A WTO Panel Openly Rejects the Appellate Body's "Zeroing" Case Law
Although most economists criticize antidumping measures as protectionist, WTO norms allow its members to impose extra duties on those exports made at less than fair value (dumped goods) if such dumping causes a material injury to domestic producers. The difference between the fair value and the prices applied in the importing country is called the "dumping margin." "Zeroing" refers to an asymmetrical calculative methodology, adopted by antidumping authorities in the course of obtaining final dumping margins, which excludes any negative margins occurring when export prices exceed normal values (such as home prices) and thus only includes positive margins occurring when home prices exceed export prices. Since the old GATT era, zeroing has been controversial because it tends to inflate final dumping margins by preventing those negative margins from offsetting positive margins. Although a GATT panel upheld this practice, GATT contracting parties (and Uruguay Round negotiators) failed to provide any clear rules on the practice.
The WTO Appellate Body (AB), in a series of decisions, has consistently struck down the zeroing practice. The AB's anti-zeroing jurisprudence has caused further controversies. For example, some members, including the United States, have expressed their strong opposition to this case law and sought to change it through negotiation. Moreover, in a surprising move, a recent WTO panel report (U.S. - Stainless Steel) explicitly defied the AB's jurisprudence and ruled in favor of certain types of zeroing. This panel's position raises an important legal question. One might argue that a WTO panel could not disobey the AB's case law on the same subject-matter. In this case, the complainant (Mexico) appealed the panel decision to the AB, which is likely to reject the panel ruling, if it still adheres to its case law in this area. If it had not been appealed, however, two conflicting case laws, i.e., one by the AB and the other by the panel, would have co-existed.
What is "Zeroing"?
In obtaining final dumping margins, antidumping authorities, such as the U.S. Department of Commerce (DOC), aggregate results of each computation (home price minus export price) as to each sub-product category (each model or type) of the same product under investigation. Yet under the zeroing methodology the DOC omits (zeroes) negative results of calculation occurring when export prices exceed home prices (normal values). Therefore, the final dumping margins tend to be bigger under zeroing than otherwise because zeroing prevents negative margins from offsetting positive ones in aggregation. In an example highly simplified for the sake of illustration, suppose that China exports two widgets (of different types) to the U.S. market. The price of the same widget in China (normal value) is $1. Widgets in the first and the second type are priced $0.5 and $1.5, respectively. Without zeroing, the final dumping margin (the sum of individual margins of comparison divided by a total sale) is 0% [(0.5-1.0) + (1.5-1.0) / (0.5 + 1.5)]. With zeroing, however, it is 25% [0 + (1.5-1.0) / (0.5 + 1.5)].
The WTO AB Has Continued to Strike Down Zeroing
For those who advocate zeroing, a "dumping margin" under the WTO Antidumping Agreement could be established "for each product type or for each individual transaction" as well as for the product as a whole. However, the AB rejected this position and ruled that the dumping margin should be established "for the product...and not for the various types or models of that product." Because zeroing picks and chooses positive results of these intermediate calculations in the situation of multiple comparisons and disregards (zeroes) negative ones, it does "not take into account the entirety of the prices of some export transactions" and thus "inflates the margin of dumping for the product as a whole." The AB viewed that this failure violated the "fair comparison" obligation under Article 2 of the Antidumping Agreement.
The AB's invalidation of the zeroing practice has been quite sweeping. It struck down zeroing employed in every different comparison context, such as the weighted-average-to-weighted-average comparison, the weighted-average-to transaction comparison, and the transaction-to-transaction comparison. The AB ruled that zeroing was also illegal in the "administrative review" process, besides in the original investigation process. An administrative review refers to a process under which upon the request of interested parties the antidumping authority (DOC) annually calculates the amount of antidumping duties owed by each individual importer by comparing the price of each individual export transaction with a monthly average normal value. For the same reason discussed above, the AB illegalized zeroing adopted in the administrative review process. 
There has been an incremental expansion in the AB's anti-zeroing jurisprudence, which resulted from complainants' strategic use of "as such" complaints. In an early stage, complaining parties challenged specific applications of the zeroing methodology ("as applied") by defending parties, as in EC - Bed Linen. However, as the AB's anti-zeroing rulings gathered momentum, complainants began to challenge the zeroing policy itself ("as such"). These "as such" challenges have been successful. In both U.S. - Zeroing (EC) and U.S. - Zeroing (Japan), the AB accepted these wholesale complaints and ruled in favor of complainants (EC and Japan). As a result, the U.S. is now obligated to repeal its zeroing policy in its entirety nearly on all fronts, both in the original investigations and in the administrative (periodic) reviews.
The Panel in U.S. - Stainless Steel Has Defied the AB's Zeroing Case Law
Despite the aforementioned AB case law invalidating zeroing, a recent panel in U.S. - Stainless Steel explicitly rejected the AB's established anti-zeroing position, in particular positions expressed in U.S. - Zeroing (EC) and U.S. - Zeroing (Japan), both of which addressed the same issue as the panel faced, i.e., a "simple zeroing" in the administrative (periodic) review. The panel acknowledged that the AB "de facto expects" the panel to follow adopted AB reports "to the extent that the legal issues are similar." However, it emphasized that panels "are not, strictly speaking, bound by previous Appellate Body or panel decisions that have addressed the same issue." Interestingly, it found support for its position in Article 19.2 of the Dispute Settlement Understanding (DSU) which prohibits the panel and the AB from "adding to or diminishing" WTO members' rights and obligations. It also viewed that its reversal of the AB's position in this issue is in pursuit of its obligation of an "objective examination" under Article 11 of the DSU.
The panel followed the same line of reasoning that two previous panels (U.S. - Zeroing (EC) and U.S. - Zeroing (Japan)) had employed, but the AB had ultimately rejected in both cases, although the panel emphasized that its analysis was "not simply an unthinking repetition of these past panel decisions." The fundamental difference between the panel and the AB on this issue centers on how to define "dumping" or "dumping margins" under the WTO Antidumping Agreement and GATT Article VI. According to the AB, dumping and dumping margins must be defined in terms of the "product as a whole" under investigation. Therefore, all intermediate, individual calculations (normal value minus export price) as to various sub-product categories must be "aggregated" to obtain the final dumping margin. In stark contrast, the panel viewed that dumping and dumping margin could be defined in terms of "a particular export sale" and that antidumping authorities need "not necessarily require an examination of different export sales at an aggregate level." The panel viewed that the AB's interpretation was not based on a "solid textual basis" and that it should adopt a "permissible interpretation" which accommodates a certain type of zeroing (simple zeroing) in this case.
The panel's unusual behavior raises many interesting legal questions. As it admitted itself, the AB expects panels to follow its jurisprudence, even in the absence of formal binding force. More importantly, such jurisprudence creates "legitimate expectations" among WTO members. It seems controversial whether a panel could reject the AB's established case law on the same issues in the presence of Article 3.2 of the DSU which underscores the "stability and predictability" of the multilateral trading system. While the AB only renders its "recommendations" in a formal sense, this technical deficiency in legal bindingness may not necessarily strip the AB's case law of their genuine normative, regulative effects. WTO members cite, quote and reference the AB's precedents to substantiate and reinforce their legal positions in the dispute. Regardless of the label, members seem to perceive this rich set of precedents as well-established "jurisprudence" which they voluntarily observe.
Now that Mexico has appealed this decision, the AB is likely to reject the panel ruling and adhere to its case law, unless it changes its course. However, if Mexico had chosen not to appeal and the panel decision had thus been adopted unmodified, two conflicting case laws would have co-existed in this area.
Considering diametrically opposite views on zeroing among major WTO members as well as the inchoate stage of WTO negotiations on rules, any legislative breakthrough on zeroing, such as an amendment of relevant provisions (Articles 2 and 9) of the Antidumping Agreement, is highly unlikely in the near future. In the meantime, WTO members opposing zeroing are liable to continue to challenge this practice before the WTO dispute settlement system. Even in the U.S., voices against zeroing, and antidumping measures in general, have recently gathered strength, as major retailers lobby against zeroing and more U.S. exports are subject to foreign antidumping measures. On the other hand, advocates of zeroing will continue to push their case. Some U.S. politicians have vowed to link the zeroing negotiation to the fate of the current Doha trade talks.
After all, the zeroing controversy raises an important legal and institutional (to some, even "constitutional") question as to the limits of the WTO tribunal's interpretation or rule-making, in particular in the area of antidumping as well as the relationship between panels and the AB. Thus far, the AB has viewed that its anti-zeroing rulings are in accordance with the standard of review under Article 17.6 (ii) of the Antidumping Agreement. However, the zeroing supporters, such as the U.S., justify the practice by the deference which they view is granted under the same provision, or by a more abstract standard, i.e., sovereignty. Whichever position may prevail in the end, it will change the nature and contour of WTO norms to a great extent.
Update: The Appellate Body Has Reversed the Panel's Departure from the Zeroing Jurisprudence (June 6, 2008)
Since the panel in the Stainless Steel case issued its highly controversial report explicitly defying the well-established Appellate Body (AB) jurisprudence invalidating simple zeroing, the AB's response has been of great interest among international trade law scholars. In April 30, 2008 the AB has finally circulated its report one month later than its usual deadline.  As widely anticipated, the AB reversed the panel's findings on the Unites States' simple zeroing practice and invalidated this methodology both "as such" and "as applied."
As explained in more detail, to permit zeroing in periodic reviews means that an antidumping authority elects not to take into account all export transactions some of which may include negative (-) results of calculation (comparison) between home market prices and export prices, and thus prevents these negative results from offsetting any positive ones. If one interprets, as the panel did, that dumping can take place on a transaction-specific basis, then those transactions whose export prices exceed home market prices (and therefore are not dumped transaction) may be 'zeroed', i.e., regarded as if there were no dumping at all. Therefore, only those transactions which produce positive (+) results will be added to produce an importer-specific dumping margin. In other words, as the U.S. claimed, there can be "multiple" dumping margins for each importer.
The AB rejected this interpretation. First, based on the textual interpretation of GATT Articles VI:1, VI:2 and VI:6(a) as well as WTO Anti-Dumping Code Articles 2.1, 2.3, 3.4, and 5.1, it emphasized that dumping (and the dumping margin) is an "export-specific" concept which should be defined in terms of a product as a whole. The AB also based its ruling on the "context" found in various other related provisions of the WTO Anti-Dumping Code, such as Articles 5.2(ii), 5.8, 6.1.1, 6.7, 6.10, 8.1, 8.2, 9.4, 9.5 and 11. Interestingly, the AB confirmed that both French and Spanish versions of Article 6.10 of the WTO Anti-Dumping Code represent one single dumping margin ("une marge" and "el margen," respectively).
More importantly, the AB highlighted that the AB reports' technical non-bindingness relied upon by the panel in departing from the AB's jurisprudence does not render subsequent panels freedom to disregard previous reports. Citing previous reports, it viewed that adopted AB reports create "legitimate expectations" among WTO members and that panels' observance with those reports would also be expected. Then, the AB justified its position with a critical observation on the value of "jurisprudence" within the WTO. Yet, the AB still left an exit for future panels to depart from its case law in the presence of "cogent reasons."
Finally, the AB did not forget to admonish the panel's unusual behavior. With a solemn tone, it emphasized that the panel's departure from the well-established AB jurisprudence in the same legal issue is against the hierarchical division of labor in DSU under which only the AB can "uphold, modify or reverse" panels' legal interpretations. Although the AB expressed its deep concern over the panel's rebellious behavior, it still refused to find that the panel violated Article 11 of DSU ("objective assessment") on the ground that the AB had already corrected the panel's errors.
About the Author
Sungjoon Cho, an ASIL member, is Assistant Professor of Law at Chicago-Kent College of Law, Illinois Institute of Technology. He currently serves as Co-Vice Chair of the ASIL International Economic Law Group. During the period of 1994-96, he represented the Korean government in negotiations under the WTO and the Organization for Economic Cooperation and Development (OECD). He is the author of Free Markets and Social Regulation: A Reform Agenda of the Global Trading System (Kluwer Law International, 2003), and The Law of the World Trade Organization through the Cases (2007) (with J. H. H. Weiler and I. Feichtner), http://www.jeanmonnetprogram.org/wto/Units/index.html.
About the International Economic Law Interest Group
The International Economic Law Interest Group promotes academic interest, discussion, research and publication on subjects broadly related to the transnational movement and regulation of goods, services, persons and capital. International law topics include trade law, economic integration law, private law, business regulation, financial law, tax law, intellectual property law and the role of law in development. Click here to learn more about the International Economic Law Interest Group.
 See generally Sungjoon Cho, The WTO Appellate Body Strikes Down the U.S. Zeroing Methodology Used in Antidumping Investigations, ASIL INSIGHTS (May 4, 2006).
 EC - Anti-Dumping Duties on Audio Tapes in Cassettes Originating in Japan, ADP/136, Apr. 28, 1995 (unadopted).
 United States - Final Anti-Dumping Measures on Stainless Steel from Mexico, Report of the Panel, WT/DS344/R, Dec. 20, 2007, para. 7.106 ("[W]e have decided that we have no option but to respectfully disagree with the line of reasoning developed by the Appellate Body regarding the WTO-consistency of simple zeroing in periodic reviews.").
 European Communities - Anti-Dumping Duties on Imports of Cotton-Type Bed Linen from India, Appellate Body Report adopted on Mar. 12, 2001, WT/DS141/AB/R, para. 12 (emphasis original) [hereinafter EC - Bed Linen (2001)].
 Id., para. 53 (emphasis original)
 United States - Final Dumping Determination on Softwood Lumber from Canada, Appellate Body Report adopted on Aug. 31, 2004, WT/DS264/AB/R, paras. 98, 101 (emphasis original).
 EC - Bed Linen (2001), supra note 4, para. 59
 United States - Final Dumping Determination on Softwood Lumber from Canada (Article 21.5 - Canada), Appellate Body Report adopted on Sep. 1, 2006, WT/DS264/AB/RW, para. 36.
 United States - Measures Relating to Zeroing and Sunset Reviews, Appellate Body Report adopted on Jan. 23, 2007, WT/DS322/AB/R, para. 169. [hereinafter U.S. - Zeroing (Japan)].
 United States - Laws, Regulations, and Methodology for Calculating Dumping Margins ("Zeroing"), Appellate Body Report adopted on May 9, 2006, WT/DS294/AB/R, para. 109 [hereinafter U.S. - Zeroing (EC)].
 U.S. - Zeroing (EC), supra note 11, paras. 205, 222; U.S. - Zeroing (Japan), supra note 10, paras. 147, 169.
 U.S. - Stainless Steel, supra note 3, paras. 7.106, 7.115. A "simple zeroing" refers to the zeroing practice adopted under "weighted average-to-transaction" (W-T) or transaction-to-transaction (T-T) comparisons between export price and normal value. The simple zeroing is often conducted in the administrative (periodic) review which starts after a year from the publication of antidumping duties. In contrast, the zeroing practice under weighted average-to-weighted average comparisons is called a "model zeroing."
 Id., para. 7.105.
 Id., para. 7.102.
 Id., para. 7.115.
 Id., paras. 7.113, 7.117.
 Id., para. 7.117.
 Id., para. 7.119.
 Id., para. 7.103.
 United States - Final Anti-Dumping Measures on Stainless Steel from Mexico, Notification of an Appeal by Mexico, WT/DS344/7, Feb. 4, 2008.
 See Jonathan Lynn, Anti-Dumping Row Roils WTO, Isolates U.S., REUTERS, Jan. 10, 2008 (quoting Brendan McGivern who observed that "it's wildly optimistic of the U.S. to think they'll get this back through negotiations"). See also Proposal on Prohibition of Zeroing, Paper from Brazil; Chile; Columbia; Costa Rica; Hong Kong, China; Israel; Japan; Korea; Mexico; Norway; the Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu; Singapore; Switzerland and Thailand, TN/RL/W/113, Jun. 6, 2003; Offsets for Non-Dumped Comparisons, Communication from the United States, TN/RL/W/208, Jun. 5, 2007 (p2); WTO, Negotiating Group on Rules, Draft Consolidated Chair Texts of the AD and SCM Agreements, TN/RL/W/213, Nov. 30, 2007; The Office of the United States Trade Representative, Joint Statement by the Office of the United States Trade Representative and the Department of Commerce's International Trade Administration, Nov. 30, 2007 (stating that the U.S. was "very disappointed with important aspects of this draft text"); Statement on "Zeroing" in the Anti-Dumping Negotiations, Statement of Brazil; Chile; China; Colombia; Costa Rica; Hong Kong, China; India; Indonesia; Israel; Japan; Korea, Rep. of; Mexico; Norway; Pakistan; Singapore; South Africa; Switzerland; Separate Customs Territory of Taiwan, Penghu, Kinmen and Matsu; Thailand; and Viet Nam, TN/RL/W/214/Rev.3, Jan. 25, 2008.
 See e.g., United States - Measures Relating to Shrimp from Thailand, Request for Consultations by Thailand, WT/DS343/1, Apr. 27, 2006; United States - Continued Existence and Application of Zeroing Methodology, Request for Consultation from the European Communities, WT/DS350/1, Oct. 2, 2006.
 See U.S., Other Countries Clash over Anti-Dumping Rules in Chair's Text, BRIDGES WEEKLY TRADE NEWS DIGEST (vol. 11, no. 44), Dec. 19, 2007; see also Lynn, supra note 24 (reporting that "the U.S. is now the third biggest target of anti-dumping actions").
 See WTO "Zeroing" Decisions Should Be Part of Doha Talks, Senators Tell Administration, 24 INT'L TRADE REP. 1478, 1478 (2007).
 See e.g., U.S. - Zeroing (EC), supra note 11, para. 134 (ruling that "Article 9.3 of the Anti-Dumping Agreement, and ArticleVI:2 of the GATT 1994, when interpreted in accordance with customary rules of interpretation of public international law, as required by Article 17.6(ii), do not, in our view, allow the use of the methodology applied by the United States in the administrative reviews at issue.").
 See Roger P. Alford, Reflections on U.S. - Zeroing: A Study in Judicial Overreaching by the WTO Appellate Body, 45 COLUM. J. TRANSNAT'L L. 196, 199-202 (2006) (criticizing that AB violated the principle of deference which, according to him, was embedded in Article 17.6 of the Antidumping Agreement reflecting the U.S. Chevron doctrine).
 See Daniel K. Tarullo, Paved with Good Intentions: The Dynamic Effects of WTO Review of Anti-Dumping Action, 2 WORLD TRADE REV. 373, 374 (2003) (warning that the AB's judicial activism on zeroing might backfire as the U.S. reluctance to trade concessions in the subsequent trade negotiations).
 United States - Final Anti-Dumping Measures on Stainless Steel from Mexico, WT/DS344/AB/R, Appellate Body Report circulated on Apr. 30, 2008.
 Id., para. 77.
 Id., para. 80.
 Id., paras. 83-86.
 Id., paras. 87-93.
 Id., para. 158.
 Id., para. 159.
 Id., para. 160.
 Id., para. 161.
 Id., para. 162.