tutorial video tutorial video Discover the CiteMap in 3 minutes
Source(s) of the information:
Source(s) of the information:

Report of the Panel

I. INTRODUCTION

1.1.
On 18 November 1999, Japan requested consultations with the United States under Article 4 of the Understanding on Rules and Procedures Governing the Settlement of Disputes (the DSU), Article 17.2 of the Anti-Dumping Agreement and Article XXII:1 of the General Agreement on Tariffs and Trade 1994 ("GATT 1994").1 The United States and Japan consulted on 13 January 2000, but failed to settle the dispute.
1.2.
On 11 February 2000, Japan requested the establishment of a panel pursuant to Article XXIII of the GATT 1994, Articles 4 and 6 of the DSU and Article 17 of the Anti-Dumping Agreement.2
1.3.
At its meeting on 20 March 2000, the Dispute Settlement Body (DSB) established a Panel in accordance with the request made by Japan in document WT/DS184/2. At that meeting, the parties to the dispute also agreed that the Panel should have standard terms of reference. The terms of reference are, therefore, the following:

"To examine, in the light of the relevant provisions of the covered agreements cited by Japan in document WT/DS184/2, the matter referred to the DSB by Japan in document WT/DS184/2, and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in those agreements".

1.4.
On 9 May 2000, Japan requested the Director-General to determine the composition of the Panel, pursuant to paragraph 7 of Article 8 of the DSU. On 24 May 2000, the Director-General composed the Panel as follows3:

Chairman: Mr. Harsha V. Singh

Members: Mr. Yanyong Phuangrach

Ms. Elena Lidia di Vico

1.5.
Brazil, Canada, Chile, the European Communities and Korea reserved their rights to participate in the panel proceedings as third parties.
1.6.
The Panel met with the parties on 22-23 August 2000 and on 27 September 2000. It met with the third parties on 23 August 2000.
1.7.
The Panel submitted its interim report to the parties on 22 January 2001.

II. FACTUAL ASPECTS

2.1.
This dispute concerns the imposition by the United States of anti-dumping measures on imports of certain hot-rolled flat-rolled-carbon-quality steel products ("hot-rolled steel") from Japan.
2.2.
On 30 September 1998, several US steel manufacturing companies, the United Steelworkers of America, and the Independent Steelworkers Union filed petitions for the imposition of anti-dumping duties on imports of certain hot-rolled steel products from Brazil, Japan, and Russia.4 The petitions also alleged that critical circumstances existed with regard to imports from Japan. Effective 30 September 1998, the United States International Trade Commission ("USITC") instituted its investigation to determine whether there is a reasonable indication that an industry in the United States is materially injured or threatened with material injury, or the establishment of an industry in the United States is materially retarded by reason of imports from the three countries of certain hot-rolled steel products that are alleged to be sold in the United States at less than fair value.5
2.3.
After an examination of the information presented in the petition filed with respect to hot-rolled steel from Japan and the amendments thereto, the United States Department of Commerce ("USDOC") initiated an anti-dumping duty investigation on 15 October 1998.6 USDOC determined that it was not practicable to examine all known producers/exporters and conducted its investigation on the basis of a sample of Japanese producers. Based on information concerning production volumes from all six Japanese producers, Kawasaki Steel Corporation ("KSC"), Nippon Steel Corporation ("NSC"), and NKK Corporation ("NKK") were selected for individual investigation and calculation of a dumping margin (i.e, the "investigated respondents"), as these three companies accounted for more than 90 per cent of all known exports of the subject merchandise during the period of investigation.
2.4.
Effective 16 November 1998, USITC issued an affirmative preliminary determination, finding a reasonable indication that the US industry was threatened with material injury by reason of hot-rolled steel imports from Brazil, Japan, and Russia.7
2.5.
Effective 30 November 1998, USDOC issued its affirmative preliminary critical circumstances determination, finding that there is a reasonable basis to believe or suspect that critical circumstances exist for imports of hot-rolled steel from Japan and Russia. USDOC also determined not to make a preliminary determination of critical circumstances with respect to imports from Brazil. Based on its determination, USDOC stated that, upon issuance of an affirmative preliminary dumping determination, Commerce would direct the US Customs Service to suspend liquidation of all entries of Japanese hot-rolled steel for a period of ninety days prior to the preliminary dumping determination.8 No specific measures were put into effect at this stage.
2.6.
Effective 19 February 1999, USDOC issued a preliminary affirmative dumping determination, finding that hot-rolled steel from Japan was sold in the United States at dumped prices.9 USDOC calculated the following preliminary margins of dumping:

KSC 67.59%

NSC 25.14%

NKK 30.63%

All Others Rate 35.06%.

The "All Others" rate, applicable to companies not investigated, was calculated as the weighted average of the margins calculated for the three investigated respondents. Pursuant to its earlier critical circumstances finding, USDOC ordered suspension of liquidation and posting of cash deposits or bonds for entries made 90 days prior to the 19 February 1999 effective date of the preliminary determination of dumping, that is, retroactive to 21 November 1998.10

2.7.
Following its preliminary affirmative dumping determination, USDOC issued several more requests for information, conducted verification at the three investigated respondents’ offices in Japan (and the US in some cases), received interested party comments, and held a public hearing on 21 April 1999. On 6 May 1999, USDOC published its final determination that respondents were selling hot-rolled steel in the United States at the following margins of dumping:

KSC 67.14%

NSC 19.65%

NKK 17.86%

All Others Rate 29.30%.11

USDOC also made a final negative determination of critical circumstances as to NSC and NKK based on the fact that they had final dumping margins below the 25 per cent threshold used to impute importer knowledge of dumping. However, USDOC continued to find that critical circumstances existed as to KSC and the "all others" companies.

2.8.
Following USDOC's preliminary determination of dumping, and while USDOC was conducting the final dumping investigation, USITC instituted and conducted the final injury investigation. Following collection of information, submission of briefs by interested parties and a public hearing held on 4 May 1999, USITC voted unanimously on 11 June 1999, that the US industry was materially injured or threatened with material injury by reason of hot-rolled steel imports from Japan.12 On 23 June 1999, USITC published its final affirmative determination of injury. USITC also made a negative determination with respect to critical circumstances, concluding that the increase in imports in a short period of time was not sufficient to warrant a finding that the imports would undermine the remedial effects of the anti-dumping duty order.13
2.9.
On 29 June 1999, USDOC published an anti-dumping duty order imposing estimated dumping duties on imports from Japan at the rates announced in its final determination.14 Since USITC had not found critical circumstances to exist, USDOC ordered the refund of any cash deposits and/or release of any guarantees provided for the period of the preliminary critical circumstances finding, 21 November 1998 - 19 February 1999.

III. PARTIES' REQUESTS FOR FINDINGS AND RECOMMENDATIONS

A. JAPAN

3.1.
Japan requests that the Panel:

(a) find that the specific anti-dumping measures imposed by the United States on hot-rolled steel from Japan are inconsistent with various provisions of the AD Agreement, as follows:

· USDOC’s application of adverse facts available to KSC’s dumping margin was inconsistent with Articles 2.3, 6.8, 9.3, and Annex II;

· USDOC’s application of adverse facts available and treatment of the facts with respect to NKK’s dumping margin were inconsistent with Articles 2.4, 6.1, 6.6, 6.8, 6.13, 9.3, and Annex II;

· USDOC’s application of adverse facts available and treatment of the facts with respect to NSC’s dumping margin were inconsistent with Articles 2.4, 6.6, 6.8, 6.13, 9.3, and Annex II;

· USDOC’s inclusion of margins based on partial facts available in the calculation of the "all others rate" was inconsistent with Article 9.4;

· USDOC’s exclusion and replacement of certain home market sales in the calculation of normal value through use of the 99.5 per cent arm’s length test was inconsistent with Articles 2.1, 2.2, and 2.4;

· USDOC’s application of a new policy with respect to preliminary critical circumstances determinations was inconsistent with Articles 10.1, 10.6, and 10.7;

· USITC’s application of the captive production provision was inconsistent with Articles 3.1, 3.2, 3.4, 3.5, 3.6 and 4.1;

· USITC’s finding of a causal connection between imports and the domestic industry’s injury was inconsistent with Articles 3.1, 3.4, and 3.5;

and to recommend that the DSB request the United States to bring these measures into conformity with the AD Agreement.

(b) find that the following actions undertaken by the United States were inconsistent with GATT 1994 Article X:3, including:

· USDOC’s accelerated proceeding;

· USDOC’s application of a revised critical circumstances policy;

· USDOC’s failure to correct, prior to the final determination, the clerical error committed in calculating NKK's preliminary margin;

· USDOC’s resort to adverse facts available with respect to respondents, coupled with USDOC’s and USITC’s decisions against applying facts available with respect to petitioners;

· USITC’s limited analysis to two years of the three-year period of investigation, in abandonment of its normal policy to analyze all three years;

and to recommend that the DSB request the United States to bring these actions into conformity with the GATT 1994;

(c) find that the United States’ anti-dumping laws, regulations, and administrative procedures governing:

· the use of adverse "facts available" are inconsistent with Article 6.8 and Annex II of the AD Agreement;

· the calculation of an "all others" rate based on partial facts available are inconsistent with Article 9.4 of the AD Agreement;

· the exclusion and replacement of certain home market sales in the calculation of normal value by the arm's length test are inconsistent with Articles 2.1, 2.2, and 2.4 of the AD Agreement;

· "critical circumstances," including the generally applicable interpretations reflected in the Policy Bulletin issued on 8 October 1998, are inconsistent with Articles 10.1, 10.6 and 10.7 of the AD Agreement;

· the focus on the merchant market sales to the exclusion of the remainder of the domestic industry when determining injury by reason of imports are inconsistent with Articles 3.1, 3.2, 3.4, 3.5, 3.6, and 4.1 of the AD Agreement;

and recommend that the DSB request the United States to ensure, as stipulated in Article XVI:4 of the WTO Agreement and Article 18.4 of the AD Agreement, the conformity of the above-listed elements of its anti-dumping laws, regulations, and administrative procedures with its obligations under the AD Agreement;

(d) recommend that, if the Panel's findings result in a determination that the imported product was either not dumped or that it did not injure the domestic industry, the DSB further request that the United States revoke its anti-dumping duty order and reimburse any anti-dumping duties collected;15

(e) recommend that, if the Panel's findings result in a determination that the imported product was dumped to a lesser extent than the duties actually imposed, the DSB further request that the United States reimburse the duties collected to the extent of the difference.

B. UNITED STATES

3.2.
The United States requests the Panel to find that:

· the information submitted to this Panel by Japan that was not made available to US authorities during the course of the anti‑dumping investigation at issue will be disregarded in this proceeding;

· Japan’s claim concerning the United States’ general practice with respect to "facts available" was not raised in Japan’s request for the establishment of a panel and is therefore not included in this Panel’s terms of reference;

· the specific anti‑dumping measures imposed by the United States on hot‑rolled steel from Japan are consistent with the provisions of the Anti‑Dumping Agreement identified by Japan under point (a);

· none of the actions identified by Japan under point (b) was inconsistent with Article X:3 of the GATT 1994;

· the United States’ anti‑dumping laws, regulations, and administrative procedures governing the issues identified by Japan under point (c) are not inconsistent with the provisions of the Anti‑Dumping Agreement identified in that paragraph.

· The specific remedies requested by Japan in its first submission, reproduced at points (d) and (e) above, are contrary to established practice and the DSU.

IV. ARGUMENTS OF THE PARTIES

4.1.
The arguments of the parties are set out in their submissions to the Panel (see Annexes, as listed above).

V. ARGUMENTS OF THE THIRD PARTIES

5.1.
The arguments of the third parties, Brazil, Canada, Chile, the European Communities and Korea, are set out in their submissions to the Panel (see Annexes, as listed above).

VI. INTERIM REVIEW

6.1.
Both parties filed comments on the interim report on 29 January 2001. The parties' comments were limited to the identification of clerical errors. Neither party requested an interim review meeting.
6.2.
In response to the parties' comments, the Panel corrected typographical and other clerical errors throughout the Report, and also corrected typographical and other clerical errors it had itself identified, consistent with WTO editorial standards.

VII. FINDINGS

A. PRELIMINARY OBJECTIONS

7.1.
The United States makes two preliminary objections.16 The United States requests that certain evidence presented for the first time before this Panel be disregarded, and objects, as falling outside the Panel's terms of reference, to a claim made by Japan concerning the US "general practice", including statutory and regulatory provisions, regarding the use of adverse facts available.

1. Exclusion of certain evidence

(a) Arguments

7.2.
The United States claims that evidence which was submitted by Japan during this proceeding, but which was not before the investigating authority during the anti-dumping investigation, may not be examined by the Panel. The United States, relying on Article 17.5(ii) of the AD Agreement, argues that we are to examine the decisions of the investigating authorities on the basis of the facts that were available to them and not on the basis of new facts revealed for the first time before the Panel. Consequently, the United States submits that we should disregard in toto four affidavits prepared for the purpose of these panel proceedings by the American attorneys of NSC, NKK, KSC and by one statistician, as well as numerous newspaper Articles that were not presented in the course of the investigation, or were presented to only one of the US authorities conducting the investigation.17 In this latter regard, the United States argues that we should disregard documents submitted by Japan concerning determinations made by the Commerce Department if those documents were not put on the Commerce Department administrative record, even if those documents were put on the USITC administrative record.
7.3.
The US argument is based in part on Article 17.5(ii) of the AD Agreement which provides that a panel shall examine the matter before it on the basis of "the facts made available in conformity with appropriate domestic procedures to the authorities of the importing Member" and the interpretation given to this provision by, inter alia, the Panel in Mexico- Anti-Dumping Investigation of High Fructose Corn Syrup from the United States18 ("Mexico-HFCS"). The United States argues that by presenting new testimony that was not before the appropriate authority, Japan seeks to have the Panel go beyond its mission under Article 17.6(i) of the AD Agreement to determine whether the establishment of the facts by the investigating authority was proper and its evaluation of those facts unbiased and objective.19 The United States further argues that the nature of the anti-dumping investigation itself directs that the Panel not consider extra-record evidence. Moreover, the United States submits, to allow only Japanese producers an opportunity to present new evidence would go against the guarantee expressed in Article 6.1 of the AD Agreement that all interested parties may present evidence.
7.4.
Japan submits that the United States has failed to provide adequate justification for the Panel to reject the challenged evidence and arguments. First, Japan disagrees with the US interpretation of Article 17.5 of the AD Agreement. Japan submits that Article 17.5(ii) provides that a panel shall base its examination on the facts made available in conformity with appropriate domestic procedures to the authorities of the importing Member. However, in Japan's view, Article 17.5(ii) does not say anything about what was or was not on the "administrative record". Japan argues that inherent in Article 17.5(ii) is the possibility that facts were "made available" to the investigating authority, but were for one reason or another not placed on the administrative record. Japan notes that Article 17.5(ii) does not refer only to information accepted by the authorities, and thus recognizes that information might be offered to the authorities but then inappropriately rejected. In Japan's view, Members must be permitted to submit evidence that explains or demonstrates how the authority's investigating procedures or determinations were unfair, unreasonable or biased. Japan maintains that such information will, more often than not, be "extra-record" evidence, since it is the investigating authority itself that determines what evidence is placed on the record. Japan argues that an authority cannot be permitted to exclude evidence inappropriately and then take advantage of the incomplete record to defend itself in the examination of its action by a WTO panel. In this regard, Japan rejects the US reliance on the distinction between the administrative records of USDOC and the USITC, arguing that information on the record of either authority may be relied upon before the Panel to challenge the determination of either authority.
7.5.
Japan also relies on statements made by the United States in other WTO proceedings concerning admissibility of amicus curiae briefs and other evidence which are contrary to its arguments in this case.20 Japan submits that the Panel should exercise its substantial discretion to accept evidence. Japan argues that the Appellate Body has made it clear that, based on Articles 12 and 13 of the DSU, it is the responsibility of the panel to determine the admissibility and relevance of the evidence proffered by the parties to a dispute.21 Japan submits that the Appellate Body statements on this matter are also valid with regard to this case since there is no conflict between Article 17.5(ii) of the AD Agreement and Articles 11 to 13 of the DSU and the provisions thus complement each other. Finally, Japan argues that the standard of Article 17.5(ii) of the AD Agreement only guides the Panel with regard to its review of actual investigations. Japan maintains that Article 17.5(ii) is not applicable to the evidence to be considered in connection with Japan's challenges to US statutory and regulatory provisions, that do not depend on the administrative record, and is also not applicable to the evidence to be considered in connection with Japan's claims under Article X:3 of GATT 1994.

(b) Finding

7.7.
The conclusion that we will not consider new evidence with respect to claims under the AD Agreement flows not only from Article 17.5(ii), but also from the fact that a panel is not to perform a de novo review of the issues considered and decided by the investigating authorities. We note that several panels have applied similar principles in reviewing determinations of national authorities in the context of safeguards under the Agreement on Safeguards and special safeguards under Article 6 of the Agreement on Textiles and Clothing. There is no corollary to Article 17.5(ii) in those agreements. Nonetheless, these panels have concluded that a de novo review of the determinations would be inappropriate, and have undertaken an assessment of, inter alia, whether all relevant facts were considered by the authorities.24 In that context, the Panel in United States - Definitive Safeguard Measures on Imports of Wheat Gluten from the European Communities ("United States– Wheat Gluten") recently observed that it was not the panel's role to collect new data or to consider evidence which could have been presented to the decision maker but was not.25
7.8.
Japan points out that it is the investigating authorities that control the receipt of information during the investigation, and thus could unjustly reject information submitted by a party, which a party might subsequently wish to present to a Panel reviewing the determination. This possibility raises an interesting question which is not really at issue before us. Japan has not made a claim that the anti-dumping measure is inconsistent with the provisions of the AD Agreement because the USITC or the USDOC wrongly rejected information submitted during the course of the investigation.26 Thus, the principal question presented by the United States' preliminary objection is whether we should exclude from our consideration in this dispute certain evidence that was not submitted to the US investigating authorities during the investigation.
7.9.
It is important to note that, in this case, Japan's claims are not limited to challenges under the AD Agreement to the final anti-dumping measure imposed by the United States. Japan also claims that certain US statutory provisions are inconsistent with the AD Agreement on their face, and claims that the United States did not administer its anti-dumping laws, regulations, decisions and rulings in a "uniform, impartial and reasonable manner", in violation of Article X of GATT 1994. There is no claim that the challenged evidence is relevant to the claims of inconsistency of certain statutes on their face.27 Japan does, however, argue that the challenged evidence is relevant to the claims under Article X of GATT 1994. In our view, the evidence to be considered in connection with Japan's Article X claim is not limited by the provisions of Article 17.5(ii) of the AD Agreement To the extent there are any limits to the evidence that may be considered in connection with Japan's claim under Article X of GATT 1994, these would derive from the provisions of the DSU itself, and not the AD Agreement.
7.10.
Under Article 13.2 of the DSU, Panels have a general right to seek information "from any relevant source". We note that, as a general rule, panels have wide latitude in admitting evidence in WTO dispute settlement.28 The DSU (as opposed to the AD Agreement) contains no rule that might be understood to restrict the evidence that panels may consider. Moreover, international tribunals are generally free to admit and evaluate evidence of every kind, and to ascribe to it the weight that they see fit. As one legal scholar has noted:

"The inherent flexibility of the international procedure, and its tendency to be free from technical rules of evidence applied in municipal law, provide the "evidence" with a wider scope in international proceedings…. Generally speaking, international tribunals have not committed themselves to the restrictive rules of evidence in municipal law. They have found it justified to receive every kind and form of evidence, and have attached to them the probative value they deserve under the circumstances of a given case".29

It seems to us that, particularly in considering allegations under Article X of GATT 1994, we should exercise our discretion to allow the presentation of evidence concerning the administration of the defending Members' anti-dumping laws, which might in any event go beyond the specific facts made available to the administering authority in accordance with appropriate domestic procedures during the course of a single anti-dumping investigation.

7.11.
This places us in the difficult situation of attempting to determine, at the outset, in the context of a preliminary objection, exactly which evidence is relevant to which of Japan's claims, and make exclusionary rulings ab initio. With respect to the newspaper articles the United States has challenged, we note that they may be relevant to Japan's claim of bias under Article X of GATT 1994, and therefore consider that it is not appropriate to exclude them at the outset. With respect to the attorneys' and statistician's affidavits, we note that while they contain certain factual statements, they also set out arguments and analysis in support of Japan's claims in this dispute, which may appropriately be brought before a panel. Thus, we have determined not to exclude the four affidavits, the newspaper articles, and the profit and web-site information contained in exhibits JP-16-23, 25-28, 32(a) - 32(f), 33, 34-38, 44, 46, 56, 105, and note 353 of Japan's second written submission. To the extent that these exhibits purport to present facts relating to the USDOC or USITC determinations different from or additional to those that were made available to those authorities in conformity with appropriate domestic procedures during the course of the investigation, we have not taken such facts into account in our review of those determinations.
7.12.
There is, however, a significant distinction between questions concerning the admissibility of evidence, and the weight to be accorded to the evidence in making our decisions. That we have concluded that it is not appropriate to exclude from this proceeding at the outset evidence put forward by Japan has no necessary implications concerning the relevance or weight of that evidence in our ultimate determinations on the substantive claims before us. Moreover, we wish to emphasize that we have conducted our examination of the challenged final anti-dumping measure and the underlying determinations of the USDOC and USITC in strict observance of the requirements of Article 17.5(ii).

2. Claim allegedly not within the Panel's terms of reference

(a) Arguments

7.13.
The United States asserts that Japan's claim that the USDOC's "general practice" concerning adverse facts available violates the AD Agreement was not made in the request for establishment of a panel. The United States maintains that Japan did not refer USDOC's "general practice" regarding facts available, which is based on the US statute, to the DSB, but only referred the specific application of this practice to the companies involved in the investigation underlying this dispute. Thus, in the US view, the Japanese claim regarding the USDOC's "general practice" on facts available falls outside the Panel's terms of reference.
7.14.
The United States notes that Japan made clear in its request for establishment of a panel those instances in which it challenged both the law on its face and the specific application of statutory provisions in the underlying investigation. The United States asserts that the broad counts of conformity under Article XVI:4 of the Marrakesh Agreement and Article 18.4 of the AD Agreement set out in section E of the request for establishment do not lead to the conclusion that USDOC's general practice on facts available, which was not mentioned in the panel request, may now be challenged. The United States finds support for its claim in the Appellate Body's statement in Korea – Definitive Safeguard Measures on Imports of Certain Dairy Products30 ("Korea-Dairy Safeguard") that "any claim that is not asserted in the request for the establishment of a panel may not be submitted at any time after submission and acceptance of that request"and similar statements in Guatemala – Anti-Dumping Investigation Regarding Portland Cement from Mexico31("Guatemala – Cement I").
7.15.
Japan maintains that it properly made a claim concerning USDOC's general practice regarding adverse facts available. Japan explains that it did not challenge US statutory provisions in this respect but rather the manner in which these provisions have been applied in practice by USDOC.32 Japan submits that its claim is based on Article XVI:4 of the WTO Agreement, and is clearly set out in section "E. CONFORMITY" of the request for establishment.33 Japan argues that by including in this section of its panel request a reference to all "above-detailed laws, regulations, and administrative rulings" and explicitly claiming them to be inconsistent with Article XVI:4 of the WTO Agreement, it made it perfectly clear to both the United States and interested third countries that the matter it was submitting to the DSB comprised not only the actions taken in the specific case but also the US anti-dumping law on which these actions were based, including the law governing the application of facts available as interpreted and applied by USDOC, which constitutes the "general practice" on facts available. Finally, Japan argues that the United States failed to demonstrate how Japan's panel request has prejudiced the United States' ability to defend itself.34 Japan therefore requests the Panel to reject the US preliminary objections.

(b) Finding

7.16.
We consider that the United States' preliminary objection raises two separate but related issues. First, has Japan identified as a measure at issue in this dispute the US "general practice" concerning facts available ? Second, and assuming that Japan has not identified the US "general practice" in this regard as a separate measure in dispute, has Japan, in the context of its challenge to the definitive anti-dumping measure, stated a claim regarding the "general practice" concerning facts available with sufficient clarity, consistent with Article 6.2 of the DSU.
7.17.
The Appellate Body has made it clear that a matter referred to the DSB consists of a measure and the claims concerning that measure.35 In this dispute, it is clear that Japan has raised the final anti-dumping measure as a measure at issue. Japan has also identified certain provisions of US laws and regulations as measures at issue -- these provisions are specifically identified in paragraphs A.3 (law governing calculation of the all others rate), A.5 (law governing preliminary critical circumstances determinations), and B.2 (law regarding treatment of captive production in injury analysis) of its request for establishment. However, based on our review of the request for establishment, we do not see that Japan has raised the US "general practice" regarding facts available as a measure at issue in this dispute.
7.18.
The "general practice" regarding which Japan asserts it has raised a claim is the USDOC's practice of, when applying adverse facts available, looking for facts that are "sufficiently adverse" to accomplish the goal of inducing respondents to provide complete and accurate information. This practice, while it is based on the US statute, is not explicitly set out in either the US statute, regulation, or any other binding policy statement of USDOC. Rather, it has been set out in the determination in this and other investigations to explain the USDOC's choice regarding the particular facts available it will consider in making its determination. Even assuming that a claim regarding the consistency of a "general practice" can be made in the WTO dispute settlement system, we are of the opinion that the request for establishment in such a case must identify such practice with sufficient clarity.36
7.19.
The US "general practice" concerning facts available is not identified on the face of the request for establishment as a measure in dispute. Japan has explicitly acknowledged that it has not challenged the US statute governing the application of facts available.37 Japan argues that its claim concerning the conformity of the US anti-dumping laws, regulations, and administrative rulings, set out in paragraph E of its request for establishment, necessarily must be understood to include a challenge to the "general practice" in question.38
7.20.
Japan did not separately set forth in the request for establishment an assertion specifically with respect to USDOC's general practice on facts available (or the statutory and regulatory provisions underlying that practice). Indeed, the phrase "general practice" does not appear in the request for establishment at all. Nor is there mention of the USDOC's interpretation or application of the statutory provisions regarding facts available in general, as opposed to its decision to apply facts available in this case. Moreover, the very fact that the other aspects of US law which are challenged on their face are spelled out in the request for establishment, as set out in paragraph 7.17 above, would lead the reader to conclude that there is no such challenge to the general practice regarding application of adverse facts available. Thus, in our view, the request for establishment does not identify USDOC's "general practice" regarding application of facts available as a measure in dispute.
7.21.
Nor can we conclude, as Japan would apparently have us do, that the general claim regarding "Conformity" set out in paragraph E of the request for establishment is sufficient to bring the "general practice" on facts available before us. That claim asserts that, by maintaining "the above-detailed laws, regulations and administrative rulings of general application" which are allegedly not in conformity with its obligations under the WTO Agreements, the United States has acted inconsistently with Article XVI:4 of the Marrakesh Agreement, as well as Article 18.4 of the AD Agreement. These two provisions generally require that Members bring their laws and regulations into conformity with the WTO Agreements. The "general practice" regarding application of facts available is not identified among the "laws, regulations and administrative rulings of general application" detailed in preceding sections of the request for establishment. The specific section of the request for establishment addressing the application of facts available, paragraph A.2, does not refer to an inconsistency in the statute, regulations, policy or "general practice" regarding application of facts available, but to the determination regarding the application of facts available under the applicable statute. We do not find that this statement is sufficient to bring into this dispute USDOC's "general practice" regarding the application of facts available. To conclude otherwise would effectively allow a Member to challenge all statutes, regulations, and "general practices" in the context of a challenge to a measure imposed pursuant to such provisions or a challenge to any one of such provisions. Such a ruling would eviscerate the obligation to set forth, in the request for establishment, with sufficient specificity, the challenged measure or measures, and the claims regarding such measure or measures.
7.22.
The Appellate Body has noted

"As a panel request is normally not subjected to detailed scrutiny by the DSB, it is incumbent upon a panel to examine the request for the establishment of the panel very carefully to ensure its compliance with both the letter and the spirit of Article 6.2 of the DSU. It is important that a panel request be sufficiently precise for two reasons: first, it often forms the basis for the terms of reference of the panel pursuant to Article 7 of the DSU; and, second, it informs the defending party and the third parties of the legal basis of the complaint." (emphasis added)39

In this case, we conclude that Japan has failed to state a claim at all with respect to the "general practice" of the USDOC concerning application of facts available. Assuming such practice could be challenged separately from a challenge to the statutory provision on which it is based, Japan has failed to present this problem in the request for establishment in this dispute. Thus, we conclude that the USDOC "general practice" regarding application of facts available is not within our terms of reference. Given that we find no claim was stated in this respect in the request for establishment at all, we consider that neither the United States nor potential third parties were informed of the legal basis of a complaint in this respect.

7.23.
As a consequence of our ruling in this regard, we will assess the consistency with the AD Agreement of the USDOC's decision to apply facts available in the investigation underlying this dispute, but will not make a general ruling as to the consistency, on its face, of the USDOC's "general practice" in the application of adverse facts available.

B. STANDARD OF REVIEW

1. Arguments

7.24.
Japan argues that Article 17.6(i) of the AD Agreement, which sets forth the standard of review to be applied to the case at hand, is a two part standard of review. In Japan's view, it requires that the Panel determine, firstly, whether the authorities' establishment of the facts was proper, which includes an assessment of whether all relevant facts were considered including those that might detract from an affirmative determination, and secondly, whether their evaluation of those facts was unbiased and objective. Japan asserts that the factual arguments in this case go directly to the US government's improper establishment of the facts and the non-objective and biased evaluation of the facts so as to favour the interest of the domestic industry. Japan also contends that Article 17.6(ii), which guides the Panel's interpretation of the AD Agreement, implicitly refers to the customary rules of interpretation of Articles 31 and 32 of the Vienna Convention on the Law of Treaties. Japan submits that these rules assume that at the end of the interpretation process the interpreter will craft one unambiguous interpretation of the provision in question. Finally, Japan claims that the general standard of review of Article 11 of the DSU applies to its challenge of the US laws and practice under Article X:3 of GATT 1994. This standard requires the Panel to make "an objective assessment of the matter before it including an objective assessment of the facts of the case and the applicability of and conformity with the covered agreements".
7.25.
The United States considers that under Article 17.6(i) of the AD Agreement, the task of the Panel is not to conduct a de novo evaluation of the facts if the authority's establishment of the facts is proper and its evaluation unbiased and objective, even though the Panel might have reached a different conclusion. The United States asserts that the role of the Panel is to examine whether the evidence before the investigating authority was such that an unbiased and objective investigating authority evaluating that evidence could properly have made the same determination. The scope of the Panel's review is limited by Article 17.5(ii) of the AD Agreement to the facts that were before the investigating authority when it made its determination, i.e. the evidence contained in the administrative record. With respect to Article 17.6(ii), the United States asserts that this provision requires panels to respect multiple permissible interpretations in their review of the legal interpretation by an investigating authority of the AD Agreement. The United States submits that Article 17.6(ii) reflects a deliberate choice by the negotiators to allow for multiple interpretations. The United States rejects Japan's argument that Articles 31 and 32 of the Vienna Convention require a panel to choose one interpretation of ambiguous language in the AD Agreement. Customary rules of interpretation, applied to Article 17.6(ii), prohibit an understanding of that provision under which the express language allowing for multiple permissible interpretations would be rendered a nullity. The United States further argues that in accordance with Article 31(3) of the Vienna Convention on the Law of Treaties, where the AD Agreement is ambiguous or silent with respect to a particular methodology, but that methodology has been subsequently adopted as standard practice by a number of signatories to the Agreement, the practice of those signatories must be taken into account with regard to determining whether that methodology constitutes a "permissible interpretation" of the Agreement. The United States therefore considers that the relevant question in every case is not whether the challenged determination rests upon the best or "correct" interpretation of the AD Agreement but whether it rests upon a "permissible interpretation" (of which there may be many). The United States finally submits that actions that are reviewed under the applicable deferential standard of review of the AD Agreement cannot also be reviewed under a different standard of review as Japan is suggesting merely because the claim has been phrased differently.

2. Finding

7.26.
Article 17.6 of the AD Agreement sets out a special standard of review for disputes arising under that Agreement. With regard to factual issues, Article 17.6(i) provides:

"in its assessment of the facts of the matter, the panel shall determine whether the authorities' establishment of the facts was proper and whether their evaluation of those facts was unbiased and objective. If the establishment of the facts was proper and the evaluation was unbiased and objective, even though the panel might have reached a different conclusion, the evaluation shall not be overturned;"

The question of whether the establishment of facts was proper does not, in our view, involve the question whether all relevant facts were considered including those that might detract from an affirmative determination. Whether the facts were properly established involves determining whether the investigating authorities collected relevant and reliable information concerning the issue to be decided - it essentially goes to the investigative process. Then, assuming that the establishment of the facts with regard to a particular claim was proper, we consider whether, based on the evidence before the US investigating authorities at the time of the determination, an unbiased and objective investigating authority evaluating that evidence could have reached the conclusions that the US investigating authorities reached on the matter in question.40 In this context, we consider whether all the evidence was considered, including facts which might detract from the decision actually reached by the investigating authorities.

7.27.
With respect to questions of the interpretation of the AD Agreement, Article 17.6(ii) provides:

"the panel shall interpret the relevant provisions of the Agreement in accordance with customary rules of interpretation of public international law. Where the panel finds that a relevant provision of the Agreement admits of more than one permissible interpretation, the panel shall find the authorities' measure to be in conformity with the Agreement if it rests upon one of those permissible interpretations."

Thus, in considering those aspects of the US determination which stand or fall depending on the interpretation of the AD Agreement itself rather than or in addition to the analysis of facts, we first interpret the provisions of the AD Agreement. As the Appellate Body has repeatedly stated, panels are to consider the interpretation of the WTO Agreements, including the AD Agreement, in accordance with the principles set out in the Vienna Convention on the Law of Treaties (the "Vienna Convention"). Thus, we look to the ordinary meaning of the provision in question, in its context, and in light of its object and purpose. Finally, we may consider the preparatory work (the negotiating history) of the provision, should this be necessary or appropriate in light of the conclusions we reach based on the text of the provision. We then evaluate whether the US interpretation is one that is "permissible" in light of the customary rules of interpretation of international law. If so, we allow that interpretation to stand, and unless there is error in the subsequent analysis of the facts under that legal interpretation under the standard of review under Article 17.6(i), the challenged action is upheld.

7.28.
While the parties have not raised any issues about burden of proof, we note that in WTO dispute settlement proceedings, the burden of proof with respect to a particular claim or defence rests with the party that asserts such claim or defence.41 The burden of proof is "a procedural concept which speaks to the fair and orderly management and disposition of a dispute".42 In the context of the present dispute, which is concerned with the assessment of the WTO consistency of a definitive anti-dumping measure imposed by the United States, Japan is obliged to present a prima facie case of violation of the relevant Articles of the AD Agreement. In this regard, the Appellate Body has stated that "... a prima facie case is one which, in the absence of effective refutation by the defending party, requires a panel, as a matter of law, to rule in favour of the complaining party presenting the prima facie case".43 Thus, where Japan presents a prima facie case in respect of a claim, it is for the United States to provide an "effective refutation" of Japan's evidence and arguments, by submitting its own evidence and arguments in support of the assertion that the United States complied with its obligations under the AD Agreement. Assuming evidence and arguments are presented on both sides, it is then our task to weigh and assess that evidence and those arguments in order to determine whether Japan has established that the United States acted inconsistently with its obligations under the AD Agreement.

C. OVERVIEW OF JAPAN'S CLAIMS

7.29.
Japan submits that the United States violated various provisions of the AD Agreement in its imposition of anti-dumping duties on imports of certain hot-rolled steel products from Japan. Japan claims that the use of adverse facts available to determine the dumping margin for the three investigated respondents is inconsistent with Articles 2.3, 2.4, 6.1, 6.6, 6.8, 6.13, 9.3, and Annex II of the AD Agreement. Japan claims that the US statute, on its face and as applied in this case, requiring the inclusion of margins calculated based on facts available in the determination of a dumping margin for all other non-investigated producers is inconsistent with Article 9.4 of the AD Agreement. Japan further considers the exclusion of certain home-market sales to affiliates from the normal value calculation on the basis of the application of the "arm's length" test, and their replacement with downstream sales, is inconsistent with Articles 2.1, 2.2, and 2.4 of the AD Agreement. Japan claims that the US preliminary determination of critical circumstances is inconsistent with Articles 10.1, 10.6, and 10.7 of the AD Agreement. Japan further claims that US "captive production" provision on its face, and as applied in this case, is inconsistent with Articles 3.1, 3.2, 3.4, 3.5, 3.6, and 4.1 of the AD Agreement, and that the USITC's analysis of injury and causation were inconsistent with Articles 3.1, 3.4, and 3.5 of the AD Agreement. Japan claims that certain actions undertaken by the United States were inconsistent with its obligations under Article X:3 of GATT 1994. Finally, Japan claims that certain laws, regulations, and administrative procedures governing various aspects of the investigations and determination in the underlying anti-dumping proceeding are not in conformity with its obligations, and thus that the United States has acted inconsistently with Article XVI:4 of the Marrakesh Agreement and Article 18.4 of the WTO Agreement.
7.30.
We note that we need not reach conclusions on all of these claims in order to resolve the dispute before us. The Appellate Body has observed that a "panel need only address those claims which must be addressed in order to resolve the matter in issue in the dispute".44 We keep in mind, however, the Appellate Body's further injunction that "[a] panel has to address those claims on which a finding is necessary in order to enable the DSB to make sufficiently precise recommendations and rulings so as to allow for prompt compliance by a Member with those recommendations and rulings "in order to ensure effective resolution of disputes to the benefit of all Members"".45

D. ALLEGED VIOLATIONS IN THE CALCULATION OF DUMPING MARGINS

1. Alleged violations of Articles 2, 6, and 9 and Annex II of the AD Agreement in the use of facts available in calculating dumping margins

7.31.
Japan claims that the use of facts available by USDOC in the case of sales by the investigated respondents was inconsistent with, inter alia, Article 6.8 and Annex II of the AD Agreement since the requirements for the use of facts available were not met. Secondly, Japan claims that USDOC 's choice of facts available of an adverse nature, based on the application of the US statute which provides that adverse inferences may be drawn if a party fails to cooperate with the investigating authority, was inconsistent with those provisions of the AD Agreement. Third, Japan argues that the specific application of facts available in this case was inconsistent with Articles 2.3 and 2.4 of the AD Agreement, and that the consequent application of the dumping margin thus calculated was inconsistent with Article 9.3 of the AD Agreement.

(a) NSC and NKK

7.32.
The price per ton of steel in coils is sometimes based on the actual weight of the coil, and sometimes on the "theoretical weight" of the coil, which is an estimated weight based on the dimensions of the product. All three investigated Japanese respondents made sales on both bases during the period of investigation. In order to be able to calculate the dumping margins in this case on the basis of a consistent unit of measurement, in the original questionnaire, USDOC requested the investigated Japanese respondents to provide a weight-conversion factor between sales made on an actual weight basis and sales made on a theoretical weight basis.

(i) Arguments

7.33.
Japan argues that NSC originally replied in good faith to the request for a weight conversion factor that "lacking an actual weight, NSC has no way of calculating the requested theoretical–to-actual weight conversion factor".46 After the preliminary determination of USDOC and while preparing for verification, NSC discovered information in its records that actual weights for sales made on a theoretical weight basis did in fact exist and were kept in a production database separate from the main sales database maintained at corporate headquarters. Thereupon, NSC submitted a conversion factor based on this information 14 days before verification.
7.34.
With respect to NKK, Japan asserts that NKK responded to the USDOC request for a conversion factor for sales made on the basis of theoretical weight that it was impracticable or impossible to provide a conversion factor with regard to such sales. After the preliminary determination, NKK discovered that KSC had used a "best estimate" as a surrogate for a weight conversion factor and that this approach had been accepted by USDOC. NKK submitted its own weight conversion factor based on the same best estimate methodology, 9 days before the commencement of verification.
7.35.
Japan claims that USDOC acted inconsistently with the AD Agreement in refusing to accept the information provided by each company. Japan maintains that USDOC erred in concluding that these companies had failed to cooperate in the investigation to the best of their ability, and therefore erred in concluding that the application of adverse facts available was appropriate. Japan submits that the issue is not whether the deadline established for responses to questionnaires by USDOC was reasonable, but whether NSC and NKK submitted the information within a reasonable period. Japan argues that flexibility when the circumstances warrant is the hallmark of reasonableness. According to Japan, USDOC should have accepted the information, which was submitted before verification, allowing USDOC and petitioners time to review the information and comment on it.
7.36.
Japan argues USDOC could have used the submitted conversion factors without undue difficulty, referring to paragraph 3 of Annex II. Japan submits that the United States failed to demonstrate that it was prejudiced in any way by the late submission of the information, which could have been verified.47 Moreover, Japan argues, USDOC violated paragraph 5 of Annex II in disregarding and rejecting information provided by the party acting to the best of its ability. Japan asserts that the United States violated paragraph 7 of Annex II by applying facts available in spite of the fact that both companies cooperated as set forth in that paragraph. In Japan's view, neither company withheld information and both cooperated with the authority so that the basic conditions for a facts available determination were not fulfilled.
7.37.
Japan also claims that NKK was not given proper notice of the need to supply a conversion factor or a proper opportunity to respond and defend itself, which Japan asserts was required by Article 6.1 of the AD Agreement, contending that USDOC officials instructed that NKK need not submit any conversion factor. Japan argues that USDOC failed to properly verify the information on weight conversion factors provided by NKK and NSC, as required by Article 6.6 of the AD Agreement. Japan also submits that USDOC failed to take into account difficulties faced by NSC and NKK as required by Article 6.13 and did not provide assistance to the companies as required by that provision.
7.38.
Japan objects to the choice of facts available used in both cases. Japan asserts that the United States, in accordance with its general practice in this regard, drew an adverse inference based on the alleged failure of the companies to cooperate, and deliberately applied the highest margin calculated as the margin for the sales of both companies made in theoretical weight. Japan submits that the use of adverse facts available in the determination of margins for the theoretical weight sales of NSC and NKK was inconsistent with the AD Agreement.
7.39.
Japan also asserts the USDOC's rejection of evidence and application of adverse facts available with regard to NSC and NKK was inconsistent with Article 2.4 of the AD Agreement, which requires that the investigating authorities make a "fair comparison... between the export price and the normal value." Japan asserts that for NSC's theoretical weight export sales, USDOC did not calculate an export price, but instead assigned the highest margin determined for that product type to those sales. Japan argues that USDOC could have used some form of conversion factor to generate surrogate export prices for those sales. Japan submits that the resort to facts available does not excuse authorities from their obligations, least of all from the obligation to make a fair comparison between export price and normal value. With regard to NKK, which had only made sales in theoretical weight in the home market, Japan submits that USDOC relied on isolated transactions with no relationship to NKK's overall average normal value as adverse facts available. Thus, Japan asserts, USDOC inflated the normal value on those sales. Finally, Japan maintains that USDOC's alleged failure to calculate the margins correctly under Article 2 in turn led to measures inconsistent with Article 9.3 of the Anti-Dumping Agreement, which requires that "The amount of the anti-dumping duty shall not exceed the margin of dumping as established under Article 2." Japan asserts that USDOC's erroneous application of facts available led to inflated dumping margins, and therefore was inconsistent with Article 9.3 of the Anti-Dumping Agreement.
7.40.
The United States submits that facts available were applied to sales of NSC and NKK in accordance with Article 6.8 of the AD Agreement since the information requested with regard to the weight conversion factor was not submitted within the reasonable deadlines set by USDOC. The United States acknowledges that USDOC Regulation section 351,301(b) provides a general rule that information submitted at least 7 days prior to verification is considered timely. However, the United States points out, an exception exists in section 351,301(c) for responses to questionnaires, which must be submitted within the deadline mentioned on the questionnaire (or any extended period granted). The requested information on weight conversion factors was not submitted within the time allowed for responses to the questionnaires. The United States argues that it is therefore clear that NSC and NKK failed to provide the information in accordance with the US regulatory deadlines, despite sufficient time to do so - including extensions, a total of 87 days was allowed for responses to the questionnaire.
7.41.
The United States argues that Article 6.8 of the AD Agreement does not provide a definition of a "reasonable period", and neither does Annex II. According to the United States, the deadlines set by USDOC were reasonable and USDOC was therefore permitted to disregard information that was provided one month after the deadlines had expired. The United States points to paragraphs 1 and 3 of Annex II as establishing timeliness as an independent condition for acceptance of information. The United States argues that it would render the timeliness requirement of Annex II meaningless if Japan's argument were accepted that, so long as the information concerned is of minor importance, it has to be accepted if it is submitted before the verification takes place.
7.42.
The United States submits that ample opportunity was given to furnish the information and present evidence as required by Article 6.1 of the AD Agreement. NSC and NKK ultimately had 87 days to respond to the questionnaires, far more than the 30 days required by Article 6.1.1 of the AD Agreement, but rather than providing the requested weight conversion factor, they merely repeatedly answered that it was either not possible or not necessary to provide the information.48 The United States argues that although paragraph 6 of Annex II requires that an opportunity be given to present evidence and to give further explanations if the information is rejected, it also refers to the time constraints on the investigating authority and cannot be construed to provide an endless opportunity to submit new information.
7.43.
The United States further argues that the application of adverse facts available in the case of NSC and NKK was permissible under paragraphs 5 and 7 of Annex II. Paragraph 5 requires the authority to accept information even if it is not ideal if the company has acted to the best of its ability. But, the United States argues, NSC and NKK did not act to the best of their ability since they could have submitted the information well within the deadlines, as demonstrated by the fact that they did so ultimately. The United States argues that when a company fails to act to the best of its ability it is deemed uncooperative and USDOC was therefore permitted under paragraph 7 of Annex II to apply adverse facts available.
7.44.
The United States argues that USDOC did not act inconsistently with Article 6.6 of the AD Agreement when it refused to verify the untimely submitted conversion factors of NSC and NKK. Article 6.6 of the AD Agreement provides clearly that the investigating authority is only required to verify the information, submitted in a timely manner, on which its findings are based. The United States argues that this was not the situation here, and consequently verification of the weight conversion factor information submitted was not required.49 Finally, the United States asserts that USDOC provided all the assistance required by Article 6.13 of the AD Agreement.50
7.45.
The United States maintains that USDOC's use of facts available, including its use of an adverse inference, for the sales affected by the weight conversion factor was fully consistent with Articles 2.4 and 9.3 of the Agreement. The United States submits that USDOC applied adverse facts available to sales of NSC and NKK made on the basis of theoretical weight in accordance with Article 2.4 of the AD Agreement. The United States explains that with regard to the theoretical weight sales of NKK, USDOC used the highest per-product weighted average price and averaged them together with the actual weight sales data, and then compared the product specific weighted average export price to those weighted average normal values which were most similar matches ensuring a fair comparison under Article 2.4 of the AD Agreement. For NSC, USDOC immediately assigned a dumping margin based on actual weight sales of the same product since, in the view of the United States, nothing in Article 2.4 of the AD Agreement or Article 6.8 requires the authorities to first calculate an export price and only then to determine the appropriate margin of dumping. Since the margins of dumping in all three cases were established in accordance with Article 2 of the AD Agreement, the United States argues that Japan's claim of a violation of Article 9.3 of the AD Agreement must automatically fail as well.
7.46.
Because NSC and NKK could have provided the information in a timely manner but did not do so, they failed to act to the best of their ability and did not cooperate with USDOC with respect to this information request. As a consequence, the United States maintains that USDOC was justified in rejecting the untimely data, and applying an adverse inference in its choice of facts available.
7.47.
In the view of the United States, the facts available provisions of the Agreement, including paragraph 7 of Annex II, allow for an adverse inference to be taken when a party does not cooperate in providing the information at issue. The United States asserts that this possibility to take adverse inferences is important to provide an incentive for exporters to respond to the questionnaires in a complete and timely manner. The United States asserts that the principle at issue here is a simple one: the right, under the Agreement, for authorities to establish reasonable deadlines for submission of information and to use the facts available when a party does not provide a response to a questionnaire within those reasonable deadlines. Where, as here, a respondent first offers the requested data long after the reasonable deadlines have passed for its submission, the Agreement does not compel the investigating authority to accept and use the late‑provided data.51 Similarly, the United States argues, the Agreement permits authorities to use an adverse inference in selecting from the facts available when a party has failed to act to the best of its ability to timely supply the requested information. The Agreement thus leaves it to investigating authorities to set and enforce deadlines for receiving information in keeping with their judgment as to when they need it, so long as they provide interested parties with a reasonable period in which to make their submissions. The United States concludes that the interpretation proposed by Japan with respect to the use and selection of the facts available would render meaningless any Member’s right to establish deadlines, which the United States considers unacceptable and not intended by the Agreement.
7.48.
With regard to the specifics of this case, Brazil considers that NSC and NKK were punished for not respecting the deadlines in the submission of a minor piece of information. Korea considers that a punitive margin was applied in the case of NSC and NKK. Moreover, Korea submits, by applying a dumping margin as the facts available, rather than simply using secondary source information for the missing conversion factor, USDOC failed to make a fair comparison as required by the AD Agreement.52

(ii) Finding

7.49.
Japan claims that USDOC acted inconsistently with the requirements of Article 6.8 and Annex II of the AD Agreement in resorting to facts available in the determination of the dumping margin for NSC and NKK.53 The United States maintains that the use of facts available was based on a permissible interpretation of Article 6.8, and was justified in the circumstances of this case. Japan also asserts that the particular "adverse" facts available used were inappropriate, as they were chosen based on an unjustified determination that NSC and NKK had failed to cooperate to the best of their ability. The United States contends, on the other hand, that adverse inferences are permissible in the choice of facts available, and that the decision to apply such an inference in this case was fully justified.
7.50.
The first question we must address in resolving this issue is under what circumstances does Article 6.8 of the AD Agreement allow an investigating authority to resort to the use of facts available. Article 6.8 provides:

"In cases in which any interested party refuses access to, or otherwise does not provide, necessary information within a reasonable period or significantly impedes the investigation, preliminary and final determinations, affirmative or negative, may be made on the basis of the facts available. The provisions of Annex II shall be observed in the application of this paragraph."

Annex II sets out additional conditions and considerations relevant to the application of facts available in a particular case.

7.51.
The conditions for applying "facts available" under Article 6.8 seem fairly clear on the face of that provision. If an interested party "refuses access to" necessary information within a reasonable period, "otherwise does not provide" necessary information with a reasonable period, or "significantly impedes the investigation" the investigating authority may make determinations on the basis of the facts available. Thus, Article 6.8 ensures that an investigating authority will be able to complete an investigation and make determinations under the AD Agreement on the basis of facts even in the event that an interested party is unable or unwilling to provide necessary information within a reasonable period.
7.52.
The question before us is whether the USDOC was justified in concluding that NKK and NSC refused access to or otherwise did not provide necessary information within a reasonable period.54 Japan argues extensively that USDOC erred in applying "adverse" facts available in this case. However, before the question of applying "adverse" facts available need be addressed, we must first assess whether USDOC was justified under Article 6.8 AD Agreement to make its determination on the basis of facts available. If USDOC acted inconsistently with Article 6.8 in resorting to facts available at all, then the specific choice of which facts it applied is, in our view, moot.
7.56.
In its final determination, USDOC explained its decision with respect to NSC:

"The evidence indicates that the requested information was routinely maintained by NSC in the normal course of business, but that obtaining it was simply not a priority. Regardless of who specifically knew about this information, the sales department or the production department, the data existed and could have easily been obtained. The fact that NSC was able to provide this information shortly after the preliminary determination also supports the conclusion that it could have done so within the time requested. Moreover, it is impossible for the Department to determine whether NSC's claims of inadvertent error are valid or merely self-serving. Thus, they are insufficient to rebut the evidence establishing that the requested information was readily available. …

Because NSC's conversion data was untimely and did not constitute a minor correction the Department informed NSC at verification that it would not accept the theoretical to actual weight conversion factors and returned the data on April 12, 1999. …

Because NSC failed to timely provide requested information, …, the Department has made its determination with respect to the theoretical weight sales on the basis of the facts available. Further, the Department finds that NSC, by not submitting a theoretical weight conversion factor it could have provided when originally requested until well after the time for response had passed, failed to cooperate by not acting to the best of its ability…The fact that NSC ultimately did provide such a factor is proof that it could have done so much earlier…."57

7.57.
It is thus clear to us that in the case of NSC the USDOC rejected information that was actually submitted to it, albeit not by the deadline specified, despite the fact that the information was available in sufficient time to allow its verification and use in the calculation of NSC's dumping margin. In our view, based on the evidence before the USDOC at the time of the determination, an unbiased and objective investigating authority evaluating that evidence could not have reached the conclusion that NSC had failed to provide necessary information within a reasonable period. Thus, we conclude that USDOC acted inconsistently with Article 6.8 in applying facts available in making its determination of NSC's dumping margin.
7.58.
With regard to NKK, USDOC stated in its final determination:

"Because NKK's conversion factor data were not timely submitted, the Department rejected these factors in a letter dated April 12, 1999. The Department, therefore, has not considered these data or retained them in the official record of the proceeding. … The Department does not agree with NKK's assertion that these data were verified. Rather, at verification, the Department specifically informed NKK and its counsel that the Department would not accept the conversion factor and would specifically instruct NKK to submit this information on the record if the Department determined that it was timely. However, any arguments as to the accuracy of these data are moot because the data in question are no longer part of the record before the Department…

Further the Department finds that NKK, by not submitting a theoretical weight conversion factor it could have provided when originally requested until well after the time for response had passed, failed to cooperate by not acting to the best of its ability. NKK's claims that it could calculate a conversion factor in February of 1999 but was unable to derive such a factor when the questionnaire responses were due, does not withstand scrutiny. Although NKK argues that it did not understand what the Department wanted when it originally requested a "conversion factor", although this was not stated at the time, and that it lacked the data necessary to calculate one,… it should have proposed to the Department the sort of conversion factor it ultimately did calculate, explaining why a more accurate one might not be practicable. Instead, NKK merely dismissed the Department's repeated requests. The fact that NKK ultimately did provide such a factor is the proof that they could have done so much earlier."58

7.59.
It is thus clear to us that in the case of NKK as well, USDOC rejected information that was actually submitted to it, albeit not by the deadline specified, despite the fact that the information was available in sufficient time to allow its verification and use in the calculation of NKK's dumping margin. In our view, based on the evidence before USDOC at the time of the determination, an unbiased and objective investigating authority evaluating that evidence could not have reached the conclusion that NKK had failed to provide necessary information within a reasonable period. Thus, we conclude that USDOC acted inconsistently with Article 6.8 in applying facts available in making its determination of NKK's dumping margin.
7.60.
Having determined that an objective and unbiased investigating authority could not, on the basis of the evidence in this dispute, have reached the conclusion that NKK and NSC failed to provide necessary information within a reasonable period, we do not consider it necessary or appropriate to address Japan's additional claims and arguments regarding the application of adverse facts available in the underlying investigation, or the consistency of USDOC's actions with Articles 2.4, 6.1, 6.6, 6.13, and 9.3 and Annex II of the AD Agreement.

(b) KSC

(i) Arguments

7.62.
Japan argues that USDOC violated Article 6.8 and Annex II of the AD Agreement in determining that KSC had failed to cooperate or withheld information, and that the use of adverse facts available was therefore justified. Japan notes that CSI was not merely an affiliated customer of KSC, it was itself a petitioner in the anti-dumping investigation of hot-rolled steel imports from Japan. Japan argues that KSC did not withhold any information from USDOC since the information requested was never in its possession but was kept by CSI. Japan points out that USDOC treated KSC and CSI as a single entity, thus assuming that KSC had sufficient control over CSI to compel its cooperation. Japan maintains that despite its efforts, KSC was unable to convince CSI to hand over the requested information. Japan argues that without the voluntary cooperation of the co-owner CVRD, KSC was powerless to compel CSI to provide the requested information. Japan submits that the Shareholders Agreement, which the United States asserts gave KSC certain power to obtain the requested information does not, in fact, reflect how the company operated in practice.59 Japan further emphasises that KSC did attempt to exercise its rights under the Shareholders Agreement, but was nonetheless unable to obtain the requested information.
7.63.
Moreover, Japan submits, USDOC failed to provide any assistance as required by Article 6.13 of the AD Agreement and paragraph 7 of Annex II. KSC invested substantial resources attempting to convince CSI and CVRD to help KSC to respond to USDOC’s request.60 KSC explained and documented that it had no way in which to force CSI to share its further manufacturing and resale data due to the structure and management of the joint venture. Japan argues that KSC explained to USDOC that the Shareholders Agreement did not operate to allow either KSC or CVRD any real control of CSI's day-to-day operations.61 Japan submits that only one person controlled the day-to-day conduct of CSI, its President Mr Gonçalves as evidenced by the decision of CSI to become a petitioner in this case.62 Japan submits that the USDOC failed to properly establish the facts by ignoring the evidence that CSI was unwilling to provide the requested information and CVRD, KSC's joint venture partner, was a competitor of KSC in the US market and thus did not have an interest in assisting KSC in the US AD investigation either. KSC therefore did not possess the information requested and did not have the power to obtain this information. Consequently, Japan maintains that the USDOC acted inconsistently with Article 6.8 of the AD Agreement in concluding that resort to adverse facts available was appropriate.
7.64.
Japan asserts that USDOC's use of the second-highest margin from KSC’s sales as facts available for the CSI transactions was inconsistent with Article 2.3 of the Anti-Dumping Agreement governing the calculation of export price. Japan argues that USDOC did not have any specific concerns regarding the unreliability of KSC's export price to CSI, but proceeded to calculate an export price, necessitating the resale and further manufacturing information, based on the assumption that because KSC and CSI were affiliated, CSI’s resale prices were necessary. When those prices were not provided, Japan asserts that USDOC did not determine an export price for KSC’s sales to CSI, and instead applied a margin from other sales. Japan maintains that this was inconsistent with Article 2.3, which permits the calculation of a constructed export price, not the imposition of a margin. Finally, Japan maintains that by wrongly applying adverse facts available, the United States ultimately applied an anti-dumping duty higher than the margin of dumping, inconsistent with Article 9.3 of the AD Agreement.
7.65.
The United States argues that the application of adverse facts available to a part of KSC's sales was permitted under the AD Agreement, since KSC failed to act to the best of its ability with regard to submitting the requested data concerning its sales through CSI, its US affiliate. The United States points out that USDOC requested the information three times, and that KSC twice requested to be excused from giving the information and on the third occasion alleged that it was not possible to submit the data due to the unwillingness of CSI. However, the United States maintains that KSC never even raised the issue before the Board of Directors of CSI, of whom two out of four are appointed by KSC, never tried to enforce certain rights under the Shareholders' Agreement in an effort to obtain the requested information, and never directly raised the issue with its joint venture partner CVRD.63 The United States submits that even if cooperation was refused by CVRD, internal means of forcing the issue and obtaining the information were available to KSC under the Shareholders' Agreement. In the US view, KSC's failure to use these means indicates that it acquiesced in the refusal of CSI to submit the requested information. As a result, KSC failed to provide the requested information within a reasonable period, and therefore USDOC was fully entitled to apply adverse facts available to that part of KSC sales to the United States that entered the United States market through CSI.
7.66.
The United States argues that the assistance requirement of Article 6.13 of the AD Agreement invoked by Japan relates in particular to small companies while KSC is one of Japan's largest corporations and one of the biggest steel producers in the world. The United States asserts that it was not USDOC's responsibility to advise KSC on the steps to take to respond to the questionnaire and argues that the information requested was clear and unambiguous. In any case, the United States submits, contrary to Japan's assertions, KSC never requested such assistance.
7.67.
Finally, the United States maintains that it did not act inconsistently with the requirements of Article 2.3 in applying, as facts available, a margin based on other KSC sales as the margin for the sales to CSI, rather than seeking to calculate an export price based on facts available. The United States maintains that KSC's refusal to provide the information necessary to construct an export price, made it necessary for USDOC to use the margin information as it was impossible to construct an export price based on available facts. The United States rejects Japan’s further argument that USDOC's determination was inconsistent with Article 9.3 of the Agreement, maintaining that USDOC correctly calculated KSC's margin, and was therefore entitled to impose a definitive measure in the amount of the margin calculated.
7.68.
With regard to the specifics of this case, Brazil questions the use of facts available in light of the fact that KSC was confronted with the refusal of CSI, which was itself a petitioner in the case, to provide the information. Korea objects to the use of fact available to KSC since it was CSI and not KSC that failed to cooperate. Chile considers that the US acted inconsistently with the AD Agreement by punishing KSC for not providing the information that was held by CSI, a petitioner in this case. According to Chile, it was unreasonable for USDOC to require cooperation between two companies having such a clear conflict of interests. Chile asserts that in any case, Article 6.8 and Annex II do not permit the choice of the most adverse facts available when a party fails to cooperate and certain information is not provided. The EC disagrees with Japan that facts available may only be used as neutral gap fillers. When selecting facts available, the authorities may take into account the degree of cooperation of the party concerned. If an exporter refuses to provide certain information, a logical inference may be drawn that the information is adverse to his interests.

(ii) Finding

7.69.
Pursuant to Article 6.8 of the AD Agreement, and as discussed above, where a party "otherwise does not provide" necessary information within a reasonable time, the investigating authorities may make their determination based on facts available. It is undisputed in this case that KSC did not provide the requested information regarding resale prices and further manufacturing costs with respect to its sales through its affiliate CSI. Thus, it appears that USDOC was justified in deciding to apply facts available with respect to the information not provided by KSC concerning CSI's further manufacturing costs, as this necessary information was not provided within a reasonable period.
7.70.
Indeed, Japan's argument with respect to the application of facts available concerning KSC focuses not on the application of facts available per se, but rather on the "adverse" nature of the facts available actually used. The parties' argument in this regard focuses on paragraph 7 of Annex II of the AD Agreement, which provides in relevant part:

"It is clear, however, that if an interested party does not cooperate and thus relevant information is being withheld from the authorities, this situation could lead to a result which is less favourable to the party than if the party did cooperate".

The parties have argued extensively concerning whether this provision allows an investigating authority to conclude that a party has failed to cooperate and therefore resort to the application of "adverse" facts available. Japan maintains that this provision merely recognizes the possibility that a failure to cooperate might result in a less favourable result for the non-cooperating party, but does not allow for any deliberate action on the part of investigating authorities to ensure such a less favourable result. The United States, on the other hand, argues that the entire purpose of allowing investigating authorities to resort to facts available would be defeated if, regardless of the failure of a party to cooperate, investigating authorities were obliged in all circumstances to seek out "neutral" facts available.

7.71.
As discussed earlier, we have concluded that Japan did not set out a claim with respect to USDOC's general practice in applying adverse facts available in its request for establishment. Therefore, we will limit our analysis to the application of facts available in the particular circumstances of this case. We consider the question whether paragraph 7 of Annex II, together with the remainder of Annex II and Article 6.8, allow for a deliberate application of facts available with the intent of obtaining a result less favourable for a particular respondent to be a question that implicates the US statute and regulations, which are not within our terms of reference in this dispute. However, it seems clear to us that any "less favourable" result under paragraph 7 of Annex II may only be appropriate in the case of an interested party who does not cooperate. The USDOC decision to apply adverse facts available in the case of KSC was based on the underlying determination that KSC had failed to cooperate by failing to act to the best of its ability to comply with the USDOC request for information concerning resale prices and further manufacturing costs. Therefore, we consider first whether an objective and unbiased investigating authority could reasonably have concluded, based on the facts before the USDOC, that KSC had failed to cooperate and that relevant information was being withheld.
7.72.
In its final determination, USDOC found, in pertinent part:

"In essence, for purposes of the [constructed export price] calculation, the [US] statute treats the exporter and the US affiliate collectively, rather than independently, regardless of whether the exporter controls the affiliate. Accordingly, KSC's argument that it does not "control" CSI is misplaced and irrelevant.

Because the statute requires that the Department base its margin calculations for the CSI sales on record information concerning the CSI sales themselves, the Department required that KSC and CSI, collectively, provide the necessary price and cost data for KSC's US sales through CSI. It is also undisputed that KSC and CSI failed to provide this necessary information….KSC and CSI have neither provided the data on CSI's sales, as requested by the Department, nor demonstrated to the Department's satisfaction that this is not possible. Therefore, the Department finds that KSC and CSI have failed to cooperate by not acting to the best of their ability to comply with the Department's requests for information with respect to the CSI sales. Therefore, we have used an adverse inference in selecting the facts available with respect to the CSI sales.

Allowing a producer and its US affiliate to decline to provide US cost and sales data on a large portion of their US sales would create considerable opportunities for such parties to mask future sales at less than fair value through the US affiliate. The fact that the affiliate is a petitioner does not allay such concerns. Thus, this fact does not constitute an exception to the principle that the Department may make an adverse inference with respect to sales for which data is not provided unless the foreign exporter and its US affiliate have acted to the best of their ability to provide such data.

While it is clear that KSC and CSI collectively have not acted to the best of their ability, we also disagree with KSC's claim that it alone acted to the best of its ability. … After careful consideration of all of the evidence on the record, the Department finds that KSC did not act to the best of its ability with respect to the requested CSI data.

CSI is a joint venture between KSC and a large Brazilian mining operation, Companhia Valle do Rio Doce ("CVRD"). Through their respective US affiliates, KSC and CVRD each own 50 per cent of CSI. KSC's claim that it acted to the best of its ability with respect to this issue rests on its assertion that it was powerless to compel CSI to provide the Department with this data, given that CSI as a petitioner in this case, refused to cooperated. Some of the most important evidence contradicting KSC on this issue, including information pertaining to the board and the Shareholders' Agreement, constitutes business proprietary information, and are discussed only in our proprietary Analysis Memorandum, which is hereby incorporated by reference. Generally, however, the record shows that, although KSC could have been much more active in obtaining the cooperation of CSI in this investigation, it limited its efforts to merely requesting the required data and otherwise took a "hands-off" approach with respect to CSI's alleged decision not to provide this data. For example, KSC officials stated that KSC did not instruct its members of the CSI board to address the issue, did not invoke the Shareholder's Agreement, and did not discuss this issue with its joint venture partner. This does not reach the "best efforts" threshold embodied in § 776(b). Furthermore, the fact that KSC has provided a great deal of information and has substantially cooperated with respect to other issues does not relieve it of the requirement to act to the best of its ability to provide the requested CSI information. With respect to the CSI sales, KSC has provided only minimal volume and value information and has not acted to the best of its ability to obtain further information. Thus, as to the missing CSI data, it cannot be said that KSC was fully cooperative and made every effort to obtain and provide the information requested by the Department. Therefore, even though full cooperation by KSC would not constrain the Department from using adverse facts available specifically with respect to the CSI sales, we do not agree with KSC's argument that it has "substantially cooperated" during this investigation….

While the Department has considered that the record supports KSC's claim that it did make some effort to obtain the data and the CSI's management rebuffed these efforts, the record also shows that KSC essentially acquiesced in CSI's decision not to provide this data. Given KSC's relationship with this 50/50 joint venture, as detailed in the Home Market Sales Verification Report, dated 26 March 1999, this did not constitute making its best efforts to obtain the data."64

7.73.
Our review of the facts on which USDOC based this conclusion, including confidential information, leads us to the conclusion that an unbiased and objective investigating authority evaluating the evidence that was before the USDOC could not reasonably have reached the conclusion that KSC had failed to cooperate and that relevant information was thus being withheld. "Cooperate" has been defined as "work together for the same purpose or in the same task."65 In our view, USDOC's conclusion that KSC failed to act to the best of its ability to comply with the request for information in this case went far beyond any reasonable understanding of any obligation to cooperate implied by paragraph 7 of Annex II. CSI was a petitioner in the investigation of hot-rolled steel imports from Japan, and thus had interests directly opposed to those of KSC. Indeed, this very fact suggests that KSC lacked the ability to control such important decisions of policy by CSI. USDOC's own conclusion that KSC "acquiesced" in CSI's refusal to provide the requested information itself suggests that KSC was not able to direct CSI's actions in this regard. CVRD, KSC's joint venture partner in CSI, was itself KSC's competitor in the US market for the steel products under investigation, and thus also had interests adverse to those of KSC. While it is conceivable that KSC could have undertaken certain measures under the Shareholders' Agreement with the possible result of forcing CSI to provide the requested information, such actions would have inevitably disrupted the on-going business relationships of the three companies. We do not consider that USDOC's conclusion that KSC's not having taken such measures justified the conclusion that it had failed to cooperate was a decision that could properly be made by an unbiased and objective investigating authority on the basis of the evidence before USDOC. In the absence of a justified conclusion that there was a lack of cooperation, there is no basis under paragraph 7 of Annex II for a result which is less favourable than would have been the case had the party cooperated.
7.74.
We therefore conclude that USDOC acted inconsistently with Article 6.8 and Annex II paragraph 7 of the AD Agreement in applying adverse facts available in making its determination of KSC's dumping margin. We therefore do not consider it necessary or appropriate to address Japan's additional claims and arguments under Articles 2.3 and 9.3 regarding the application of adverse facts available in the determination of KSC's dumping margin in the underlying investigation.

2. Alleged violations of Article 9.4 of the AD Agreement on the face of US law and in the calculation of an "all others" rate including margins established on the basis of facts available

(a) Arguments

7.75.
Japan argues that Article 9.4 of the AD Agreement provides in clear and mandatory terms that any margins based on facts available shall be disregarded for the purposes of determining an all others rate. Japan submits that the US statute, section 735(c)(5) of the Tariff Act of 1930 (as amended), impermissibly limits the exclusion provided for under Article 9.4 to only margins based entirely on facts available. Japan argues that the statutory distinction between margins based entirely on facts available, which are excluded, and margins based only partially on facts available, which can still form the basis for an "all others" rate, is, on its face, inconsistent with the AD Agreement In Japan's view, Article 9.4 is clear and explicit in prohibiting the inclusion of any margins based even in part on facts available in the calculation of the all others rate, and the US interpretation of Article 9.4 is impermissible. Japan notes that facts available may be used in determining a company's dumping margin under Article 6.8 of the AD Agreement due to a particular action or inaction of that particular company, and asserts that Article 9.4 of the AD Agreement establishes that other non-investigated companies should not be punished for the lack of cooperation of another company.
7.76.
Second, Japan argues that, applying the statute in its determination of an all others rate, USDOC violated Article 9.4 of the AD Agreement since it used as the basis for the calculation of the all others rate the margins calculated for the three investigated respondents, each of which was based in part on facts available. Japan submits that the USDOC determination of an "all others" rate on the basis of margins that were calculated based in part on facts available was inconsistent with Article 9.4 of the AD Agreement.
7.77.
The United States argues that Article 9.4 of the AD Agreement permits the inclusion of margins partially based on facts available in establishing an all others rate. Likewise, the United States argues, merely because a factor in the calculation of the overall margin for each of certain investigated producers is de minimis or zero does not mean that such margins cannot be used in the determination of an all others rate under Article 9.4 of the AD Agreement. The United States argues that margins are only "established under the circumstances referred to in paragraph 8 of Article 6", and to be disregarded in the determination of a margin of dumping for non-investigated producers or exporters as provided for in Article 9.4, if they are entirely based on facts available. Thus, when only an element included in the overall calculation of the margin is determined in this manner, as was the case in the investigation on imports of hot-rolled steel products from Japan, the United States maintains that such margins may be included in the calculation of the all others rate.
7.78.
The United States argues that the term "margin" is used throughout the AD Agreement as referring to the overall margin for a producer or exporter, and not just to portions of it. In this regard, the United States notes that Article 5.8 of the AD Agreement, which provides that an investigation is to be terminated immediately when the margin of dumping is de minimis, is generally understood not to require termination merely because one transaction of a producer shows a de minimis margin. Thus, the United States maintains that only if the overall margin for the producer is de minimis does Article 5.8 require termination of the investigation. The United States refers to the customary rule of interpretation that the terms of an Agreement be given the same meaning throughout the Agreement to argue that the same meaning should apply in Article 9.4, and therefore only if the overall margin was based on facts available does Article 9.4 require its exclusion from the all others rate. The United States also argues that, to the extent the term margin might be ambiguous, it is clear that Members are free to adopt any permissible interpretation.
7.79.
The United States also argues that the interpretation argued by Japan would be unworkable, since it is often the case that parts of the margin will be based on facts available. For instance, in this case, there were no margins calculated that did not rest, in part, on facts available, and thus Japan's interpretation of Article 9.4 would make it impossible to calculate an all others rate consistently with the AD Agreement. Moreover, this would undermine the purpose of Article 6.10 of the AD Agreement which explicitly allows the investigation of less than all exporters where a full investigation of all producers or exporters would be too burdensome for the investigating authority. The United States also rejects Japan’s allegation that the non-investigated producers would be "punished" for the respondents' lack of cooperation if they were to receive an all others rate based in part on margins established on facts available. In the United States' view, Article 9.4 establishes a neutral formula for the calculation of the all others rate, which the USDOC properly applied in this case.
7.80.
Brazil considers that Article 9.4 of the AD Agreement provides in a clear and unequivocal manner that margins established on the basis of facts available are to be excluded in the determination of an all others rate. The distinction in US law between margins based entirely or only partially on facts available is not present in the AD Agreement and it cannot be argued that it is a permissible interpretation of the Agreement, Brazil submits, since it completely changes the meaning of the relevant provision in the AD Agreement.
7.81.
Chile argues that the US practice to include margins partially based on facts available in the determination of an all others rate violates Article 6.13 of the AD Agreement and 9.4 of the AD Agreement and unduly inflates the margin for the non-investigated exporters or producers.
7.82.
The EC agrees with Japan that by providing for the exclusion from the all others rate only of those dumping margins which are based entirely on facts available, US law is inconsistent with Article 9.4. The EC would nevertheless argue that Article 9.4 does not require investigating authorities to disregard the dumping margin in every instance where facts available have been used, as long as the resort to facts available was limited and no adverse inferences were drawn.

(b) Finding

7.83.
Article 9.4 of the AD Agreement establishes a maximum level for anti-dumping duties to be applied to imports from exporters or producers not included in the investigation, as provided for in Article 6.10 of the AD Agreement. It provides:

"When the authorities have limited their examination in accordance with the second sentence of paragraph 10 of Article 6, any anti‑dumping duty applied to imports from exporters or producers not included in the examination shall not exceed:

(i) the weighted average margin of dumping established with respect to the selected exporters or producers or,

(ii) where the liability for payment of anti‑dumping duties is calculated on the basis of a prospective normal value, the difference between the weighted average normal value of the selected exporters or producers and the export prices of exporters or producers not individually examined,

provided that the authorities shall disregard for the purpose of this paragraph any zero and de minimis margins and margins established under the circumstances referred to in paragraph 8 of Article 6. The authorities shall apply individual duties or normal values to imports from any exporter or producer not included in the examination who has provided the necessary information during the course of the investigation, as provided for in subparagraph 10.2 of Article 6" (emphasis added).

7.84.
Section 735(c)(5) of the Tariff Act of 1930, as amended, and the implementing regulations of the USDOC establish the method to be applied by USDOC in determining the estimated "all others" rate, which is the rate applicable to uninvestigated producers. US law generally provides that USDOC shall determine the estimated weighted average dumping margin for each exporter and producer individually investigated, and then determine the estimated all-others rate for all exporters and producers not individually investigated.66 Section 735(c)(5) specifies, with respect to the all others rate:

"(A) General rule

For purposes of this subsection and section 1673b(d) of this title, the estimated all-others rate shall be an amount equal to the weighted average of the estimated weighted average dumping margins established for exporters and producers individually investigated, excluding any zero and de minimis margins, and any margins determined entirely under section 1677e of this title [i.e., any margins determined entirely on the basis of facts available].

(B) Exception

If the estimated weighted average dumping margins established for all exporters and producers individually investigated are zero or de minimis margins, or are determined entirely under section 1677e of this title, the administering authority may use any reasonable method to establish the estimated all-others rate for exporters and producers not individually investigated, including averaging the estimated weighted average dumping margins determined for the exporters and producers individually investigated".67

7.85.
Thus, US law creates a distinction between rates based "entirely" on facts available, which are excluded from the calculation of an "all others" rate, and individual rates based on facts available in part, which, under the US law, can be used in the calculation of an "all others" rate. In this case, USDOC investigated three respondents and calculated an all others rate applicable to the remaining Japanese producers, by taking the weighted average of the margins calculated for the three investigated respondents.68 USDOC relied on facts available with respect to some elements of the calculation in determining the overall margin for each of the three investigated respondents.
7.86.
In order to resolve the issue that is before us, we must determine whether Article 9.4 "admits of" the interpretation put forward by the United States, and set out in the relevant US law. Specifically, we must consider whether the phrase "shall disregard…margins established under the circumstances referred to in paragraph 8 of Article 6" can, as United States contends, be understood to be limited to margins established entirely under the circumstances referred to in paragraph 8 of Article 6. In our view, it can not.
7.87.
We note first that the exclusion provided for in Article 9.4 is mandatory - there are no exceptions provided for. Thus, any margins that fall within the categories established in Article 9.4 must not be considered in determining the maximum duty applicable to uninvestigated producers. Second, we note that the category of "margins established under the circumstances referred to in [Article 6.8]" is not qualified in any way. Thus, in our view, the provision is clear and explicit on its face. Any margin established under the circumstances referred to in Article 6.8 must be disregarded in determining the maximum anti-dumping duty applicable to uninvestigated producers.
7.88.
The question then becomes when is a margin "established under the circumstances referred to in [Article 6.8]". In the United States' view, a dumping margin refers only to the overall margin established for a particular product from a particular source. Consequently, a margin is only "established" based on facts available if the overall margin attributed to a particular producer or exporter was determined on the basis of facts available. Any "interim" use of facts available on the way to determining the margin does not, in the United States' view, result in a margin established under the circumstances referred to in Article 6.8
7.89.
We note, however, that Article 6.8 itself does not refer to the establishment of margins per se, but rather specifies that in certain circumstances a determination may be made on the basis of facts available. We can perceive of no textual basis in Article 6.8 to suggest that a determination should be considered made under the circumstances referred to in that Article only if the determination is made entirely on the basis of facts available. We generally agree with the United States that a "margin" is the overall margin for a particular product from a particular source.69 Where we part company from the United States is in our understanding of what it means to "establish a margin under the circumstances referred to in [Article 6.8]". The establishment of a dumping margin is a complex calculation comprising many elements. However, the "determination" with respect to the margin of dumping is the end result of all the calculation steps - the final margin that may be applied to the dumped products from the particular source. In our view, a margin determined under the circumstances referred to in Article 6.8 includes a margin determined on the basis of a calculation in which some element was established on the basis of facts available.70
7.90.
We therefore conclude that the US statute governing the calculation of the all others rate, section 735(c)(5) of the Tariff Act of 1930, as amended, is, on its face, inconsistent with Article 9.4 of the AD Agreement insofar as it requires the consideration of margins based in part on facts available in the calculation of the all others rate.71 Having found the statute governing the United States' actions in this regard inconsistent with the AD Agreement, and there being no dispute that the USDOC applied that statute in its determination in this case, we must perforce conclude that the calculation of the all others rate in this case was inconsistent with the United States' obligations under Article 9.4 of the AD Agreement. In addition, having determined that the statute is inconsistent on its face with the relevant specific provision of the AD Agreement, we consequently conclude that the United States acted inconsistently with Article 18.4 of the AD Agreement and Article XIV:4 of the Marrakesh Agreement in maintaining that provision after the entry into force of the AD Agreement.

3. Alleged violations of Article 2 of the AD Agreement in the exclusion of certain home market sales to affiliates and their replacement with downstream sales in USDOC's determination of normal value

(a) Arguments

7.91.
Japan argues that USDOC's exclusion of certain home market sales to affiliates from the determination of normal value, based on the application of the "99.5 per cent" or "arm’s length" test, and the replacement of such sales with re-sales by the affiliates to unaffiliated customers, is inconsistent with Articles 2.1, 2.2 and 2.4 of the AD Agreement. Japan challenges USDOC’s established practice in this regard on its face, and USDOC’s application of that practice in the investigation of imports of hot-rolled steel from Japan.
7.92.
Japan agrees with the general definition of the term "ordinary course of trade" used by the United States -- "the conditions and practices which, for a reasonable period of time prior to the exportation of the subject merchandise, have been normal" for sales of the foreign like product.72 In addition, Japan appears to agree that sales to affiliated purchasers may not be in the ordinary course of trade. However, Japan argues that the "arm's length" test applied by the United States is an unreasonable basis for determining whether such sales are in the ordinary course of trade, and that Article 2 does not allow a Member to treat sales that fail the "arm's length" test as "outside the ordinary course of trade". Japan argues that there is nothing in the AD Agreement that supports the premises of the "arm's length" test - that sales made to affiliates73 at average prices more than 0.5 per cent below the average prices for the same product sold to unaffiliated customers are outside the "ordinary course of trade". According to Japan, a 0.5 percentage point average price differential is too small a difference upon which to base a finding that sales to affiliates are not made in the ordinary course of trade. Japan submits that Article 2.2 of the AD Agreement makes clear that the exclusion of sales as outside the ordinary course of trade is a rigorous undertaking, and that the "arm's length" test is too mechanical and not consistent with the rigorous tests applicable to determining whether sales below cost may be considered outside the ordinary course of trade.
7.93.
Second, Japan argues that Article 2.2 of the AD Agreement prescribes what an authority shalldo if there are no home market sales in the ordinary course of trade. In Japan's view, Article 2.2 does not permit the replacement of home-market sales to an affiliate with the affiliate's re-sales.74 Article 2.2 of the AD Agreement provides that in such a case, the authorities must compare export price either with sales to a third country or with a constructed value (cost of production plus a reasonable amount for administrative, selling and general costs and for profits). Japan submits that only Article 2.3 of the AD Agreement concerning export price expressly provides for the possibility that the investigating authority may construct a price on the basis of the price at which the product is first resold to an independent buyer. On the basis of the principle "expressio unius est exclusio alterius", Japan argues that the absence of such a power in Article 2.2 of the AD Agreement implies that the USDOC practice is not permitted in the context of a determination of normal value.
7.94.
Finally, Japan asserts that USDOC practice to exclude sales that fail the "arm's length" test violates the requirement of Article 2.4 of the AD Agreement to make a "fair comparison" between normal value and export price. Japan argues that a "fair" comparison does not permit statistically arbitrary rules that reject low-priced sales from the calculation of normal value thereby artificially inflating the dumping margin. Japan submits that there are two main problems with the test, first that it tests only for lower prices and considers higher prices to be normal, and second that it fails to account for the degree of variability in prices, producing absurd outcomes. Japan submits that a standard deviation analysis that captures both the frequency and the magnitude of the variation or some other statistically valid test could ensure a fair comparison.75
7.95.
Japan submits that the use of downstream sales to replace home market sales that failed the "arm's length" test, also violates the requirement of Article 2.4 that a fair comparison be made with the respondent's US sales. Japan argues that prices of downstream sales can only be higher than the prices of a producer's own direct sales76, and the downstream home-market sales are often made at a different level of trade and therefore cannot be compared in a fair manner to export sales made directly to unaffiliated customers.
7.96.
The United States argues that Article 2.1 of the AD Agreement, which requires that normal value be based on sales made in the ordinary course of trade, allows for more than one permissible interpretation. The United States submits that the USDOC's "arm's length" test of sales to affiliates is one way of examining whether sales were made in the ordinary course of trade. The United States asserts that it is generally recognized that sales to affiliates are suspect and it is expressly recognized in Article 2.3 of the AD Agreement that association may lead to prices that are unreliable. The United States points out that other Members have a similar practice of doubting the reliability of prices of sales to affiliates.77
7.97.
Since Article 2.1 of the AD Agreement does not specify how to determine whether sales are made in the ordinary course of trade, the United States asserts that the "arm's length" test is one permissible way of making this determination, on the basis of consideration whether sales to affiliates are made at prices that are comparable to those of sales to unaffiliated customers. In the United States view, in the absence of guidance in the AD Agreement on how to assess whether sales are outside the ordinary course of trade, it cannot be argued that a difference of 0.5 percentage points between the prices of sales to affiliated and unaffiliated customers is too small. The United States submits that the authority is free under the AD Agreement to consider a difference of 0.5 per cent significant.78
7.98.
The United States considers that USDOC's "arm's length" test, which compares the average price of sales to each affiliated customer to the average price of sales of the same product by the same producer to all unaffiliated customers, is preferable to the alternative suggested by Japan, because it focuses on the relationship between the seller and the customer, not on a particular product. The United States believes that the standard deviation analysis suggested by Japan would lower the threshold and provide no certainty that sales included in the calculation of normal value are not affected by the relationship between the seller and the buyer. Moreover, USDOC's weighted average methodology is consistent with the way dumping margins are normally calculated under the AD Agreement.79 The United States further asserts that USDOC may otherwise consider aberrationally high prices, as well as prices that fail the "arm's length" test, to be outside the ordinary course of trade - these are simply not what is being tested for by the "arm's length" test.80
7.99.
The United States maintains that nothing in the AD Agreement precludes the replacement of sales excluded as a result of the application of the "arm's length" test with downstream sales to unaffiliated customers. The United States submits that Article 2.2 of the AD Agreement does not give a definition of the ordinary course of trade and only deals with the situation where there are no such sales or insufficient sales, which was not the case here. There is no provision that prescribes how sales to affiliates should be treated. Article 2.2.2 of the AD Agreement concerning sales below cost is merely one example of sales that are not considered to be made in the ordinary course of trade and does not determine the treatment of sales to affiliates. In sum, the United States considers that the use of downstream sales to replace sales to affiliates that failed the arm's length test falls within the definition of normal value of Article 2.1 of the AD Agreement.
7.100.
The United States considers that Article 2.2 of the AD Agreement expresses a clear preference for actual home market sales over sales to a third country or a construction based on cost of production. In this context, the United States asserts that the downstream sales it uses to replace excluded sales to affiliates are, in fact, "sales in the ordinary course of trade" in the home market, and as such, are preferable to the alternatives provided for in Article 2.2. According to the United States, Japan’s argument in fact leads to the absurd result that as soon as sales are made to affiliates, normal value would have to be either constructed based on cost of production or based on sales to third countries.
7.101.
The United States submits that USDOC establishes normal value in a permissible way when it excludes sales to affiliates that fail the "arm's length" test, and when it replaces those sales with re-sales to unaffiliated customers, and then compares the determined normal value and export price in accordance with the requirements of Article 2.4, i.e. at the same level of trade, and making due allowance for differences that affect price comparability.81 Therefore, the United States asserts that Japan’s claim that USDOC’s "arm's length" test and its use of downstream sales in certain circumstances violates the "fair comparison" requirement of Article 2.4 of the AD Agreement must fail.
7.102.
Brazil submits that there is no legal basis in the AD Agreement for USDOC's "arm's length" test, and objects to the very low threshold of affiliation applied by the United States, which, in Brazil's view, will very often lead to the exclusion of perfectly normal sales under the "arm's length" test. It considers that Japan correctly presents the WTO inconsistencies of USDOC's "arm's length" test. Brazil submits that the test excludes prices that are clearly comparable under any reasonable standard. Moreover, Brazil argues, the averaging methodology used in the test removes prices that might be higher than most sales to unaffiliated customers and products that might prove to be a more appropriate match to export sales.
7.103.
Brazil also argues that there is no textual basis for USDOC' s replacement of certain affiliated party sales with the resale price of the downstream sale. Brazil argues that the silence in Article 2.2 as to the use of resale prices in determining normal value is meaningful in light of the express provisions concerning the use of resale prices in determining export price under Article 2.3 of the AD Agreement. Brazil is of the opinion that the substitution of higher downstream resale prices increases the normal value so as to vitiate any fair comparison. Brazil concludes that the fact that USDOC only disregards the lower priced sales that fall outside the "arm's length" test and replaces certain related-party sales with higher downstream prices demonstrates the bad faith in which the United States has implemented the AD Agreement.
7.104.
Korea argues that the "arm's length" test is inconsistent with the AD Agreement. In Korea's view, the only basis for considering home market sales as outside the ordinary course of trade is set out in Article 2.2.1 concerning sales below cost, and even then only under certain conditions. In Korea's view, the "arm's length" test mixes together all models of subject merchandise sold to affiliated and unaffiliated parties, without taking into account differences in prices and/or product that existed independent of the factor of affiliation. Moreover, Korea considers the test biased since USDOC disregards only lower priced sales, guaranteeing that higher-priced sales remain in the database for the calculation of normal value.
7.105.
Chile supports Japan's view that the "arm's length" test, because it excludes only lower priced sales to affiliated customers, does not allow for a fair comparison between normal value, which is artificially inflated as a result of the application of the test, and export price. Moreover, Chile considers that a difference of 0.5 per cent in price does not constitute a sufficiently significant price difference.
7.106.
According to the EC, the "arm's length" test applied by the US authorities is not a "permissible" interpretation of the terms "in the ordinary course of trade" in Article 2.1. The EC considers that it is unreasonable and contrary to Article 2.1 for the US authorities to treat in all circumstances a 0.5 per cent difference in average prices as irrefutable evidence that sales are not made in the ordinary course of trade.

(b) Finding

7.107.
The parties are in general agreement that sales between affiliated parties may not be in the ordinary course of trade, and therefore not included in the determination of normal value.82 However, Japan disagrees with: (i) the "arm's-length" test applied by the USDOC in determining whether affiliated party sales are not in the ordinary course of trade, and (ii) the methodology applied by the USDOC in using the resale price of the affiliated purchaser as a substitute price for sales excluded from the calculation of normal value on the basis of the application of the "arm's length" test.83

(i) The use of the "arm's-length" test by USDOC in determining whether affiliated party sales are in the ordinary course of trade

7.108.
Turning to the first issue, we note that Article 2.1 of the AD Agreement specifies that a product is to be considered as dumped if the export price is less than "the comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country." However, the AD Agreement does not define the concept of "ordinary course of trade", either in Article 2.1 or elsewhere, and establishes no general tests for determining whether sales are made in the ordinary course of trade, or not.84 It seems clear to us, and the parties do not dispute, that investigating authorities must determine whether sales in the home market are made in the ordinary course of trade in order to determine which of these sales are to be considered in the determination of normal value. The parties also seem to be in general agreement that sales to affiliates may, in some circumstances, be made not in the ordinary course of trade. It thus seems indisputable that an investigating authority may "test" home market sales to affiliated customers to decide whether they are made in the ordinary course of trade and consequently are to be considered in determining normal value. The difference between the parties is in their position as to whether the USDOC's "arm's length" test is an appropriate basis for making this decision.
7.109.
The "arm's length" test, as we understand it, is intended to test for differences in pricing to affiliated customers as compared with pricing to unaffiliated customers. In the United States' view, such differences demonstrate that sales to affiliated customers are not in the ordinary course of trade. We can certainly accept, as Japan appears to accept, that a pattern of prices to affiliated customers that is different from the pattern of prices to unaffiliated purchasers might support a conclusion that sales to affiliated customers are not in the ordinary course of trade. However, a test intended to distinguish sales that are "in the ordinary course of trade" from those that are not must be based on a permissible interpretation of that term as used in the Agreement.
7.110.
Our concern with the "arm's length" test arises because it does not, in fact, test for differences in prices of sales to affiliated customers as compared with unaffiliated customers, which might indicate that sales are not made in the ordinary course of trade. Rather, the "arm's length" test only tests whether prices to affiliated customers are lower, on average, than prices to unaffiliated customers.85 There is no reason to suppose, and the United States has not proposed any, that affiliation only results in sales that are outside the ordinary course of trade because they are lower priced on average than sales to unaffiliated customers. One example of prices to affiliated customers that are higher as a result of affiliation, and might be considered not in the ordinary course of trade, would be where prices between affiliates are established in order to allocate profits, and consequently tax burdens, among affiliates. These prices might, on average, be higher than prices to unaffiliated customers, but would not be caught by the USDOC's "arm's length" test.
7.111.
The United States argues before us that it would, if the situation arose, test for "aberrationally high" prices to affiliated customers. However, merely that the United States might apply a different test in other circumstances does not mean that the "arm's length" test is based on a permissible interpretation of "sales in the ordinary course of trade". Moreover, it is clear that the "arm's length" test was applied in this case without consideration of any particular factual circumstances. USDOC stated, in the preliminary determination

"Sales to affiliated customers in the home market not made at arm's length prices (if any) were excluded from our analysis because we considered them to be outside the ordinary course of trade. See 19 CFR 351,102. To test whether these sales were made at arm's length, we compared on a model-specific basis the prices of sales to affiliated and unaffiliated customers net of all discounts, rebates, billing adjustments, movement charges, direct selling expenses, and packing. Where, for the tested models of subject merchandise, prices to the affiliated party were on average 99.5 per cent or more of the price to unaffiliated parties, we determined that sales made to the affiliated party were at arm's length and used those sales in determining [normal value]. …86

7.112.
The result of application of the "arm's length" test, in this case and in general, is the exclusion from the determination of normal value of prices that are, on average, lower. As a result, the application of the "arm's length" test cannot but skew the normal value upward, thereby making a finding of dumping, or a higher margin of dumping, more likely.87 This reinforces our view that the "arm's length" test does not rest on a permissible interpretation of the term "sales in the ordinary course of trade".

(ii) The replacement of excluded sales with sales by affiliated purchasers in the determination of normal value

7.113.
Turning to the second issue, the replacement, in the calculation of normal value, of "excluded" sales by downstream sales, we note that Article 2.1 establishes that normal value is the "comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country". The United States argues that the downstream prices it uses as replacements for excluded home market sales to affiliated companies meet this definition, and that therefore, the USDOC decision to use these prices in establishing normal value is based on a permissible interpretation of the Agreement. We do not agree.
7.114.
It is important to keep in mind the overall object and purpose of the AD Agreement, to establish rules for the imposition of anti-dumping duties. Among these is the obligation, set out in Article 6.10, to "as a rule, determine an individual margin of dumping for each known exporter or producer concerned of the product under investigation". To this end, investigating authorities routinely collect information from the known exporters and producers concerned regarding their home market and export sales, in order to enable the calculation of a dumping margin. In our view, the "comparable price, in the ordinary course of trade" on the basis of which normal value is to be determined under Article 2.1 must be the price of the sales by each known exporter or producer for which a dumping margin is calculated. The "replacement" prices used in this case in the calculation of normal value for investigated Japanese producers were the prices of sales made by affiliates of the companies being investigated for purposes of determining whether dumping was occurring and if so, the margin of dumping. While it may be true that those sales were, in the broad sense, in the ordinary course of trade, in our view they are simply not sales which may be taken into account in determining normal value for the companies for which dumping margins were being established, as they are not sales in the ordinary course of trade of those companies.
7.115.
We consider that the overall structure of Article 2 supports our conclusion in this regard. Article 2.1 defines dumping as a situation where export price is lower than normal value, which is the "comparable price, in the ordinary course of trade, for the like product when destined for consumption in the exporting country". Article 2.2 then provides alternative methods for establishing normal value when there are no such sales, the volume of such sales is too low to permit a proper comparison, or the particular market situation does not permit a proper comparison. Article 2.3 next provides alternative methods for establishing export price in situations where there is no export price or the export price is deemed unreliable.
7.116.
Once normal value and export price have been determined in accordance with the provisions of Articles 2.1 through 2.3, Article 2.4 then establishes rules governing the comparison of normal value and export price.88 Thus, Article 2 as a whole sets out the basic rules for all aspects of the determination of dumping. Of course, investigating authorities will have to make numerous decisions along the way to determining a margin of dumping for each investigated company, not all of which are specifically addressed in Article 2 itself. However, these decisions must, in all cases, not be inconsistent with the specific requirements of Article 2, as well as the rest of the AD Agreement.
7.117.
The alternative methods for the calculation of normal value and export price provided for in Articles 2.2 and 2.3 of the AD Agreement are not the same. We see no basis on which to conclude that because Article 2.3 allows for the construction of an export price on the basis of a first resale to an independent buyer, a similar action must be allowed for the determination of normal value. It is by no means clear to us that calculation of a constructed normal value by a method parallel to that provided for constructed export price would be acceptable under the AD Agreement.89 In any event, however, that is not what USDOC did in this case. There was no attempt to make allowances for costs, including duties and taxes, incurred between the original sale to the affiliated purchaser and the first resale to an independent buyer, as is required when export price is constructed pursuant to Article 2.3. The consideration of level of trade does not compensate for this lack.90 In our view, the replacement of excluded sales by investigated companies to affiliates with the downstream sales by those affiliates in the calculation of normal value is inconsistent with the AD Agreement.
7.118.
There was no allegation in this case that there were no or insufficient sales in the ordinary course of trade by the investigated companies to allow for the calculation of normal value on the basis of those sales, as required by Article 2.1. Neither party contends that there was a need to calculate normal value according to one of the alternate methods provided for in Article 2.2. Thus, in our view, in order to be consistent with Article 2.1, normal value was to be determined on the basis of the prices of sales made by the investigated companies themselves, in the ordinary course of trade. We can see no basis in the AD Agreement for the replacement of certain excluded home market sales by downstream sales of the goods in the calculation of normal value for the investigated respondents in this case. We therefore conclude that the "replacement" of excluded sales to affiliates with the sales by those affiliates to downstream purchasers in this case was not consistent with Article 2.1 of the AD Agreement.

(iii) Additional findings

7.119.
Having found that the "arm's length" test does not relate to a permissible interpretation of the term "sales in the ordinary course of trade", we conclude that its application in this case led to a determination as to whether certain sales were made in the ordinary course of trade inconsistent with Article 2.1 of the Agreement. As a consequence of our finding in this regard, we do not consider it either necessary or appropriate to consider whether that test also is inconsistent with the more general obligation of fair comparison set out in Article 2.4 of the AD Agreement.
7.120.
Similarly, having found that the USDOC acted inconsistently with Article 2.1 in replacing, in the determination of normal value, certain "excluded" sales by investigated companies with downstream sales made by purchasers affiliated with the investigated companies, we do not consider it necessary to go on to consider whether the replacement of excluded sales with sales by affiliates was consistent with Articles 2.2 and 2.4 of the AD Agreement.

E. ALLEGED VIOLATIONS IN THE PRELIMINARY DETERMINATION OF CRITICAL CIRCUMSTANCES

1. Arguments

7.121.
Japan claims that USDOC's preliminary critical circumstances finding is inconsistent with Articles 10.1, 10.6 and 10.7 of the AD Agreement because (i) USITC had preliminarily found only a threat of injury to the industry while Article 10.6 of the AD Agreement requires evidence of current injury; and (ii) the preliminary determination of critical circumstances was not supported by sufficient evidence as required by Article 10.7 of the AD Agreement. Japan moreover asserts that the evidentiary standard in the US statute governing preliminary critical circumstances findings on its face is inconsistent with the "sufficient evidence" standard of Article 10.7 of the AD Agreement. Finally, Japan argues that the US statute does not require evidence of all the conditions set forth in Article 10.6 of the AD Agreement.
7.122.
Japan submits that the USDOC preliminary critical circumstances determination inevitably also is inconsistent with Article 10.1 of the AD Agreement since it allowed for the possibility that AD duties would be levied retroactively in spite of the fact that the requirements of Articles 10.6 and 10.7 of the AD Agreement had not been satisfied.
7.123.
Japan asserts that the investigating authorities cannot predicate a finding of critical circumstances on a mere threat of injury. Japan therefore claims that USDOC’s preliminary critical circumstances determination, which Japan asserts was based on a preliminary finding by the USITC of threat of injury, is inconsistent with Article 10.6 of the AD Agreement. Japan asserts, in support of its view, that Article 10.6 of the AD Agreement uses the term "injury", while Articles 10.2 and 10.4 of the AD Agreement contain a clear distinction between "injury" and "threat" thereof, and allows for retroactive imposition of duties only in the case of current material injury.
7.124.
Second, Japan argues that USDOC's preliminary critical circumstances determination is inconsistent with Article 10.7 of the AD Agreement because it was not based on "sufficient evidence" that the requirements of Article 10.6 of the AD Agreement were satisfied. Japan claims that USDOC based its critical circumstances determination on information contained in the petition and in certain press reports. In Japan's view, such information is one-sided and necessarily biased and can therefore never constitute sufficient evidence as required by Article 10.7 of the AD Agreement. In particular, Japan asserts that USDOC lacked sufficient evidence of the existence of dumping as required by Article 10.6(ii) of the AD Agreement and the chapeau of Article 10.6, since it based its conclusion entirely on information contained in the petition. Japan further argues that USDOC did not have sufficient evidence of injury to the industry caused by dumped imports, since it relied on press reports and ignored the preliminary findings of the USITC which stated that the "industry was relatively healthy during much of the period examined". Japan finally argues that USDOC lacked sufficient evidence of "massive dumped imports over a relatively short period". Japan submits that USDOC departed from its normal practice of assessing the period before and immediately after the filing of a petition, and instead picked a period of five months preceding and following April 1998 as the basis for determining whether there were massive dumped imports over a relatively short period. Japan argues that this date was arbitrarily chosen on the basis of press reports that allegedly announced the likely filing of a petition for anti-dumping measures by US producers.91
7.125.
In addition, Japan alleges that the US statutory provision governing preliminary critical circumstances determinations, section 733(e) of the Tariff Act 1930, as amended, is inconsistent on its face with Article 10.7 of the AD Agreement. That Article requires that the authorities "have sufficient evidence" that the conditions set forth in paragraph 6 of Article 10 of the AD Agreement are satisfied. Japan argues that section 733(e) of the Tariff Act of 1930, as amended, sets a lower evidentiary standard by requiring only a "reasonable basis to believe or suspect" that certain conditions are satisfied, rather than "sufficient evidence" that those conditions are satisfied.92 Moreover, Japan argues that the US statutory provisions governing critical circumstances determinations do not require all of the findings of fact required by Article 10.6 of the AD Agreement. Japan refers in particular to the absence in the US statute of a requirement to make a preliminary finding of dumping and of an assessment of whether the remedial effect of the AD duty is undermined by the dumped imports. Japan submits that the US statute also does not require sufficient evidence of the causal link between massive imports and injury.
7.126.
The United States submits that neither the US statute nor USDOC’s preliminary critical circumstances determination is inconsistent with Article 10 of the AD Agreement. The United States claims that, contrary to Japan's assertion, Article 10.6 of the AD Agreement and 10.7 of the AD Agreement expressly authorise that preliminary critical circumstances determinations be made based on a threat of material injury to the domestic industry.93 The United States refers in this regard to the ordinary meaning of the word injury. The United States argues that the term "injury" is defined in footnote 9 of the AD Agreement as "material injury to a domestic industry or threat of material injury to a domestic industry, unless otherwise specified". The United States notes that Article 10.6 of the AD Agreement does not "otherwise specify". Thus, in the US view, "injury" in Article 10.6 of the AD Agreement includes both material injury and threat thereof.
7.127.
The United States argues that Articles 10.6 and 10.7 of the AD Agreement permit "such measures be taken as may be necessary to collect AD duties retroactively" to be taken at any time after the initiation of the investigation.94 The United States submits that, in accordance with Article 10.7 of the AD Agreement, USDOC made a preliminary critical circumstances determination on the basis of sufficient evidence that the conditions set forth in Article 10.6 of the AD Agreement were satisfied.95 The United States asserts that USDOC had sufficient evidence that the importers knew or should have known that the exporter was practising dumping. The United States emphasises that the several hundreds of pages of exhibits to the petition are not "mere allegations" of dumping, but contain substantial factual information on the export price and normal value of the subject products and thus constitute evidence. On the basis of this information, knowledge of dumping was imputed to the importers on the basis of dumping margins in excess of 25 per cent.96 The United States argues that since the AD Agreement does not dictate how to determine whether the importers were aware that products were being dumped, it is both reasonable and permissible to deduce such knowledge from the degree of the dumping margin as preliminary established.97
7.128.
The United States asserts that USDOC also had sufficient evidence of massive imports over a short period of time. USDOC compared two six month periods and established that there was an increase in imports of 100 per cent. The United States asserts that nothing in the AD Agreement dictates which date to choose to assess whether there have been massive imports over a short period. Therefore, USDOC was permitted to choose the date on which it became common knowledge that anti-dumping proceedings would be initiated in the near future, and the date of April 1998 was therefore reasonable. The United States argues that because petitioners wait to submit their petition in order to gather more evidence does not mean that they should be deprived of their remedy against massive dumped imports that entered the country in anticipation of the anti-dumping investigation.98
7.129.
The United States refutes Japan’s challenge to the consistency of section 733(e) of the Tariff Act of 1930, as amended. First, the United States submits that it is clear that section 733(e) does not mandate any WTO inconsistent action and can therefore not be found to be inconsistent on its face with the AD Agreement. Moreover, the United States submits, the evidentiary standard of "a reasonable basis to believe or suspect" is similar to that of "sufficient evidence" and both are used interchangeably by USDOC.99 The United States asserts that it is not a lower evidentiary standard. The United States further argues that it is general USDOC practice to make all the determinations as required by Article 10.6 of the AD Agreement concerning massive imports, knowledge of dumping and injury and the causal link between the dumping and the injury.
7.130.
Brazil supports Japan's argument that the US critical circumstances determination was inconsistent with Article 10.6 of the AD Agreement, as in Brazil's view a preliminary finding of material injury to the industry and not just threat thereof is required. Brazil argues that the USDOC determination was not based on sufficient evidence as required by Article 10.7 of the AD Agreement but on mere allegations of the petitioners. Moreover, Brazil submits that the evidentiary standard in the US statutory provisions is lower than that set forth in the AD Agreement and that US law does not require a finding of all the elements of fact of Article 10.6 of the AD Agreement.
7.131.
Korea agrees with Japan's view that USDOC's critical circumstances determination was not based on sufficient evidence of current injury, but only of threat of injury, and is therefore inconsistent with Articles 10.6 and 10.7 of the AD Agreement. Korea asserts that this interpretation, that evidence of current material injury is necessary, comports with the limited object and purpose of Article 10.6 of the AD Agreement which is to assure that the remedial effects of the final duties are not eviscerated. Korea argues that, if only threat of injury exists, the remedial effect will not be undermined since the prospective application of the duties will precisely prevent injury from occurring.
7.132.
Chile is of the opinion that information from petitioners is not "sufficient evidence" and the USDOC critical circumstances determination therefore is inconsistent with Articles 10.7 of the AD Agreement.

2. Finding

7.133.
We recall certain of the facts that are relevant to our examination of the matter before us. On 8 October 1998, USDOC issued a policy bulletin stating that the USDOC would, if adequate evidence of critical circumstances was available, issue preliminary critical circumstances determinations prior to preliminary dumping determinations.100 On 30 November 1998, USDOC issued an affirmative preliminary critical circumstances determination regarding imports of hot-rolled steel from Japan.
7.134.
Although USDOC made a preliminary determination of critical circumstances, no measures "necessary to collect anti-dumping duties retroactively" were actually taken until the preliminary determination of dumping by USDOC, effective 19 February 1999.101 USDOC made a second and final critical circumstances determination as part of its final dumping determination on 6 May 1999. Under US law, however, it is the USITC, in its final determination of injury, which determines whether critical circumstances exist that warrant the retroactive application of duties to 90 days prior to the date of application of provisional measures. USITC in its final injury determination of 23 June 1999 made a negative critical circumstances finding. USITC concluded that "we do not find that the record evidence indicates that the subject imports from Japan would seriously undermine the remedial effects of the order".102 Therefore, anti-dumping duties were ultimately not collected retroactively.
7.135.
Japan is challenging the consistency of the USDOC preliminary critical circumstances determination with Articles 10.6 and 10.7 of the AD Agreement. Japan claims that by violating these two provisions, USDOC also acted inconsistently with Article 10.1 of the AD Agreement.
7.136.
Article 10.1 of the AD Agreement reads as follows:

"Provisional measures and anti-dumping duties shall only be applied to products which enter for consumption after the time when the decision taken under paragraph 1 of Article 7 and paragraph 1 of Article 9, respectively, enters into force, subject to the exceptions set out in this Article"

7.137.
Articles 10.6 and 10.7 of the AD Agreement provide that:

"10.6 A definitive anti‑dumping duty may be levied on products which were entered for consumption not more than 90 days prior to the date of application of provisional measures, when the authorities determine for the dumped product in question that:

(i) there is a history of dumping which caused injury or that the importer was, or should have been, aware that the exporter practises dumping and that such dumping would cause injury, and

(ii) the injury is caused by massive dumped imports of a product in a relatively short time which in light of the timing and the volume of the dumped imports and other circumstances (such as a rapid build‑up of inventories of the imported product) is likely to seriously undermine the remedial effect of the definitive anti‑dumping duty to be applied, provided that the importers concerned have been given an opportunity to comment.

10.7 The authorities may, after initiating an investigation, take such measures as the withholding of appraisement or assessment as may be necessary to collect anti‑dumping duties retroactively, as provided for in paragraph 6, once they have sufficient evidence that the conditions set forth in that paragraph are satisfied".

7.138.
Japan’s claims concern only the US preliminary critical circumstances determination, which is governed by Article 10.7 of the AD Agreement, which authorizes preliminary measures necessary to collect duties retroactively. We will first address Japan's claims concerning the US statutory provisions on critical circumstances. Japan argues that the US statute does not require a finding of all the conditions necessary for the retroactive imposition of duties and sets an evidentiary standard which is lower than the "sufficient evidence" standard of the AD Agreement. Secondly, we will discuss Japan's claim that USDOC acted inconsistently with Article 10.6 of the AD Agreement by making an affirmative critical circumstances determination in the absence of a finding of current material injury to the industry. Finally, we will address Japan's claim that USDOC's preliminary critical circumstances determination was not supported by sufficient evidence of all the conditions of Article 10.6 of the AD Agreement and therefore was inconsistent with Article 10.7 of the AD Agreement.

(a) Are the US statutory provisions concerning critical circumstances consistent with the Agreement with respect to the evidentiary standard it sets forth and the conditions of application it requires?

7.139.
Section 733(e)(1) of the Tariff Act of 1930, as amended, requires USDOC to make certain preliminary determinations in a case in which a petitioner requests the imposition of anti-dumping duties retroactively for 90 days prior to a preliminary determination of dumping. The statute provides:

"If a petitioner alleges critical circumstances in its original petition, or by amendment at any time more than 20 days before the date of a final determination by [USDOC], then [USDOC] shall promptly (at any time after the initiation of the investigation under this part) determine, on the basis of the information available at that time, whether there is a reasonable basis to believe or suspect that—

(A) (i) there is a history of dumping and material injury by reason of dumped imports in the United States or elsewhere of the subject merchandise, or

(ii) the person by whom, or for whose account, the merchandise was imported knew or should have known that the exporter was selling the subject merchandise at less than its fair value and that there was likely to be material injury by reason of such sales, and

(B) there have been massive imports of the subject merchandise over a relatively short period".103

7.140.
Japan argues that the evidentiary standard set forth in the statute is inconsistent with the requirements of Article 10.7, and also that the statute does not require all of the conditions of the AD Agreement for making a preliminary critical circumstances determination.
7.141.
It is well established in GATT/WTO practice that a statute is inconsistent on its face with a Member’s WTO obligations only if it is mandatory and requires WTO inconsistent action or prohibits WTO consistent action.104 The Appellate Body recently stated, in United States – Anti-Dumping Act of 1916 :

"88. As indicated above, the concept of mandatory as distinguished from discretionary legislation was developed by a number of GATT panels as a threshold consideration in determining when legislation as such – rather than a specific application of that legislation – was inconsistent with a Contracting Party's GATT 1947 obligations. The practice of GATT panels was summed up in United States – Tobacco as follows:

… panels had consistently ruled that legislation which mandated action inconsistent with the General Agreement could be challenged as such, whereas legislation which merely gave the discretion to the executive authority of a contracting party to act inconsistently with the General Agreement could not be challenged as such; only the actual application of such legislation inconsistent with the General Agreement could be subject to challenge. (emphasis added)

89. Thus, the relevant discretion, for purposes of distinguishing between mandatory and discretionary legislation, is a discretion vested in the executivebranch of government". (footnotes omitted)105

We therefore consider whether the statute in question requires USDOC to take action which contravenes the US obligations under the WTO AD Agreement.

7.142.
The statute provides that if the petitioner alleges critical circumstances, USDOC shall promptly determinewhether there is a reasonable basis to believe or suspect that critical circumstances exist. On the basis of this determination such measures may be taken as necessary to collect anti-dumping duties retroactively. We do not believe that this provision of the US statute requires USDOC to take WTO inconsistent action. Nor does it preclude USDOC from acting consistently with the Agreement.
7.143.
First, the evidentiary standard set forth in the US statute is "a reasonable basis to believe or suspect". Article 10.7 of the AD Agreement on the other hand uses the term "sufficient evidence". It is a well accepted principle of international law that for the purposes of international adjudication national law is to be considered as a fact.106 The analysis of the consistency of the US statute with Article 10.7 must take into account, therefore, its application in practice, as interpreted and applied by the administering and judicial authorities. We recognize that the actual terms used in the US statute differ from those of the Agreement. However, we believe that the consistency of this evidentiary standard is not determined by a semantic difference. Rather, we must examine how this standard has been applied in practice.
7.144.
In our view, "sufficient evidence" refers to the quantum of evidence necessary to make a determination. "A reasonable basis to believe or suspect" on the other hand, seems to refer to the conclusion reached on the basis of evidence presented, that is, a legal mindset that certain facts exist, based on the evidence presented. It appears that in past cases the US authorities have applied the standard as set out in the statute interchangeably with a standard expressed as "sufficient evidence" and have made affirmative determinations when sufficient evidence was adduced that the conditions of application were satisfied.107 We therefore consider that the US statute, as it has been applied is not inconsistent with the requirement of the AD Agreement that the investigating authority must have sufficient evidence of the conditions of Article 10.6 before taking measures necessary to collect the duties retroactively.108
7.145.
Japan further argues that the US statute does not require evidence that all the conditions of Article 10.6 of the AD Agreement are satisfied, as required by Article 10.7. Japan claims in particular that the statute does not require sufficient evidence of dumping, injury and causation, and that it does not require evidence that massive dumped imports are likely to seriously undermine the remedial effect of the duty. We recall that the question we must address in this regard is whether the statute requires action inconsistent with, or prevents actions consistent with, the requirements of the Agreement.
7.146.
In our view, the US statute allows the investigating authority to make its determinations consistently with the AD Agreement in this respect. We recognise that the statute does not explicitly set out the same requirements as are set out in Article 10.6. However, this does not imply that USDOC is precluded from taking these elements into consideration, in so far as necessary. In our view, the text of the US statute in this regard does not preclude USDOC from determining whether there is sufficient evidence that the conditions set out in paragraph 10.6 are satisfied. The question then becomes whether USDOC did so in this case. We will discuss this question below.
7.147.
We note that Article 10.7 requires that there be sufficient evidence that the conditions of Article 10.6 are satisfied. Article 10.6 of the AD Agreement of course presupposes a final dumping and injury determination, without which no definitive dumping duties may be applied in any case. Rather than being conditions set out in Article 10.6, we consider that findings of dumping and injury are a precondition for any definitive duty to be applied. Article 10.7 of the AD Agreement provides that certain preliminary measures may be taken "after initiation". This implies that at the time of the critical circumstances determination, the authority has already determined, under Article 5.3, that the petition contained sufficient information of dumping, injury, and a causal link to justify the initiation of the investigation. For a preliminary critical circumstances determination, Article 10.7 requires, in addition, sufficient evidence of the specific conditions of Article 10.6 as set forth in 10.6 (i) and (ii). It does not, however, in our view necessarily require additional or different evidence of dumping or injury from that on which the decision to initiate was based.
7.148.
We note that the US statute governing preliminary critical circumstances determinations does not expressly refer to the question whether massive dumped imports seriously undermine the remedial effect of the duty. However, we do not consider that the Agreement requires that a separate determination be made with regard to this aspect of Article 10.6 at the preliminary stage of considering whether to take action under Article 10.7. Rather than a "condition" of Article 10.6 of which there must be sufficient evidence in order to act under Article 10.7, in our view, this requirement establishes the conclusion that must be reached in order to justify retroactive application of the anti-dumping duty under Article 10.6.109 Consideration of this question at the preliminary stage of deciding whether to apply measures under Article 10.7 would, in our estimation, at best be speculative. Our view is reinforced by the fact that the possible undermining of the remedial effect of a definitive anti-dumping duty is not a question of which evidence would be available at the very early stages of an investigation, after initiation, when the determination under Article 10.7 may be made and authorized precautionary measures taken. The conclusion that the remedial effect of a definitive duty would be undermined by the effect of massive dumped imports can only meaningfully be addressed at the end of the investigation, when it has been determined that the imposition of a definitive anti-dumping measure is warranted, based on a final determination of dumping, injury, and causal link. To require investigating authorities to undertake what is likely to be an impossible, meaningless task under Article 10.7 is not, in our view, necessary or appropriate.
7.149.
Moreover, in this respect, we note the US regulation set out in 19 CFR § 351,206 (h). It provides that, in assessing whether imports of the subject merchandise have been massive, USDOC is to examine the volume and value of the imports, the seasonal trends and the share of domestic consumption accounted for by the imports, and establishes that imports over a relatively short period of time may be determined based on the knowledge of exporters that an anti-dumping proceeding was likely or had been initiated. We recall that Article 10.6 (ii) of the AD Agreement provides that injury must be caused by massive dumped imports "which in light of the timing and the volume of the dumped imports and other circumstances (such as a rapid build-up of inventories) is likely to seriously undermine the remedial effect of the definitive anti-dumping duty to be applied". Thus, the Agreement requires that the likelihood that the remedial effect of the duty will be undermined be assessed in light of timing and volume of the dumped imports. In our view, by requiring that the assessment of massive dumping in a relatively short period be made in light of the exporters' knowledge of an initiation or a likely initiation, USDOC addresses whether massive imports are likely to seriously undermine the remedial effect of the duty.
7.150.
On the basis of the foregoing, we conclude that the US statute, section 733(e) of the Tariff Act of 1930, as amended, is not, on its face, inconsistent with Articles 10.1, 10.6 and 10.7 of the AD Agreement. Having reached this conclusion, we also find that the United States has not acted inconsistently with its obligations under Article XVI:4 of the Marrakesh Agreement and Article 18.4 of the AD Agreement in maintaining this statutory provision.

(b) Is the USDOC preliminary critical circumstances determination concerning hot-rolled steel from Japan inconsistent with Articles 10.6 and 10.7 of the AD Agreement ?

7.151.
Japan further challenges the specific preliminary critical circumstances determination made by USDOC in the investigation of imports of hot-rolled steel from Japan. As a preliminary matter, we note that we understand Japan to argue that a Member is precluded from making a preliminary determination of critical circumstances in the absence of a preliminary determination of material injury to the domestic industry. According to Japan, USDOC's preliminary determination of critical circumstances thus violated Article 10.6 of the AD Agreement since USITC had found threat of injury to the industry, but not current material injury, in its preliminary determination. However, Article 10.6 sets out the conditions for retroactive application of "definitive anti-dumping duties"(emphasis added). In the case of imports of hot-rolled steel products from Japan, no duties were actually levied retroactively, since USITC in its final determination of injury found that the conditions of Article 10.6 were not satisfied. In our view, Article 10.6 does not directly govern the determination at issue here – rather, USDOC's preliminary critical circumstances determination must be judged against the obligations set out in Article 10.7. Those obligations, while related to the obligations set out in Article 10.6, are not necessarily identical. This is not to say that the basis of the final injury determination is irrelevant to whether definitive duties may be levied retroactively under Article 10.6. It is only that in this case, since we are not considering whether there is a violation of Article 10.6, we need not determine whether, under Article 10.6, duties can only be levied retroactively if there is a final determination of material injury and not where there is a final determination of threat of injury. It is a different question, which we discuss below, whether a preliminary determination of critical circumstances under Article 10.7 requires sufficient evidence of current material injury to the domestic industry, or whether sufficient evidence of threat of injury may be enough.
7.152.
We first address what constitutes "sufficient evidence" for the purposes of a determination under Article 10.7. Second, we must determine what are the conditions of paragraph 6 of Article 10 of which sufficient evidence is required by Article 10.7. Finally, we consider whether USDOC's determination that there was sufficient evidence of the required elements of Article 10.6 to make an affirmative preliminary critical circumstances determination is one that an objective and unbiased investigating authority could make on the basis of the evidence that was before USDOC in this case.
7.153.
Article 10.7 of the AD Agreement does not define "sufficient evidence". However, Article 5.3 also reflects this standard, in requiring that the authorities examine the accuracy and adequacy of the evidence provided in the application "to determine whether there is sufficient evidence to justify the initiation of an investigation". The Article 5.3 requirement of "sufficient evidence to initiate an investigation" has been addressed by previous GATT and WTO panels. Their approach to understanding this standard has been to examine whether the evidence before the authority at the time it made its determination was such that an unbiased and objective investigating authority evaluating that evidence could properly have made the determination.110 These Panels have noted that what will be sufficient evidence varies depending on the determination in question. The Panel in Mexico – HFCS quoted with approval from the Panel's report in the Guatemala – Cement I case that "the type of evidence needed to justify initiation is the same as that needed to make a preliminary or final determination of threat of injury, although the quality and quantity is less".111
7.154.
The question before us is whether USDOC had sufficient evidence of the conditions of Article 10.6 to entitle USDOC to take such measures as may be necessary to collect AD duties retroactively. We are of the view that what constitutes "sufficient evidence" must be addressed in light of the timing and effect of the measure imposed or the determination made. Evidence that is sufficient to warrant initiation of an investigation may not be sufficient to conclude that provisional measures may be imposed. In a similar vein, the possible effect of the measures an authority is entitled to take under Article 10.7 of the AD Agreement informs what constitutes sufficient evidence. Whether evidence is sufficient or not is determined by what the evidence is used for. In sum, whether evidence is sufficient to justify initiation or to justify taking certain necessary precautionary measures under Article 10.7 is not a standard that can be determined in the abstract. We will therefore consider the impact of a finding of sufficient evidence for the purposes of Article 10.7 and examine the evidence on which USDOC relied in making the challenged preliminary critical circumstances determination.
7.155.
Article 10.7 provides that once the authorities have sufficient evidence that the conditions of Article 10.6 are satisfied, they may take such measures as, for example, the withholding of appraisement or assessment, as may be necessary to collect anti-dumping duties retroactively. We read this provision as allowing the authority to take certain necessary measures of a purely conservatory or precautionary kind which serve the purpose of preserving the possibility of later deciding to collect duties retroactively under Article 10.6. Unlike provisional measures, Article 10.7 measures are not primarily intended to prevent injury being caused during the investigation. They are taken in order to make subsequent retroactive duty collection possible as a practical matter. Measures taken under Article 10.7 are not based on evaluation of the same criteria as final measures that may be imposed at the end of the investigation. They are of a different kind - they preserve the possibility of imposing anti-dumping duties retroactively, on the basis of a determination additional to the ultimate final determination.
7.156.
Our understanding in this regard is confirmed by the fact that, unlike provisional measures, which can only be imposed after a preliminary affirmative determination of dumping and injury, Article 10.7 measures may be taken at any time "after initiating an investigation". In light of the timing and effect of the measures that are taken on the basis of Article 10.7, we consider that the Article 10.7 requirement of "sufficient evidence that the conditions of Article 10.6 are satisfied" does not require an authority to first make a preliminary affirmative determination within the meaning of Article 7 of the AD Agreement of dumping and consequent injury to a domestic industry. If it were necessary to wait until after such a preliminary determination, there would, in our view, be no purpose served by the Article 10.7 determination. The opportunity to preserve the possibility of applying duties to a period prior to the preliminary determination would be lost, and the provisional measure that could be applied on the basis of the preliminary affirmative determination under Article 7 would prevent further injury during the course of the investigation. Moreover, the requirement in Article 7 that provisional measures may not be applied until 60 days after initiation cannot be reconciled with the right, under Article 10.6, to apply duties retroactively to 90 days prior to the date on which a provisional measure is imposed, if a preliminary affirmative determination is a prerequisite to the Article 10.7 measures which preserve the possibility of retroactive application of duties under Article 10.6.
7.157.
The question that remains is whether USDOC had sufficient evidence that all the conditions of paragraph 6 of Article 10 of the AD Agreement were satisfied. Japan argues that USDOC did not have sufficient evidence of dumping and material injury caused by the dumped imports. Japan also submits that USDOC did not have sufficient evidence that massive dumped imports were likely to seriously undermine the remedial effect of the duty.
7.158.
We note that Japan did not challenge the initiation of the investigation, which was, pursuant to Article 5.3, based on a determination that there was sufficient evidence of dumping, injury, and a causal link. We can perceive of no reason, given the precautionary nature of the measures that may be taken under Article 10.7, why that same information might not justify a determination of sufficient evidence of dumping and consequent injury in the context of Article 10.6 as required by Article 10.7.
7.159.
Turning to the conditions of which there must be sufficient evidence, we note that Article 10.6 requires authorities to determine that, for the dumped product in question,

"(i) there is a history of dumping which caused injury or that the importer was, or should have been, aware that the exporter practices dumping and that such dumping would cause injury, and

(ii) the injury caused by massive dumped imports of a product in a relatively short time which in light of the timing and the volume of the dumped imports and other circumstances (such as a rapid build-up of inventories of the imported product) is likely to seriously undermine the remedial effect of the definitive anti-dumping duty to be applied, provided that the importers concerned have been given an opportunity to comment".

7.160.
USDOC determined that the importers knew or should have known that exporters were dumping and that such dumping would cause injury. USDOC normally considers dumping margins of 25 per cent or more and a USITC preliminary determination of material injury to impute knowledge of dumping and the likelihood of consequent material injury. USDOC determined that the information in the petition indicated that the estimated dumping margins were over 25 per cent for the Japanese respondents. The evidence of dumping in the petition was, in our view, sufficient for an unbiased and objective investigating authority to reach this conclusion. We note, in this regard, that Japan has not alleged that an imputed knowledge of dumping is, per se, inconsistent with Article 10.7, but rather argues that USDOC did not have sufficient evidence of dumping at all, for the purposes of Article 10.7.
7.161.
In this case, USITC had made a preliminary determination of threat of material injury. Consequently, USDOC looked to the information regarding injury to the domestic industry in the petition, and considered press reports regarding increasing imports, declining prices, and shifts in purchasing to import sources, as well as the USITC preliminary determination. On this basis USDOC found sufficient evidence that importers knew or should have known that material injury caused by dumping was likely. In our view, the evidence relied on by USDOC in this regard was sufficient for an unbiased and objective investigating authority to reach this conclusion.
7.162.
In any event, we note that Article 10.6 itself refers to a determination that an importer knew or should have known that there was dumping that would cause injury. The term "injury" is defined in footnote 9 to Article 3 of the Agreement to include threat of material injury or material retardation of the establishment of an industry, unless otherwise specified. Article 10.6 does not "otherwise specify". Consequently, in our view, sufficient evidence of threat of injury would be enough to justify a determination to apply protective measures under Article 10.7.
7.163.
The role of Article 10.7 in the overall context of the AD Agreement confirms this interpretation. This provision is clearly aimed at preserving the possibility to impose and collect anti-dumping duties retroactively to 90 days prior to the date of application of provisional measures. Thus, Article 10.7 preserves the option provided in Article 10.6 to impose definitive duties even beyond the date of provisional measures. Assume arguendo Article 10.7 were understood to require sufficient evidence of actual material injury. In a situation in which, at the time Article 10.7 measures are being considered, there is evidence only of threat of material injury, no measures under Article 10.7 could be taken. Assume further that in this same investigation, there was a final determination of actual material injury caused by dumped imports. At that point, it would be impossible to apply definitive anti-dumping duties retroactively, even assuming the conditions set out in Article 10.6 were satisfied, as the necessary underlying Article 10.7 measures had not been taken.112 Thus, in a sense, Article 10.7 measures serve the same purpose as an order at the beginning of a lawsuit to preserve the status quo - they ensure that at the end of the process, effective measures can be put in place should the circumstances warrant.
7.164.
The third condition of Article 10.6 of which sufficient evidence is required by Article 10.7, is that the injury be caused by massive dumped imports in a relatively short period of time. In this case, USDOC assessed the question whether there were massive dumped imports in a relatively short time by comparing imports during a period of five months preceding and following April 1998. That date was established based on press reports which, USDOC concluded, established that importers, exporters, and producers knew or should have known that an anti-dumping investigation was likely.113 USDOC found an increase of imports of hot-rolled steel of more than 100 per cent between the period December 1997-April 1998 and May-September 1998.114
7.165.
The Agreement does not determine what period should be used in order to assess whether there were massive imports over a short period of time. Japan asserts that the latter part of Article 10.6 (ii) of the AD Agreement, referring to whether the injury caused by massive imports is likely to seriously undermine the remedial effect of the duty, implies that the period for comparison is the months before and after the initiation of the investigation. Japan argues that since the duty cannot be imposed retroactively to the period before the initiation, the remedial effect of the duty cannot be undermined by massive imports before initiation.
7.166.
We disagree with this conclusion. Article 10.7 allows for certain necessary measures to be taken at any time after initiation of the investigation. In order to be able to make any determination concerning whether there are massive dumped imports, a comparison of data is obviously necessary. However, if a Member were required to wait until information concerning the volume of imports for some period after initiation were available, this right to act at any time after initiation would be vitiated. By the time the necessary information on import volumes for even a brief period after initiation were available, as a practical matter, the possibility to impose final duties retroactively to initiation would be lost, as there would be no Article 10.7 measures in place. Moreover, as with the situation if a Member were required to wait the minimum 60 days and make a preliminary determination under Article 7 before applying measures under Article 10.7, the possibility of retroactively collecting duties under Article 10.6 at the final stage would have been lost.
7.167.
Moreover, in our view, it is not unreasonable to conclude that the remedial effect of the definitive duty could be undermined by massive imports that entered the country before the initiation of the investigation but at a time at which it had become clear that an investigation was imminent. We consider that massive imports that were not made in tempore non suspectu but at a moment in time where it had become public knowledge that an investigation was imminent may be taken into consideration in assessing whether Article 10.7 measures may be imposed. Again, we emphasize that we are not addressing the question whether this would be adequate for purposes of the final determination to apply duties retroactively under Article 10.6
7.168.
We have carefully considered the information on which USDOC based its preliminary critical circumstances determination. We consider that an objective and unbiased investigating authority could, on the basis of the evidence before USDOC, determine that there was sufficient evidence that the conditions set forth in Article 10.6 were satisfied, and its preliminary critical circumstances determination is therefore consistent with Article 10.7. We therefore find that the preliminary critical circumstances determination was not inconsistent with Article 10.1 of the AD Agreement either since it complied with the conditions of Article 10.7 of the AD Agreement.

F. ALLEGED VIOLATIONS IN THE DETERMINATION OF INJURY AND CAUSATION

1. Alleged violations of Articles 3 and 4 of the AD Agreement on the face of the captive production provision and in its application by USITC in this case

7.169.
Japan claims that the US captive production provision both on its face and as applied by USITC in the case of imports of hot-rolled steel from Japan violates Articles 3 and 4 of the AD Agreement. We will first consider the on-its-face challenge of the US statutory provision. If necessary we will thereafter consider Japan's arguments concerning the application of the provision in this case.

(a) Does the captive production provision on its face violate Articles 3 and 4 of the AD Agreement

(i) Arguments

7.170.
Japan argues that the captive production provision of US law, section 771(7)(c)(iv) of the Tariff Act of 1930, on its face violates Articles 3 and 4 of the AD Agreement, which require that an authority consider a domestic industry in its entirety throughout its injury and causation analysis. Japan alleges that under the captive production provision, the USITC must focus its injury analysis on the merchant market and potentially may find material injury on the basis of the merchant market even if the industry as a whole is not experiencing material injury. Japan submits that given the mandatory nature of the captive production provision — and, therefore, the lack of discretion the USITC has in whether to apply the provision — it is inconsistent with Articles 3 and 4 on its face, regardless of its application in this case.
7.171.
Japan submits that the US statute is inconsistent with Article 4.1 of the AD Agreement. In Japan's view, the definition of domestic industry in Article 4.1 requires authorities to consider domestic producers as a whole and their overall output. Japan argues that the captive production provision and its mandatory focus on merchant market data necessarily precludes any balanced assessment of the data of an industry as a whole and more specifically ignores the attenuated nature of import competition in the captive market.115
7.172.
Japan argues that the captive production provision exaggerates the market share of imports relative to all domestic production and thus is inconsistent with Article 3.2 of the AD Agreement. Japan argues that the captive production provision narrows the analysis to the merchant market.116 Japan claims that the captive production provision also violates Article 3.4 of the AD Agreement since it requires the authority to evaluate the key factors mentioned in Article 3.4 based on a narrow segment of the industry, rather than the industry as a whole as provided for in that provision. Japan argues that the statutory provision leaves no discretion to consider fully both the merchant market and the overall industry, nor does it require an explanation of how the merchant market relates to the industry as a whole or is representative of it.
7.173.
Japan claims that the captive production provision is also inconsistent with Article 3.5 of the AD Agreement, which requires the establishment of a causal link between dumped imports and injury to the industry, because it requires the USITC to ignore the "shielding" effect of captive production and to focus instead on the injury to that portion of the industry serving the merchant market. Japan submits that the provision thus makes it impossible for USITC to consider fully "all relevant evidence before the authorities" as Article 3.5 of the AD Agreement requires.
7.174.
Japan further submits that the captive production provision violates the requirement in Article 3.6 of the AD Agreement to analyse the effect of imports on all domestic production. In sum, Japan claims that the captive production provision does not allow for an objective examination as required by Article 3.1 since an examination can only be objective if it takes into consideration all information concerning the industry as a whole. Finally, Japan alleges that the captive production provision violates Article XVI:4 of the Marrakesh Agreement since it mandates an analytic approach that focuses on one segment of the industry in violation of Articles 3 and 4 of the AD Agreement.
7.175.
The United States argues that the captive production provision, which requires USITC to, in certain circumstances, focus primarily on the merchant market, is consistent with the AD Agreement. The US emphasises that in all cases, including in the case at hand, USITC must render a determination with regard to injury to the industry as a whole, and cannot ignore the captive segment of the domestic industry.
7.176.
The United States argues that the US statute explicitly requires that the USITC examine the industry as a whole. The definition given in the US statute of the domestic industry is similar to that of Article 4.1 of the AD Agreement. The captive production provision does not require the exclusion of any other segments of the market. Nor does it require that emphasis be placed on some factors more than on others. The United States submits that the refined analysis suggested by the captive production provision is consistent with Articles 3.1 and 3.4 of the AD Agreement. It only operates as an analytical tool to reveal the impact of imports on a segment of the industry when this segment is a significant indicator of the state of the industry as a whole, and it thus improves the required overall industry analysis.
7.177.
The United States further argues that the captive production provision is consistent with Articles 3.4, 3.5 and 3.6 of the AD Agreement. Article 3.4 of the AD Agreement distinguishes between the effect of imports on sales and their effect on output, which is precisely the sort of distinction made by the captive production provision. The United States rejects Japan's argument that the captive production focus violates Articles 3.5 and 3.6 of the AD Agreement, which require consideration of the effect of dumped imports on domestic production as a whole, since the US statutory provisions also require such an overall industry analysis.
7.178.
In sum, the United States claims that the captive production provision is consistent with the AD Agreement and therefore does not violate Article XVI:4 of the Marrakesh Agreement either.
7.179.
Canada submits that failure to allow investigating authorities to differentiate between production that is internally transferred and production that is sold into the domestic market in competition with dumped imports, in appropriate circumstances, would deprive Article 3 of the Antidumping Agreement of its proper application and result in investigating authorities being unable to accurately determine whether a domestic industry had been injured, or threatened with injury.
7.180.
The EC agrees with the US that, where a significant portion of domestic output of the like product is for captive use, it is not inconsistent with the AD Agreement to focus the injury analysis on the "merchant" or "free market", since it is there that the immediate injurious effects of the dumped imports takes place. The EC considers that such a focus is even needed in order to avoid that the effects of dumped imports become obscured through the use of aggregate data.
7.181.
Chile considers that Articles 3 and 4 of the AD Agreement clearly support Japan's claim that an authority is to examine injury with regard to the industry as a whole.
7.182.
Brazil supports Japan's claim that an authority is required to examine the domestic industry as a whole, not merely part of it, when determining injury and causation. Brazil considers that consideration of only one segment of an industry is simply not permitted under the AD Agreement. Brazil is therefore of the view that the US captive production provision, which requires the authority to ignore the captive portion of the industry, is inconsistent with the AD Agreement.
7.183.
Korea asserts that Article 3.4 of the AD Agreement requires an analysis of "all relevant economic factors and indices bearing on the state of the domestic industry", i.e. the industry as a whole. It considers that an authority may not unduly emphasize a particular segment of the industry at the expense of the industry as a whole.

(ii) Finding

7.184.
Section 771(7)(c)(iv) of the Tariff Act of 1930 provides that, in a case in which domestic producers internally transfer significant production of the domestic like product for the production of a downstream article, and under certain specified circumstances, the USITC, in its injury analysis shall focus primarily on the merchant market for the domestic like product in determining market share and factors affecting financial performance.117 This provision is commonly referred to as the captive production provision since it distinguishes between the merchant market, the segment of the market consisting of commercial shipments on the open market, and the captive segment of the market - production which is internally consumed by the producer in the production of downstream products. In the investigation underlying this dispute, the USITC found that the domestic industry comprised US producers of hot-rolled carbon steel flat products. The USITC further found that these same producers used hot-rolled steel they had produced in the manufacture of downstream products such as cut to length, tubular, cold-rolled, and plated or galvanized steel. This "captive" consumption of hot-rolled steel by the domestic producers thereof was the subject of substantial argument by the parties to the investigation. In particular, its effect on the domestic industry underlies the dispute regarding the US captive production provision and its application in this case.
7.185.
Japan alleges that the US captive production provision violates Articles 3.1, 3.2, 3.4, 3.5, 3.6 and 4.1 of the AD Agreement concerning the determination of injury to the domestic industry. The thrust of Japan's argument is that the captive production provision's "primary focus" on the merchant market is inconsistent with the Agreement's requirement to determine injury to the "domestic industry" which is defined in Article 4 as domestic producers as a whole of the like products.
7.186.
In relevant part, Article 3 provides as follows:

"Determination of Injury9

3.1 A determination of injury for purposes of Article VI of GATT 1994 shall be based on positive evidence and involve an objective examination of both (a) the volume of the dumped imports and the effect of the dumped imports on prices in the domestic market for like products, and (b) the consequent impact of these imports on domestic producers of such products.

3.2 With regard to the volume of the dumped imports, the investigating authorities shall consider whether there has been a significant increase in dumped imports, either in absolute terms or relative to production or consumption in the importing Member. With regard to the effect of the dumped imports on prices, the investigating authorities shall consider whether there has been a significant price undercutting by the dumped imports as compared with the price of a like product of the importing Member, or whether the effect of such imports is otherwise to depress prices to a significant degree or prevent price increases, which otherwise would have occurred, to a significant degree. No one or several of these factors can necessarily give decisive guidance.

3.4 The examination of the impact of the dumped imports on the domestic industry concerned shall include an evaluation of all relevant economic factors and indices having a bearing on the state of the industry, including actual and potential decline in sales, profits, output, market share, productivity, return on investments, or utilization of capacity; factors affecting domestic prices; the magnitude of the margin of dumping; actual and potential negative effects on cash flow, inventories, employment, wages, growth, ability to raise capital or investments. This list is not exhaustive, nor can one or several of these factors necessarily give decisive guidance.

3.5 It must be demonstrated that the dumped imports are, through the effects of dumping, as set forth in paragraphs 2 and 4, causing injury within the meaning of this Agreement. The demonstration of a causal relationship between the dumped imports and the injury to the domestic industry shall be based on an examination of all relevant evidence before the authorities. The authorities shall also examine any known factors other than the dumped imports which at the same time are injuring the domestic industry, and the injuries caused by these other factors must not be attributed to the dumped imports. Factors which may be relevant in this respect include, inter alia, the volume and prices of imports not sold at dumping prices, contraction in demand or changes in the patterns of consumption, trade restrictive practices of and competition between the foreign and domestic producers, developments in technology and the export performance and productivity of the domestic industry.

3.6 The effect of the dumped imports shall be assessed in relation to the domestic production of the like product when available data permit the separate identification of that production on the basis of such criteria as the production process, producers' sales and profits. If such separate identification of that production is not possible, the effects of the dumped imports shall be assessed by the examination of the production of the narrowest group or range of products, which includes the like product, for which the necessary information can be provided.

9 Under this Agreement the term "injury" shall, unless otherwise specified, be taken to mean material injury to a domestic industry, threat of material injury to a domestic industry or material retardation of the establishment of such an industry and shall be interpreted in accordance with the provisions of this Article.

7.187.
In relevant part, Article 4.1 of the AD Agreement provides as follows:

"For the purposes of this Agreement, the term "domestic industry" shall be interpreted as referring to the domestic producers as a whole of the like products or to those of them whose collective output of the products constitutes a major proportion of the total domestic production of those products,"

7.188.
In addressing Japan's claim that the US statute is inconsistent with the AD Agreement on its face, we must resolve two questions. First, we must determine what is required by the AD Agreement, that is, whether the investigating authority is in all cases required to make a determination of injury to the domestic industry as a whole. If so, we must then consider whether the primary focus on the merchant market with respect to market share and financial performance set out in the "captive production" provision of the US statute is inconsistent, on its face, with this requirement?
7.189.
We consider that the definition of the domestic industry of Article 4.1 of the AD Agreement provides a clear answer to the first question. The domestic industry consists of the domestic producers as a whole of the like products, or of those producers whose collective output constitutes a major proportion of the total domestic production of those products. The terms "domestic industry" and domestic producers are also used interchangeably in Articles 3.1 and 3.4 of the Agreement. Article 3.1 of the AD Agreement provides that a determination of injury has to involve inter alia an objective examination of the "impact of these imports on domestic producers of such like products". Article 3.4 of the AD Agreement expands on this obligation and provides that the "examination of the impact of the dumped imports on the domestic industry concerned" shall include an evaluation of all relevant economic factors having a bearing on the state of the industry. Article 3.5 of the AD Agreement requires that a causal relationship be demonstrated "between the dumped imports and the injury to the domestic industry". We conclude that the requirement to make a determination of injury to the domestic industry read in light of the definition of the domestic industry of Article 4.1 of the AD Agreement, implies that the injury must be analysed with regard to domestic producers as a whole of the like product or to those whose collective output constitutes a major proportion of the total domestic production of those products.
7.190.
In our view, the AD Agreement thus clearly requires an investigating authority to make a final determination as to "injury" as defined in the Agreement to the industry as a whole. However, the Agreement does not prescribe a particular method of analysis. Specific circumstances might well call for specific attention to be given to various aspects of the industry's performance or to specific segments of the industry, as long as the end-result of this analysis is consistent with the Agreement's requirement to examine and evaluate all relevant factors having a bearing on the state of the industry and demonstrate a causal relationship between the dumped imports and the injury to the domestic industry.118
7.191.
We thus must examine whether the US "captive production" provision is on its face inconsistent with the established requirement of the Agreement to determine injury for the industry as a whole, as Japan is alleging. We note that the United States agrees with Japan that the AD Agreement requires a determination concerning injury with respect to the industry as a whole. According to the United States, the US statute is fully consistent with this obligation as it requires USITC to consider the industry as a whole. The United States asserts that the captive production provision, on its face, only affects some statutory factors required to be considered under 19 U.S.C. § 1677(7)(C)(iii) and does not affect the general requirement to determine injury for the domestic industry as a whole, which is set forth in 19 U.S.C. § 1677(4)(A) and which governs the entire determination of injury.
7.192.
The question before us is whether the captive production provision and the required "primary focus" on one segment of the market, the merchant market, with respect to market share and financial performance of the industry, is inconsistent with the obligations imposed on WTO Members in conducting an injury analysis for the purpose of an anti-dumping investigation. It is established GATT/WTO practice that the consistency of a law on its face may be challenged independently from any application thereof only in so far as the law is mandatory and not discretionary in nature. In other words, only if a law mandates WTO inconsistent action or prohibits WTO consistent action can the legislation be challenged on its face in a dispute settlement proceeding.119
7.193.
We do not doubt that the captive production of the US statute is mandatory in nature and may thus be challenged before a panel. The language of the provision ("shall focus primarily") makes it clear that USITC is required by statute to focus primarily on the merchant market in certain circumstances and under certain conditions. The question remains however whether the statute mandates action that is inconsistent with the United States' obligations under the AD Agreement.
7.194.
We recall that in relevant part, the captive production provision provides that "the Commission, in determining market share and the factors affecting financial performance set forth in clause (iii), shall focus primarily on the merchant market for the domestic like product". The key to answering the question posed lies in the ordinary meaning of the words "focus primarily". Japan argues that the use of the word "focus" skews the analysis to the merchant market at the expense of the rest of the domestic industry and the modifier "primarily" narrows the focus even more.
7.195.
The verb "to focus" is defined as "to concentrate"120 on something. "Primary" is defined as "of the first importance, chief"121. Literally, the captive production provision thus requires USITC to concentrate in chief on the merchant market when considering market share and financial performance of the industry. Such a specific direction to focus the analysis of certain factors with attention for a particular segment of the domestic market does not, in our view, necessarily imply that the overall injury analysis is not performed with respect to the industry as a whole. The statute does not require a general and exclusive focus on the merchant market when considering market share and industry performance, but only a "primary" focus.122 It certainly does not require a determination of injury based only on consideration of the merchant market.
7.196.
We believe that the context of the captive production provision confirms our view. The general obligation for injury determinations is set out in section 735(b)(1)(A) of the Tariff Act of 1930, as amended. That provision requires USITC to make a final determination of whether "an industry in the United States is materially injured or is threatened with material injury by reason of imports". Section 771(4)(A) of the Tariff Act of 1930, as amended, defines the relevant industry as "the producers as a whole of a domestic like product, or those producers whose collective output of a domestic like product constitutes a major proportion of the total domestic production of the product".123 US law specifically requires USITC, in making this determination, to consider "the impact of imports of such merchandise on domestic producers of domestic like products".124 In addition to the volume of imports and the effect of imports on prices, which the statute provides "shallbe considered",125 the statute further provides that "such other economic factors as are relevant to the determination regarding whether there is material injury by reason of dumped imports may be considered".126 The statute also sets out, in subsection C, entitled "Evaluation of Relevant Factors", specific elements to be considered when evaluating the volume of imports, the effects of imports on prices and their impact on the affected domestic industry, thus expanding on the more general obligation set out in subsection (B)(i) to consider the volume of imports and their effect on prices and impact on domestic producers.127 The captive production provision is set forth in section 771(7)(C)(iv) of the Tariff Act of 1930, and is thus an additional instruction with respect to the "Evaluation of Relevant Factors". We do not agree with Japan's position that the captive production provision constitutes an exception to the obligation to make a determination of material injury to the domestic industry as a whole. In our view, it is an instruction to "focus primarily" on certain "other economic factors as are relevant", and defines the circumstances in which such factors are relevant. Those factors are the market share and factors affecting financial performance in the merchant market, and the circumstances are the factual situation of the industry with regard to captive production. However, we can find no basis in the text of the US law to conclude that the captive production provision eliminates the general obligation on USITC to make a determination regarding material injury to the domestic industry. Nor does it, in our view, diminish the obligation to examine all relevant economic factors having a bearing on the state of the industry as a whole in making a final determination of injury caused by dumped imports. Finally, we note that US law explicitly provides that "The presence or absence of any factor which the [USITC] is required to evaluate under subparagraph (C) or (D) shall not necessarily give decisive guidance with respect to the determination by the [USITC] of material injury."128 The captive production provision, being set out in subparagraph (C) would thus fall within the scope of this instruction.
7.197.
Thus, we understand US law to require USITC to make a determination whether there is material injury to the domestic industry, and to provide guidance on the analysis to be undertaken in making that determination. The captive production provision is one of these latter sections, and thus defines an analytic step that must, in certain circumstances, be undertaken along the way to making the statutorily required determination of material injury to the domestic industry as a whole. It does not affect the nature of the determination of injury that must be made, only the analysis underlying that determination. While there is no guarantee that this analysis will result in a determination consistent with US obligations under the AD Agreement, it does not require any action inconsistent with those obligations.
7.198.
This is our reading of the statutory captive production provision. Equally important, this is our understanding of how the relevant US authorities have interpreted and applied this provision. We recall that, for the purposes of international law, domestic legislation is to be considered as a fact.129 In this respect, we believe it is of great importance that the Statement of Administrative Action notes that "the captive production provision does not require USITC to focus exclusively on the merchant market".130 The SAA is "an authoritative expression by the Administration concerning its views regarding the interpretation and application of the Uruguay Round agreements, both for purposes of US international obligations and domestic law … it is the expectation of Congress that future Administrations will observe and apply the interpretations and commitments set out in this Statement".131 In this case, the SAA confirms our conclusion based on the text and context of the captive production provision that the primary focus on the merchant market does not imply an exclusive focus on the merchant market in determining injury, and therefore does not mandate USITC to act in violation of the Agreement's established obligation to assess injury for the industry as a whole.132 As the Panel noted in United States - Sections 301 - 310 of the Trade Act of 1974, "The SAA thus contains the view of the Administration, submitted by the President to Congress and receiving its imprimatur, concerning both interpretation and application and containing commitments, to be followed also by future Administrations, on which domestic as well as international actors can rely."133 Moreover, this is the interpretation applied by the USITC itself, as set out in its decision in this investigation.134
7.199.
We therefore find that the captive production provision is not on its face inconsistent with Articles 3 and 4 of the AD Agreement. Having reached that conclusion, we also conclude that the United States did not act inconsistently with its obligations under Article XVI:4 of the Marrakesh Agreement and Article 18.4 of the AD Agreement in maintaining this statutory provision. We next turn to the question whether the USITC, in applying that provision in the investigation underlying this dispute, acted inconsistently with Articles 3 and 4 of the AD Agreement.

(b) Was USITC’s application of the captive production provision in this case consistent with Articles 3 and 4 of the AD Agreement?

(i) Arguments

7.200.
Japan claims that the application of the captive production provision in this case violated various provisions of Articles 3 and 4 of the AD Agreement. Japan argues that three Commissioners considered the captive production provision applicable and focused primarily on the merchant market in their analysis. Japan asserts that a fourth Commissioner de facto considered the merchant market data in parallel with data on the industry as a whole. Japan claims that this focus on the merchant market fundamentally altered the results of the investigation and distorted the Commissioners' judgement.135
7.201.
Japan claims that the specific determination in this case based on the captive production provision violates Articles 3 and 4 of the AD Agreement. Japan asserts that the USITC did not make an objective examination as required by Article 3.1 of the AD Agreement since it did not focus on domestic producers as a whole. The USITC’s focused analysis of injury also violated Articles 3.2, 3.4, 3.5 and 3.6 of the AD Agreement since it failed to examine all relevant evidence concerning the industry as a whole.
7.202.
Japan submits that although USITC mentioned both merchant market data and overall industry data, this in no way diminishes USITC's impermissible emphasis on merchant market data and it asserts that under a balanced analysis, USITC would have considered both the merchant and captive segments of the industry. This, according to Japan, would be the only way in which USITC could relate its segmented approach to the industry as a whole.
7.203.
The United States submits that the USITC analysis in this case was not inconsistent with the AD Agreement by virtue of the application of the captive production provision by three of the six Commissioners. The United States notes that all six Commissioners made affirmative determinations, five of current material injury and one of threat of material injury, while only three applied the captive production provision. This implies, according to the United States, that the application of the provision in this case did not change the outcome, which was in any case affirmative. Moreover, the United States maintains that USITC did not fail to make its determination on the basis of the domestic industry as a whole - indeed, information on the relevant economic factors was considered with respect to both the merchant market (the primary focus under the provision) and the industry as a whole. The United States asserts that USITC found that, both in the merchant market and with regard to the industry overall, consumption rose as did the volume of imports. The United States further notes that the declining financial trends that the USITC established in the merchant market also appeared in the overall industry analysis. Contrary to Japan’s claim that USITC did not "relate its merchant market findings to producers as a whole," the United States argues that the USITC determination shows how a primary focus on the merchant market for certain factors is consistent with such an analysis of the industry as a whole. The United States further submits that USITC compared the performance in the merchant market with overall performance of those domestic producers (integrated producers) most shielded from import competition and USITC found their operating income to be falling both from merchant market sales and overall.136 In sum, the United States claims that the captive production provision was irrelevant to the affirmative finding of USITC.

(ii) Finding

7.204.
The question before us is whether the USITC's determination of injury is consistent with the requirements of Articles 3 and 4 of the AD Agreement, in light of the focus on the merchant market by some Commissioners with respect to some factors examined, or whether that focus so taints the determination that we cannot conclude that an objective and unbiased investigating authority could make the determination the USITC made, on the basis of the facts on the record and in light of the explanations given. Under the applicable standard of review, we are not to overturn the evaluation of the administering authority if the establishment of the facts was proper and the evaluation unbiased and objective, even though we might have reached a different conclusion.
7.205.
The USITC report forms the basis for our examination of the consistency of the USITC's injury analysis with the requirements of the WTO AD Agreement concerning injury to the industry as a whole. We consider that the definition of the domestic industry used for the purposes of the investigation is a first important indicator of the scope of the investigation. The USITC report explains that

"[I]n defining the domestic industry, the Commission's general practice has been to include in the industry all of the domestic production of the like product, whether toll-produced, captively consumed or sold in the domestic merchant market. Based on our finding that the domestic like product consists of all hot-rolled steel, we define the corresponding domestic industry as all producers of hot-rolled steel in the United States, as we did in the preliminary determination"137. (emphasis added)

7.206.
USITC examined whether the domestic industry so defined was injured or threatened with material injury by reason of imports of hot-rolled steel from Japan. The report discusses various conditions of competition before entering into the examination of the volume of the imports, their effect on prices and the overall impact of the dumped imports on the domestic industry. USITC considered captive production to be one of the relevant conditions of competition. It stated that "the domestic industry captively consumes the majority, i.e. over 60 per cent of its production of the domestic like product in the manufacture of downstream articles".138 Based in part on this conclusion, Vice Chairman Miller and Commissioners Hillman and Koplan found that the captive production provision was applicable, and applied it in making their affirmative determination of material injury to the domestic injury producing hot-rolled steel caused by dumped imports. The three other Commissioners considered that not all of the statutory conditions for applying the captive production provision were fulfilled and thus did not find the provision applicable. Nonetheless, two of these Commissioners made an affirmative determination of material injury to the domestic injury producing hot-rolled steel caused by dumped imports, and the third made an affirmative determination of threat of material injury to that industry caused by dumped imports.
7.207.
The report contains data concerning both the industry as a whole and the merchant market in particular. USITC appears to have discussed these data independently from the application of the captive production provision, which in any case only requires a focus on the merchant market with regard to market share and factors affecting financial performance. As Chairman Bragg, Commissioner Crawford and Commissioner Askey, who did not apply the captive production provision, note in their Views Regarding The Captive Production Provision:

"even in circumstances in which the captive production provision does not apply, the Commission has the discretion to consider the significant volume of captive production as a condition of competition. Accordingly, we have examined data both for the domestic industry as whole and for merchant market operations for purposes of our determination".139

7.208.
We believe that the alleged distorting effect of the captive production provision should be examined in particular with regard to the USITC's analysis of market share and factors affecting financial performance, since these are the factors with respect to which a primary focus on the merchant market is required. The relevant section of the USITC report on market share discusses market share held by imports in the merchant market as well as in the overall US market and concludes that in both cases market share held by subject imports more than doubled from 1996 to 1997 and again from 1997 to 1998. In relevant part, USITC concluded as follows:

"In the merchant market, the share held by subject imports increased from 5.0 per cent of apparent US consumption as measured by volume sold in 1996, to 10.2 per cent in 1997, and then increased again to 21.0 per cent in 1998. For the industry as a whole, the share held by subject imports increased from 2.0 per cent of apparent US consumption, as measured by volume sold in 1996, to 4.2 per cent in 1997, and then increased again to 9.3 per cent in 1998".140

7.209.
The section in the report dealing with "impact of the subject imports on the domestic industry" discusses various economic factors having a bearing on the state of the industry. The report notes for example that capacity increased, while capacity utilization declined from 94.5 per cent in 1996 to 87.5 per cent in 1998. With regard to production and sales, USITC concludes that both merchant market data and overall industry data show a decline from 1997 to 1998. With regard to the domestic industry's financial performance indicators, USITC made the following analysis:

"From 1997 to 1998, as apparent consumption increased significantly, operating income declined by more than half. On merchant market sales, the ratio of operating income to net sales declined from 5.9 per cent in 1997 to 0.6 per cent in 1998 and overall, the ratio declined from 5.5 per cent in 1997 to 2.6 per cent in 1998".141

7.210.
The USITC conclusion on material injury by reason of imports of hot-rolled steel products from Japan was as follows:

"Accordingly, in light of the domestic industry's declining production, shipments, market share, prices, capacity utilization, and financial condition, in the face of increasing subject import volume and market share and declining subject import prices, we determine that the domestic industry producing hot-rolled steel is materially injured by reason of LTFV imports from Japan".142

7.211.
It is clear that USITC considered data for the domestic industry as a whole as well as merchant market data. On its face, the report sets out a complete and substantially motivated analysis of the state of the domestic industry as a whole. The report discusses data for the industry as a whole with regard to all relevant factors, including market share and financial perfomance of the industry, the two factors to which the captive production provision applies.
7.212.
We considered the data contained in the report in order to assess whether the evaluation of the USITC of the facts and data concerning market share and financial performance was that of an unbiased and objective investigating authority. We note that the USITC report includes two tables detailing the same sort of information for the industry as a whole and for the merchant market.143 These tables appear to support the conclusions of the report that the trends that are apparent in the merchant market also appear in the overall US market, albeit sometimes less pronounced.
7.213.
Japan asserts that the application of the captive production provision's primary focus for certain factors on the merchant market by three of the Commissioners so influenced their overall evaluation that it cannot be said with certainty what their conclusion would have been had they not applied the captive production provision. We do not consider it appropriate to engage in speculations about what could have or might have been. Upon careful examination, we consider that the USITC determined that the domestic industry producing hot-rolled steel as a whole, defined in the report as the domestic producers as a whole of hot-rolled steel in the United States, was materially injured, or threatened with material injury. We further consider that the determination was one that could properly be reached by an objective and unbiased investigating authority on the basis of the information before the USITC, and in light of the explanations given in its analysis. The mere fact that the analysis also included a discussion with regard to a certain segment of the industry most affected by the subject imports, in our view, does not at all necessarily imply that the analysis was faulty. Quite the contrary is true. As the Panel in Mexico – HFCS stated:

"There is certainly nothing in the AD Agreement which precludes a sectoral analysis of the industry and/or market. Indeed, in many cases, such an analysis can yield a better understanding of the effects of imports, and more thoroughly reasoned analysis and conclusion".144

Again, however, such an analysis does not excuse the investigating authority from making the determination required by the AD Agreement concerning injury to the domestic industry as a whole.

7.214.
We conclude that the analysis performed by USITC established injury with regard to the industry as a whole, in spite of, or regardless of, the application of the captive production provision by three of the Commissioners. We note that in any case all six commissioners made an affirmative injury or threat of injury determination whether they applied the captive production provision or not. This to us confirms our view that the application of the captive production provision did not undermine the examination of injury to the industry as a whole which is required under the AD Agreement.
7.215.
We therefore find that the USITC's analysis was consistent with the obligations of the United States under Articles 3.1, 3.4, 3.5, 3.6 and 4.1 of the AD Agreement in so far as it examined and determined injury to the domestic industry as a whole.

2. Alleged violations of Article 3 of the AD Agreement in the USITC's injury and causation analysis.

(a) Arguments

7.216.
Japan submits that the USITC injury and causation analysis is inconsistent with Articles 3.1, 3.4 and 3.5 of the AD Agreement since it focused on data for only two years of the normal three-year period of investigation and ignored or marginalized alternative causes of injury.
7.217.
First, Japan submits that the USITC eschewed its traditional three-year analysis and instead compared industry data for 1998 with those for 1997.145 Japan points to a recommendation of the AD Committee to argue that an investigating authority is to examine imports, prices and the industry performance over a three-year period of investigation, and asserts that this was the USITC's longstanding practice. Japan alleges that if applied in the hot-rolled steel case, a three-year analysis would have revealed that virtually all the major domestic industry performance indices improved between 1996 and 1998. According to Japan, the base year 1997, which Japan asserts was used by USITC in this investigation, happened to be the best year the industry had experienced in a decade and any comparison with this record-breaking year almost guaranteed an affirmative determination of injury. Japan asserts in particular that the USITC's analysis reveals an unexplained shift from a three-year to a two-year analysis for financial performance.146 In support of its argument, Japan refers to the views of Commissioner Askey who considered the entire three-year period of investigation in her analysis, and found no material injury to the domestic industry by reason of imports, but only threat of injury.
7.218.
Japan submits that by manipulating the period of investigation, USITC violated Article 3.1 by failing to base its material injury determination upon "positive evidence" and an "objective examination". Moreover, Japan argues that USITC violated Article 3.4 of the AD Agreement by failing to consider and to "make apparent" its consideration of the Article 3.4 factors for the first year of the period.147 Japan further alleges that the USITC determination was also inconsistent with Article 3.5 of the AD Agreement by failing to conduct a proper causation analysis that covered the full three years period and took into account the injury trends for this three year period.
7.219.
Japan also claims that USITC acted inconsistently with Article 3.5 of the AD Agreement by inadequately analyzing "other" causes of injury. Japan refers in particular to the strike at General Motors (the largest steel consumer in the US) in 1998, the increased capacity of and production by low-cost mini-mills, and faltering demand for pipe and tube due to collapsing oil prices. According to Japan, USITC did not consider the price effects of non-subject imports, as explicitly required by Article 3.5 of the AD Agreement. Japan asserts that the USITC mentions certain other relevant causal factors but fails to reconcile the facts and arguments presented by the parties.
7.220.
Japan further submits that USITC failed to isolate the injury caused by these alternative factors in order to ensure that such injury is not attributed to dumped imports. According to Japan, the United States – Wheat Gluten case made it clear that an anti-dumping investigating authority must ensure that when injury caused by alternative factors is subtracted, the remaining injury still rises to the level of "material injury".148
7.221.
The United States asserts that the USITC conducted an objective examination of data covering a period of investigation of three years and thoroughly examined possible known alternative causes of injury. USITC based its causal analysis on an evaluation of the changes in all relevant factors over a period of three years and the data used also covered three years.149 The United States disagrees with Japan's assertion that 1997 was an exceptionally good year for the US industry which would preclude any fair comparison with information for that year. According to the United States, many factors started to decline in 1996 and continued to decline in 1997 and 1998. Moreover, the United States argues, in 1998 productivity was higher and costs lower than in 1997, but nevertheless domestic industry performance indicators indicated a sharp decline in 1998. The United States submits that USITC reliance on recent trends is not unique but rather customary150 since the most recent data are in general more relevant and probative for the state of the industry. The United States asserts that it was appropriate for the USITC to place more weight on the most recent trends. The United States argues that the USITC's comparison of 1997 and 1998 data reflected its evaluation of the probative value of the 1996 – 1998 data in view of the changes in demand in the market that occurred since 1996.
7.222.
Second, the United States argues that, in accordance with Article 3.5 of the AD Agreement, the USITC examined all relevant factors and ensured that injury caused by other factors was not attributed to dumped imports. The United States asserts that Article 3.5 of the AD Agreement does not, however, require that a separate determination be made of the effects of the alternative causes. Nor, in the US view, is it required to quantify injury from other causes. According to the United States, the United States - Wheat Gluten Panel report is not relevant to this case since it was not concerned with the application of the AD Agreement but rather provided what the United States considers to be an incorrect interpretation of the Safeguards Agreement. The report of the Panel in United States – Imposition of Anti-Dumping Duties on Imports of Fresh and Chilled Atlantic Salmon for Norway ("United States -Atlantic Salmon"), which according to the United States should guide the Panel in its interpretation of the non-attribution provision of the AD Agreement, found that the authority is not required to demonstrate that dumped imports are the sole cause of material injury to an industry. Neither is an authority required to identify the extent of the injury caused by alternative factors.151 The USITC examined other known alternative causes and thus complied with the requirement of Article 3.5 of the AD Agreement. The United States submits that all known alternative causes suggested by Japan were extensively discussed by USITC and USITC did not attribute to the dumped imports the effects of other known factors.
7.223.
According to Brazil it must be demonstrated that imports in and of themselves were a cause of material injury to the industry, otherwise, the prohibition against attribution to dumped imports of injury caused by other factors in Article 3.5 of the AD Agreement would be meaningless. Brazil submits that USITC did not ensure against the attribution to imports of the effects of other factors as it made no efforts to isolate the effects of the other factors affecting the industry.
7.224.
Chile argues that USITC acted inconsistently with the AD Agreement and failed to conduct an objective examination by analyzing data from the most recent two years only and by not examining possible alternative causes of injury more thoroughly.

(b) Finding

7.225.
We will first consider Japan's claims concerning the allegedly WTO inconsistent focus of USITC on data for two years of the three-year period of investigation. We will then consider the USITC treatment of alternative causes of injury in light of the requirements of Article 3.5 of the AD Agreement.

(i) Did USITC properly discuss and evaluate data covering the whole period of investigation ?

7.226.
We note with regard to Japan's claim concerning USITC's alleged focus on two years of the three-year period of investigation that the AD Agreement does not specify the period of investigation and thus does not prescribe that the data used in the injury analysis have to cover three years.152 While the United States does not dispute that a three-year period of investigation should be considered for the purpose of making an injury determination, it asserts that the USITC in this case did consider a three-year period of investigation (1996 – 1998) and analysed all relevant economic factors having a bearing on the state of the industry on the basis of data covering this three-year period. Japan acknowledges that the USITC gathered data for the entire three-year period and that those data are mentioned in the USITC report in various tables and annexes. However, Japan argues, USITC failed to adequately factor this information into its determination and failed to compare the state of the industry at the end of the period of investigation in 1998 with the state of the industry in 1996.
7.227.
We note that throughout the USITC report there are various instances in which USITC does discuss trends in the data for the three-year period. For example, the USITC report discusses data from three years when examining the conditions of competition153 and the evolution in the volume of imports and the market share held by imports.154 The price effects of subject imports are also evaluated over the entire period of investigation 1996 – 1998.155 In the section of the report concerning impact of the subject imports on the domestic industry, the factors regarding capacity and capacity utilization are likewise discussed for the entire three-year period of investigation.156
7.228.
Japan's argument thus appears mainly based on the section of the USITC report that examines the impact of the subject imports on the domestic industry, and in particular, the data concerning financial performance of the industry.157 We note that the USITC report discusses production and sales as well as financial performance of the industry by comparing data for 1998 with data from 1997, without explicitly mentioning the 1996 values. In relevant part, the USITC report reads as follows:

"The domestic producers' production and shipments declined from 1997 to 1998, both on a merchant market and overall basis.98 The domestic industry's financial performance likewise deteriorated significantly. From 1997 to 1998, as apparent consumption increased significantly, operating income declined by more than half.99 On merchant market sales, the ratio of operating income to net sales declined from 5.9 per cent in 1997 to 0.6 per cent in 1998, and overall, the ratio declined from 5.5 per cent in 1997 to 2.6 per cent in 1998.100 101 This decline was due largely to declines in unit values of the industry's hot-rolled steel shipments and sales. As described above, unit values fell significantly in 1998 as subject imports increased in volume and market share".

98 CR & PR at Tables C-1 and C-2

99 CR & PR at Tables C-1 and C-2

100 CR & PR at Tables C-1 and C-2. In addition the domestic industry's productivity improved and COG's declined from 1997 to 1998. The domestic industry's productivity (measured in short tons per 1,000 hours worked) increased from 864.8 in 1996, to 905.3 in 1997 and to 938.7 in 1998. As discussed in our analysis of the price effects of the subject imports, the domestic industry's unit COG's declined from 1996 to 1998, but not by as much as the decline in the industry's unit values. CR & PR at Table C-1.

101 CR & PR at Table C-1. Aside from productivity, which increased during the investigation period, a number of the industry's other employment indicators declined somewhat during the period of investigation. CR & PR at Table III-5 (the number of workers declined from 33,965 in 1996, to 33,518 in 1997, to 32,885 in 1998; hours worked declined from 73,597 in 1996, to 71,634 in 1997, to 68,574 in 1998; wages paid were essentially flat from 1996 to 1998; hourly wages increased somewhat from $23.04 in 1996 to $24.13 in 1997, to $24.46 in 1998; unit production costs were $26.65 in 1996 and 1997 and declines somewhat to $26.06 in 1998). US producers' inventories were also relatively stable during the investigation period, both on an absolute basis and relative to production and shipments. CR & PR at Table III-4. Capital expenditures declined significantly from $1.7 billion in 1996, to $908 million in 1997, and to $715 million in 1998. CR & PR at Table VI-7. We also note that one firm filed for bankruptcy protection in September 1998 and another in February 1999. See CR & PR at Table III-1 nn.1 & 3; Petitioners' Prehearing brief at 51-52, 54; Respondents' Joint Prehearing Brief at 143. Both firms ***. See Questionnaire Responses of Geneva and Acme Metals, Inc." (footnotes in original)158