expressly reproduced by the Anti-dumping Agreement,Footnote ⁴ denoting a scenario wherein an exporter charged different prices to different purchasers, regions, or time periods in order to mask its dumping.Footnote ⁵ In such a scenario, to help “unmask” this type of dumping, an investigating authority may also determine the dumping margin by comparing a normal value established on a weighted average basis to prices of individual export transactions, provided that it offers an explanation as to why these differences cannot be taken into account appropriately by the use of a W-W or T-T comparison.Footnote ⁶

Zeroing, instead, is a controversial practice for the calculation of the margin of dumping, criticized by most WTO members, but tirelessly defended by the United States. It can be used under the W-W, T-T, or W-T methodologies and, with some generalization, entails ignoring sales where the export price is higher than the normal price, thus effectively treating them as zero in the calculation of the dumping margin. The practice has been criticized because it risks increasing the gap between export prices and prices of a product in its home market, which in turn, inflates the overall margin of dumping for a product and results in the imposition of higher anti-dumping duties. The Appellate Body has ruled against zeroing in every appeal it has heard.Footnote ⁷

The dispute here followed a U.S. Department of Commerce investigation of softwood lumber products from Canada that resulted in the imposition of anti-dumping measures applying the Department's “differential pricing methodology” (DPM). Canada claimed that the agency had acted inconsistently with WTO anti-dumping rulesFootnote ⁸—in particular, with Article 2.4.2 of the Antidumping Agreement, which establishes the W-T methodology, allowing a comparison of the weighted average normal value with individual export transactions to identify instances of targeted dumping.Footnote ⁹ Canada further alleged that the Commerce Department investigation was inconsistent with Article 2.4 of the Antidumping Agreement because it used zeroing under the W-T methodology in its application of DPM.

The first claims rested on two grounds. First, Canada alleged that the Commerce Department had aggregated export price variations in softwood lumber products across the three unrelated “categories” of purchasers, regions, and time periods to find a single pattern (para. 3.1), a method that the Appellate Body had previously rejected.Footnote ¹⁰ The Panel started from a textual analysis of the provision, the relevant portion of which reads as follows:

The Panel focused on the use of the word “among,” which showed “that an investigating authority may not compare prices between purchasers and regions because these two categories are not of the same type” (para. 7.43). The Panel showed sympathy toward the United States’ arguments in favor of a more holistic approach designed to unmask targeted dumping (para. 7.44), but rejected them. Noting its agreement with the similar findings of the Appellate Body in US—Washing Machines, it reasoned that the “pattern” clause does not permit an investigating authority to find a single pattern that aggregates differences in export prices across purchasers, regions, and time periods, thus finding in Canada's favor (paras. 7.45–7.49).

The second ground for Canada's first claims related to the identification of both higher-priced and lower-priced export sales among different purchasers, regions or time periods as part of the purported pattern (para. 7.50). The Panel thus had to determine whether a pattern can include export prices that differ significantly because they are significantly higher than others in one of those categories. Canada had relied on the authority of US—Washing Machines for the proposition that such pattern must be limited to the export transactions whose prices are found to differ significantly because they are significantly lower (para. 7.10), further arguing that such an interpretation naturally followed from the use of the word “pattern” and was systemically consistent with the Antidumping Agreement's focus on sale below normal value (para. 7.51). The United States countered that nothing in the pattern clause indicated a focus on either lower-priced or higher-priced export sales, but just export prices that “differ significantly” (para. 7.52).

The Panel started from the text of Article 2.4.2, observing that while “the focus of the pattern is on export prices which ‘differ significantly’ and thus not on all export prices,” no further qualification was included (para. 7.55). While acknowledging that even lower or higher export prices still needed to qualify as a pattern, it departed from US—Washing Machines, based on the Panel's understanding of the second sentence of the Article (paras. 7.56–7.57). The Panel saw it as designed “to unmask dumping targeted to certain purchasers, regions or time periods,” referring to the scenario where “significantly lower prices to certain purchasers, or certain regions, or in certain time periods are masked by significantly higher export prices to certain other purchasers, or to certain other regions, or in certain other time periods” (para. 7.57). Thus, as the panel explained, an investigating authority should be permitted to adopt a methodology that considers significantly higher-priced sales in order to unmask instances of targeted dumping (para. 7.58). The Panel noted that the contrary position taken by the Appellate Body in U.S.—Washing Machines flowed from its reliance on “contextual considerations,” such as the definition of dumping in the Antidumping Agreement (para. 7.59), which the Panel found unconvincing (paras. 7.60–7.61). Thus, the Panel found that Canada had not established that the United States had acted inconsistently with the second sentence of Article 2.4.2 (para. 7.66).

Canada's second claim related to the application of zeroing under the W-T methodology and its permissibility under Article 2.4.2. The panel started its analysis by recalling that the Appellate Body had rejected the issue in U.S.—Washing Machines, but was also quick to point out that one Appellate Body member had appended a dissenting opinion and that the majority had, in any event, conceded that “the W-T methodology is an exceptional methodology which is designed to unmask targeted dumping” (para. 7.68).

The Panel, agreeing with prior jurisprudence, clarified the scope of application of the W-T methodology, affirming that it was restricted to pattern transactions (paras. 7.78–7-84). However, it disagreed with the Appellate Body on the exclusion of nonpattern transactions when an investigating authority makes dumping determinations pursuant to Article 2.4.2 (paras. 7.85–7.100). Starting from the definitions of “margin of dumping” in the Antidumping Agreement and the General Agreement on Tariffs and Trade (GATT), it noted that the Appellate Body had relied on these to conclude that “dumping and margins of dumping must be determined for the product as a whole” and that, on its face, Article 2.4.2 did not appear to create an exception allowing the exclusion of transactions outside of the identified pattern (paras. 7.88–7.89). In fact, the Panel concluded, such transactions must be taken into account to properly determine whether and to what extent dumping is taking place (para. 7.90). The Panel further specified that that while an investigating authority would be permitted to apply the W-T methodology to the pattern transactions, only the W-W or T-T methodologies could be used for nonpattern transactions (para. 7.99).

The Panel then recalled its finding that higher-priced export transactions were included among the pattern transactions, as these may be masking lower-priced export sales (para. 7.101). The question remained whether, having identified such higher-priced export transactions, an investigating authority was permitted to “unmask” them by treating their value as zero. Again, the Panel noted the silence of Article 2.4.2 on the matter as well as the need to resolve the question in light of the function of its second sentence (para. 7.102). It found that provision permitted an investigating authority to compare a weighted average normal value with the prices of “individual”—rather than “all”—export transactions. The use of this term suggested that an investigating authority could distinguish those individual export transactions that mask others from those individual export transactions that are being masked. Accordingly, these transactions could be treated differently when making dumping determinations under the W-T methodology and, specifically, treated as zero. Doing otherwise would essentially “re-mask” them (para. 7.103).

The Panel found support for this conclusion in the exceptional nature of the W-T methodology, predicated on its function of unmasking targeted dumping (para. 7.104). If zeroing were to be impermissible under the W-T methodology, the result would be that the dumping margin calculated under it would be, in most cases, mathematically equivalent to that “based on the application of the W-W methodology to all export transactions, provided the weighted average normal values used under the W-W and W-T methodologies are the same” (para. 7.100), thereby rendering the W-T methodology useless for its intended function (para. 7.105). Recalling the Appellate Body's statement in U.S.—Gasoline that “an interpreter is not free to adopt a reading that would result in reducing whole clauses or paragraphs of a treaty to redundancy or inutility” (para. 7.106), the Panel found that Canada had failed to establish that the use of zeroing under the W-T methodology was inconsistent with Article 2.4.2, second sentence (para. 7.108).Footnote ¹²

Finally, the Panel addressed the question of whether the use of zeroing by the United States was inconsistent with Article 2.4 of the Antidumping Agreement. The Panel recalled that, in U.S.—Washing Machines, the Appellate Body had found zeroing under the W-T methodology inconsistent with the provision. However, it observed that this view was based on the finding that zeroing was prohibited under Article 2.4.2, thereby undermining the relevance of the precedent (para. 7.110). The Panel had instead already found that zeroing was permissible under Article 2.4.2, second sentence, and Canada failed to provide independent grounds for inconsistency with Article 2.4. As a result, the Panel found in favor of the United States on the point (para. 7.112). As to the claims under Articles 1 and 2.1 of the Antidumping Agreement as well as Articles VI:1 and VI:2 of the GATT, the Panel exercised judicial economy (paras. 7.113–7.115).

* * * *

Although the W-T methodology and zeroing are controversial, it is uncommon for a Panel to decline to follow previous Appellate Body reports, and to do so expressly.Footnote ¹³ Not only is this departure surprising in light of the de facto doctrine of precedent espoused by the Appellate Body, but it is all the more significant in light of the peculiar circumstances in which it occurred. The Panel Report was circulated in April 2019, at a time when debates in world trade law were dominated by discussions of the United States’ criticism of the Appellate Body, especially on the matter of precedent, and anxieties over its fate. The two issues were deeply interrelated. Indeed, the handling of precedent was a key grievance of the United States, which complained of that the Appellate Body had championed a de facto doctrine of stare decisis, while also indulging in judicial legislation by going beyond the dispute at hand and issuing “advisory opinions.”Footnote ¹⁴ With the continued effort by the United States to block the appointment of new Appellate Body members, the mandate of two of the remaining three members expired on December 10, 2019. Accordingly, the Appellate Body, which requires three members to hear appeals, ceased to function. While Canada promptly appealed the Panel Report well before that fateful day, the dispute was not among the ones that the Appellate Body would, under Rule 15 of its Working Procedures,Footnote ¹⁵ carry over.Footnote ¹⁶ Accordingly, the decision remains unadopted, and its treatment of the issue of precedent removed from scrutiny by the Appellate Body.

It would be facile to consider the Panel Report in U.S.—Differential Pricing Methodology an instance of defiance against a weakened Appellate Body. While the Panel overtly departed from prior Appellate Body decisions, acknowledging the differences in approaches, it defended this decision by reference to the very criteria set out by the Appellate Body regarding the precedential status of its reports.

This approach to precedent had been laid out in U.S.—Oil Country Tubular Goods Sunset Reviews, where the Appellate Body held that “following the Appellate Body's conclusions in earlier disputes is not only appropriate, but is what would be expected from panels, especially where the issues are the same.”Footnote ¹⁷ Subsequently, in U.S.—Stainless Steel, the Appellate Body further specified that “ensuring ‘security and predictability’ in the dispute settlement system, as contemplated in Article 3.2 of the [dispute settlement understanding (DSU)], implies that, absent cogent reasons, an adjudicatory body will resolve the same legal question in the same way in a subsequent case.”Footnote ¹⁸ To support its conclusion, it cited “the hierarchical structure contemplated in the DSU,” in which “panels and the Appellate Body have distinct roles to play,” further arguing that “failure to follow previously adopted Appellate Body reports addressing the same issues undermines the development of a coherent and predictable body of jurisprudence.”Footnote ¹⁹ In the present case, the Panel was careful to signal that its departure was not gratuitous, but rather the result of its “objective assessment of the facts,” the covered agreements, and, ultimately, the presence of “cogent reasons” to reject Appellate Body precedent. By doing so, the Panel arguably did not so much diminish the Appellate Body's stance on the power of precedent as lend support to it—at least formally.

The parties’ submissions and the interim review processFootnote ²⁰ shed additional light on the matter. The precedential status of Appellate Body reports had been the subject of extensive discussion by the United States and Canada—unsurprising, since the case largely hinged on the application of U.S.—Washing Machines as a putative controlling authority. In its submission, Canada had allocated considerable space to an endorsement of the U.S.—Stainless Steel approach,Footnote ²¹ as did third parties.Footnote ²² In turn, the United States had, consistent with its recent practice,Footnote ²³ expressed harsh criticism toward the Appellate Body's stance on the matter.Footnote ²⁴

Interestingly, this criticism seeped into the requests made to the Panel in the interim review process. Notably, the United States took issue with two points unconcerned with the substance of the decision: first, it objected to a sentence stating that the prohibition of zeroing under the W-W and T-T methodologies was “well established by now in WTO jurisprudence,” arguing that this might suggest that “WTO rights and obligations originate in WTO panel or Appellate Body reports, rather than the covered agreements.”Footnote ²⁵ Second, and more crucially, the United States requested that the reference to “cogent reasons” in the report be omitted.Footnote ²⁶ The Panel acceded to the first request, which was arguably consistent with its own line of reasoning, but not to the second.Footnote ²⁷

The Panel's insistence in maintaining a reference to the “cogent reasons” standard is noteworthy. Regrettably, and somewhat ironically, the Panel's reliance on the standard also reproduced its most critical weakness—the absence of clear-cut criteria to determine which reasons might qualify as “cogent” for a decision maker confronted with a putative controlling authority. Some previous panels had tried to define such criteria. In China—Rare Earths, the Panel relied on the dictionary definitions to indicate that “[t]he word ‘cogent’ means ‘[a]ble to compel assent or belief; esp. (of an argument, explanation, etc.) persuasive, expounded clearly and logically, convincing.’” It also looked at the pronouncements of other international adjudicatory bodies, and concluded that the threshold was a high one.Footnote ²⁸ Even more clearly, the Panel in U.S.—Countervailing and Anti-dumping Measures (China) stated that the expression “cogent reasons” could include as a multilateral interpretation under Article IX:2 of the WTO Agreement, a demonstration “that a prior Appellate Body interpretation proved to be unworkable in a particular set of circumstances falling within the scope of the relevant obligation at issue,” a demonstration that the prior interpretation leads to a normative conflict, or proof that the prior interpretation was “based on a factually incorrect premise.”Footnote ²⁹

The elaboration of principles for panels to depart from precedent has not, thus far, been accompanied by actual defiance. Indeed, the panels in the two cases cited above both concluded that no “cogent reasons” existed warranting a departure from Appellate Body precedent. It is perhaps a missed opportunity that the Panel would do the opposite here, finding reasons warranting a departure, but not elaborating expressly on them.

The approach taken by the Panel signals a willingness to depart from wrongly decided holdings. This is consistent with the general orthodoxy in international adjudication, where, barring exceptions,Footnote ³⁰ decisions of international adjudicatory bodies do not constitute binding precedents. The mention of the “cogent reasons” standard also provides a thread of coherence with the Appellate Body's approach espoused in U.S.—Stainless Steel. Yet, it is difficult to escape the belief that the “cogent reasons” standard must stand for something closer to the proposition that “a decision to overrule should rest on some special reason over and above the belief that a prior case was wrongly decided.”Footnote ³¹ If the security and predictability of the system is what counts, it might well be “more important that the applicable rule of law be settled than that it be settled right.”Footnote ³² Thus, deeper engagement with the issue was to be expected, especially on the matter of zeroing, which featured a long line of adverse authorities.

The Panel's solution confirms that Appellate Body precedent remains the starting point for any analysis of legal questions. It might also strike a balance between the deference to the “cogent reasons” standard and consistent criticism by the United States. Unsurprisingly, Canada's appellate submission closes with the claim that “[w]hile the Panel acknowledged the existence of the ‘cogent reasons’ standard that arises from the text of the DSU, it failed to provide any such reason that would justify departing from adopted Appellate Body legal interpretations and reasoning.”Footnote ³³ Conversely, while defeated on the matter of whether the pattern clause permits the identification of a single pattern aggregating differences in export prices across purchasers, regions, and time periods, the United States expressed satisfaction that it prevailed on the permissibility of zeroing under the Antidumping Agreement and, even more crucially, on the issue of precedent.Footnote ³⁴ The possibility cannot be discounted that the Panel's decision was an attempt to respond, within the boundaries set by the Appellate Body, to the criticism levelled by the United States.Footnote ³⁵ By doing so, the Panel may have planted the seed for a softened notion of precedential strength—a compromise, to be sure, but one that might have been embraced by the United States and the Appellate Body alike, thereby improving the chances for the long-term survival of the dispute settlement system.

In light of the Appellate Body's previous holdings on these matters, and its record in chastising “rogue” panels, its pronouncement on Canada's appeal would have been of considerable interest. Having been appealed, the Panel Report remains in a limbo,Footnote ³⁶ and it remains to be seen whether panels tasked with resolving similar matters will find guidance in it.Footnote ³⁷