The WTO First-Instance Ruling, in Softwood Lumber, on the USA’s Methodology in Anti-Dumping
On April 9, 2019, the WTO Dispute Settlement Body issued its Panel Report regarding the United States’ particular use of Differential Pricing Methodology (DPM) in reaching its conclusion that the Canadian softwood lumber industry had been dumping into the United States. The Report has to do not with the Department of Commerce’s (DOC) findings that Canadian softwood lumber is subsidized (which are subject to other challenges) but with its finding that dumping can be found by applying a particular methodology.
Canada announced on April 15 that, though the Panel held in Canada’s favour regarding the DOC’s justification of its choice of methodology, it will be filing an appeal to the WTO Appellate Body on what can be done with that methodology. Considering the Appellate Body’s significant backlog, caused in large measure by the United States’ refusal to cooperate in the naming of new Appellate Body members to fill vacancies there, even putting in danger the Appellate Body’s capacity to field a quorum, this appeal might take some time. Presently, the terms of two Appellate Body Members are set to expire in December of 2019, at which point (in the absence of new appointments) the Appellate Body’s minimum quorum of three Members would be lost.
Some have held the Panel decision out as a successful, from the United States’ point of view, reversal of years and years of WTO rulings against the use of “zeroing” in calculations of margins of dumping. A reading of the Report indicates that this is, if not false, an ambitious overstatement.
Dumping is, simplistically, selling abroad at prices lower than the prices at which one sells at home. The vendor’s domestic prices are used (unless they are unreliable in which case proxy prices are found) to generate a “normal value” for the goods, which is then compared to the “export price”. If the export price is lower, it is dumped, and the margin of dumping is the difference between the domestic price and the lower export price.
If there are a number of export sales during the period, some of them may be found to be dumped and some others may be found not to be dumped. An aggregate margin of dumping must then be found. “Zeroing” is the practice whereby the investigating agency simply ignores the undumped sales. It inflates the margin of dumping by calculating a weighted average using only what is below the normal-value criterion. Indeed, if there are a sufficient number of high export prices, which would normally outweigh the dumped sales, it can turn a year of what would be, in the weighted-average aggregate, undumped sales into a year of dumping.
The dumping thus found and its margin can be used, as is the case in Softwood Lumber where the investigation dives into the details of selected exporters only, to spread dumping findings and resulting dumping-margin-based duty to other exporters whose exports have not been examined at all. That is, an uninvestigated exporter can be saddled with duty without ever having had a single of its sales found to have been dumped, even if no other exporter’s sales are found to have been dumped on a real weighted average, if the other exporter is found, by the arithmetic artificiality of zeroing, to have been dumping even if by a margin divorced from overall reality.
The United States has several times sought to defend the DOC’s penchant (other countries’ investigative agencies, including Canada’s, have also on occasion used zeroing). Several WTO decisions find it to lead to unfair results. This Panel Report does open a door to zeroing, but a narrow one and one which might well close again on appeal.
In short, the Panel found that zeroing is permissible but only in a third, exceptional methodology, which can only be chosen in a particular circumstance and then only if a cogent explanation can be provided that it is impossible to apply either of the other two methodologies. It also held that, in this case, the DOC had impermissibly leapt to this third methodology.
The Panel therefore found (par. 8.1) that the United States had acted inconsistently with the WTO’s Anti-Dumping Agreement, that this had presumptively nullified or impaired benefits to which Canada was entitled (8.4), and recommended (8.5) that the United States bring its anti-dumping measure into conformity.
Not only is it difficult to parlay this into a victory for the United States, whose choice of methodology was in the end found inconsistent with its treaty obligations, but the holding that zeroing would have been permissible as a limited part of the third, exceptional methodology was accompanied by reiteration that zeroing is wholly unacceptable in either of two other, primary, methodologies. There is, furthermore, good reason to expect that the Appellate Body will reverse, for there are sound logical reasons for finding zeroing to be just as inapposite and unfair in the narrow context of the third, exceptional methodology as in the broader context of the two primary methodologies.
The DOC chose not to use the W-W methodology, comparing weighted average normal values to weighted average export prices, and chose not to use the T-T methodology, comparing normal value and export price on a transaction-by-transaction basis. These two methodologies are, according to the first sentence of Article 2.4.2 of the Anti-Dumping Code, the two methodologies “normally” used to determine margin of dumping. (Report par. 7.13) The Appellate Body also refers to them as the “normal methodologies”. (Report par. 7.19) The Panel here also adopts that vocabulary, more or less throughout.
The Panel followed and confirmed the previous numerous rulings to the effect that zeroing can be used in neither of these two primary methodologies. (Report par. 7.18 and 7.19) Zeroing would fail to take into account the prices of "all" comparable export transactions, as is required under the W-W methodology described in the first sentence of Article 2.4.2. It would also fail to determine the dumping margin for the product as a whole.
The DOC, though, resorted to the W-T methodology, and the United States argued that, for this exceptional methodology, zeroing is permitted.
The Panel found, relying on the clear wording of Article 2.4.2, that the investigating agency can only bring itself within the exceptional W-T methodology if two conditions are met, so as to guarantee resort to the W-T methodology only when it is necessary to unearth targeted dumping which might escape the two primary methodologies:
- 7.14. The second sentence of Article 2.4.2 sets out the third methodology to determine the dumping margin. It states that "[a] normal value established on a weighted average basis may be compared to prices of individual export transactions" if the following two conditions are met by the investigating authority:
- it finds "a pattern of export prices which differ significantly among different purchasers, regions or time periods" (we refer to this part of the second sentence of Article 2.4.2 as the pattern clause); and
- it explains why "such differences cannot be taken into account appropriately by the use of a [W-W or T-T] comparison" .. .
- 7.15. The W-T methodology is considered an exception because its use is permitted only when these two conditions are met. The function of the second sentence is to identify and unmask dumping targeted to certain purchasers, to certain regions, or in certain time periods.
The DOC wrote that it had identified a single “pattern of export prices which differ greatly among different purchasers, regions and time periods” rather than a (or several) “pattern of export prices which differ greatly among different purchasers, regions or time periods” (Art. 2.4.2 of Anti-Dumping Agreement). The Panel found that this way of approaching things allowed an agglomeration of trends and pricing which necessarily makes it more difficult rather than easier to discern whether there is targeted dumping.
- 7.42. The issue before us is whether the pattern clause permits an investigating authority to aggregate export price variations across purchasers, regions and time periods to find a single pattern of export prices which differ significantly among different purchasers, regions and time periods. The Appellate Body and the panel in US – Washing Machines concluded that it does not. Canada relies on these findings of the Appellate Body to contend that in the underlying investigation the USDOC failed to find a pattern among different purchasers, regions or time periods because it aggregated differences in prices across three unrelated categories, i.e. purchasers, regions and time periods. The United States disagrees with Canada's arguments.
- 7.43. We note that an investigating authority must identify a pattern of "export prices which differ significantly among different purchasers, regions or time periods". …, the comparisons must be made between export prices to different purchasers, or different regions, or different time periods, i.e. with respect to categories of the same type.
- 7.44. … According to the United States, the text of the pattern clause on its face contemplates a pattern of export prices that transcends multiple purchasers, regions or time periods, noting that if the pattern clause prohibited the identification of such a single pattern, instead of using the preposition "among" only once, it would have used this preposition three times (as in among purchasers, or among regions, or among time periods) . We, however, disagree with the United States' arguments.
That is, the Panel rejected the basis on which the United States argued that zeroing was permitted, by rejecting the manner in which the United States brought itself into the third, exceptional category.
The Panel’s writing is not as clear as it might have been but can be summarized this way: one cannot argue that the two primary methodologies are insufficient to discern the targeted dumping if the pattern-finding methodology one uses to get into the third, exceptional methodology is one which needlessly obfuscates differences rather than shedding light on them.
- 7.49. … [W]e find that the USDOC acted inconsistently with the second sentence of Article 2.4.2 … because in applying the DPM, and specifically under the ratio test, it aggregated differences in export prices across unrelated categories, i.e. purchasers, regions and time periods to identify a single pattern of export prices which differed significantly among different purchasers, regions and time periods.
The Panel did say that, if the United States had properly justified its recourse to the W-T methodology (which one supposes is still possible if the United States decides to bring its measure into compliance), then it could use zeroing as part of that exceptional methodology, but only to the transactions fitting within the pattern which justify recourse to the W-T methodology. Non-pattern transactions must be handled by one or the other of the normal, primary methodologies and they must be handled without zeroing.
- 7.78. We consider, as past panels and the Appellate Body have, that the second sentence of Article 2.4.2 does not permit an investigating authority to apply the W-T methodology to all export transactions. Instead, this methodology may be applied only to pattern transactions. … an investigating authority must apply the W-W or the T-T methodology to those non-pattern transactions. The intermediate result calculated by applying the W-T methodology to pattern transactions must be aggregated with the intermediate result calculated by applying the W-W or the T-T methodology to the non-pattern transactions. The intermediate result based on non-pattern transactions may not be excluded, irrespective of whether that result is positive or negative. We have reached these conclusions based on the legal analysis set out below.
(See also 7.85 and following.)
This is of course another reason for insisting that the investigating agency not use a blunderbuss pattern to excuse its recourse to the exceptional W-T methodology, but one which ensures that the eventual zeroing, if there is to be any, is not too broadly applied.
The Panel thus rejects the United States’ argument that, if zeroing were not permitted in the exceptional methodology, for all transactions, it would no longer be exceptional (par. 7.72) and would anyway just produce the same result as the normal, primary methodologies. It does, however, uphold the United States’ argument that zeroing must be permitted for within-pattern-transactions for that exceptional methodology to have meaning and to allow it to unmask targeted dumping, and it does so even though it knows that it is acting contrary to Appellate Body findings:
- 7.106. Considering the raison d'être of the W-T methodology is to unmask targeted dumping, the inability of this methodology to do so will render this methodology inutile. We recall 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. Therefore, contextual considerations also support our view that the second sentence of Article 2.4.2 does not prohibit zeroing under the W-T methodology. Based on the above, we find that an investigating authority is permitted to use zeroing while applying the W-T methodology to the pattern transactions.
- 7.107. We are aware that our conclusions in this Report differ from those of the panel and the Appellate Body in US – Washing Machines as well as the panel in US – Anti-Dumping Methodologies (China). This is the result of our objective assessment of the facts of this case, and the applicability of, and conformity with, the relevant covered agreements. We have carefully considered these reports of the panels and the Appellate Body, and found convincing or cogent reasons to arrive at conclusions different from those of the Appellate Body in US – Washing Machines as well as the panels in US – Washing Machines and US – Anti-Dumping Methodologies (China).179
If an appeal is heard, though, there are good reasons to believe that the Appellate Body will again close the narrow opening which the Panel Report has now theoretically opened. The unfairness of a zeroing methodology which pretends that undumped sales never happened is just as striking in the third methodology (even if only for pattern-transactions) as in the two primaries, as is its failure to assess the true effect in the importing market of the exporter’s exports overall.
More fundamentally, it is simply not compelling that the exceptional methodology, however one gets to it, is incapable of unmasking targeted dumping unless the investigating agency uses zeroing. Even were it the case, a dumping margin which can be found only by using an arithmetic trick to blind the investigator to undumped sales can only be regarded as suspect, and one which brings the anti-dumping regime into disrepute.