This is from a dissenting opinion in today's Appellate Body report in U.S. - Washing Machines, with the dissenter arguing that zeroing should be permitted in the context of "pattern" transactions under the second sentence of AD Agreement Article 2.4.2:
5.1.10 Separate opinion of one Appellate Body Member regarding zeroing under the W-T comparison methodology
5.191. This dissent is limited to whether zeroing is permitted for "pattern transactions". My agreement with the other sections of this Report is subject to my views as expressed in this separate opinion.
5.192. The second sentence of Article 2.4.2 of the Anti-Dumping Agreement says that "[a] normal value established on a weighted average basis may be compared to prices of individual export transactions" 390 if an investigating authority finds the requisite "pattern" and provides the requisite explanation. This text has no qualifier, and it does not specify how the investigating authority is to do the comparison between a weighted average normal value and prices of individual export transactions.
5.193. The second sentence of Article 2.4.2 is an exception and has the function of "unmasking targeted dumping" and addressing it. Since the text of the second sentence does not say how that is to be done, the question before the Appellate Body in this appeal – the first to confront squarely the meaning of the second sentence – should be, what are the limits, if any, that the Anti-Dumping Agreement places on what an investigating authority may do to "unmask" and deal with "targeted dumping".
5.194. My distinguished colleagues of the majority have developed an interpretation that would allow investigating authorities to base W-T analyses solely on all "pattern transactions", but that would prohibit them from zeroing when doing so. In effect, investigating authorities may confine their examination to "pattern transactions", but, in doing so, they must combine the comparison results of all sales prices within the "pattern", that is, combine the comparison results of those prices above normal value with the comparison results of those below normal value within the "pattern".
5.195. The majority's interpretation would permit investigating authorities to deal with "targeted dumping" only partially, and possibly ineffectively. Within the "pattern", prices above normal value will cancel out – or "re-mask" – partly or completely, the "targeted dumping" that results from prices below normal value.
5.196. In my view, such an incomplete approach is not required by the text of the second sentence read in the context of the entire Article 2.4.2 and in light of the object and purpose of the Anti-Dumping Agreement, and it unduly restricts the regulatory leeway that should be accorded to investigating authorities to deal with "targeted dumping". Accepting that, when applying the second sentence of Article 2.4.2, investigating authorities are to focus only on "pattern transactions", I would permit investigating authorities also to zero those "pattern transactions" that are priced above normal value, and to calculate dumping only on the basis of "pattern transactions" priced below normal value. Doing so would deal fully with "targeted dumping" by dividing the full amount of such dumping – instead of an amount diminished by non-dumped prices – by the full value of an exporter's sales.
5.197. Let us test this. Are the "in-pattern", below-normal-value sales "dumped", so that they can properly be identified as the relevant "targeted dumping"? Article 2.1 of the Anti-Dumping Agreement defines "dumping" as occurring when a product is "introduced into the commerce of another country at less than its normal value, if the export price of the product exported from one country to another is less than the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country." The second sentence of Article 2.4.2 says that "[a] normal value established on a weighted average basis may be compared to prices of individual export transactions if the authorities find a pattern …".391 That is, literally, what "in-pattern" zeroing, as I am proposing it, would do.
5.198. Does the proposal of allowing "in-pattern" zeroing under the second sentence of Article 2.4.2 accord with the context of Article 2.4.2 and the function of the second sentence? The majority has said that: (i) "the effet utile of the second sentence in addressing 'targeted dumping' is fulfilled once an investigating authority has identified the relevant 'pattern' … and has established dumping and margins of dumping by applying the W-T comparison methodology exclusively to 'pattern transactions'"392; (ii) "[z]eroing the negative intermediate comparison results within the pattern is neither necessary to address 'targeted dumping', nor is it consistent with the establishment of dumping and margins of dumping as pertaining to the 'universe of export transactions' identified under the second sentence of Article 2.4.2" 393; and (iii) the majority's interpretation ─ but not the allowance of in-pattern zeroing ─ is consistent with the "fair comparison" requirement in Article 2.4 of the Anti-Dumping Agreement.394 In my view, none of these statements is backed by convincing authority or is self-evident.
5.199. Does previous Appellate Body jurisprudence prohibit "in-pattern" zeroing under the second sentence of Article 2.4.2? On the contrary, in US – Softwood Lumber V (Article 21.5 – Canada), the Appellate Body emphasized the exceptional nature of the W-T comparison methodology. 395 In US – Zeroing (Japan), the Appellate Body stated that "[t]he asymmetrical methodology in the second sentence is clearly an exception to the comparison methodologies which normally are to be used."396 Also in US – Softwood Lumber V (Article 21.5 – Canada), in finding that a prohibition on the use of zeroing under the T-T comparison methodology would not render the second sentence of Article 2.4.2 meaningless, the Appellate Body stated that, "on the contrary, … the use of zeroing under the two comparison methodologies set out in the first sentence of Article 2.4.2 would enable investigating authorities to capture pricing patterns constituting 'targeted dumping', thus rendering the third methodology inutile." 397 In the same case, the Appellate Body, in discussing the interpretation of the T-T comparison methodology in the first sentence of Article 2.4.2, further stated that "the reference to 'export prices' in the plural, without further qualification, suggests that all of the results of the transaction-specific comparisons should be included in the aggregation for purposes of calculating the margins of dumping."398 In my view, the fact that the second sentence, unlike the first sentence of Article 2.4.2, uses the phrase "prices of individual export transactions"399 indicates that not all the transaction-specific comparisons arising from the export prices that form part of the "pattern" need to be aggregated in order to calculate dumping. In US ‒ Stainless Steel (Mexico), the Appellate Body emphasized that "[t]he Appellate Body has so far not ruled on the question of whether or not zeroing is permissible under the comparison methodology in the second sentence of Article 2.4.2."400 The Appellate Body made this statement after having recalled the US – Softwood Lumber V (Article 21.5 – Canada) jurisprudence about the exceptional nature of the second sentence, the application of the W-T comparison methodology to individual export transactions falling within the "pattern", and the relationship between zeroing and the effet utile of the second sentence.
5.200. Regarding the text of the second sentence of Article 2.4.2 in other official languages, the French version reads:
Une valeur normale établie sur la base d'une moyenne pondérée pourra être comparée aux prix de transactions à l'exportation prises individuellement si les autorités constatent que, d'après leur configuration, les prix à l'exportation diffèrent notablement entre différents acheteurs, régions ou périodes, et si une explication est donnée quant à la raison pour laquelle il n'est pas possible de prendre dûment en compte de telles différences en utilisant les méthodes de comparaison moyenne pondérée à moyenne pondérée ou transaction par transaction.401
5.201. By referring to "[les] prix de transactions à l'exportation prises individuellement", which translates literally in English as "the prices of export transactions taken individually", the French text of the second sentence of Article 2.4.2 puts emphasis on the selection of individual transactions.
5.202. Thus, I believe that allowing an investigating authority to zero within the "pattern" under the second sentence of Article 2.4.2 not only is a permissible interpretation within the meaning of the second sentence of Article 17.6(ii) of the Anti-Dumping Agreement, but that it is a more defensible interpretation within the meaning of the first sentence of that provision.
5.203. For these reasons, I disagree with the finding of the majority that zeroing within the "pattern" under the W-T comparison methodology of the second sentence of Article 2.4.2 of the Anti-Dumping Agreement is not permissible. Consequently, I also disagree with the findings of the majority on zeroing under Article 2.4 of the Anti-Dumping Agreement and under Article 9.3 of the Anti-Dumping Agreement and Article VI:2 of the GATT 1994.
So the U.S. still loses on this issue under the majority opinion, and every kind of zeroing continues to be found in violation of WTO obligations, but there is a formal dissent on the Appellate Body record in support of this type of zeroing.
Here's my question: Will prospective AB Members be asked about this issue by governments during the selection process? If so, what is the best way to answer?