Lawyers, other representatives, expert(s), tribunal’s secretary

Report of the Panel

I. INTRODUCTION

A. COMPLAINT OF THE EUROPEAN COMMUNITIES

1.1.
On 10 November 2000, the European Communities requested consultations1 with the United States pursuant to Article 4 of the Understanding on Rules and Procedures Governing the Settlement of Disputes ("DSU"), Article XXII:1 of the General Agreement on Tariffs and Trade 1994 ("GATT 1994"), and Article 30 of the Agreement on Subsidies and Countervailing Measures ("SCM Agreement"), regarding countervailing duties ("CVDs") imposed by the United States on imports of certain corrosion-resistant carbon steel flat products originating in Germany, in particular, the sunset review of these CVDs.
1.2.
The European Communities and the United States held consultations on 8 December 2000, but failed to reach a mutually satisfactory solution.
1.3.
On 5 February 2001, the European Communities requested further consultations2 with the United States, regarding certain aspects of the procedure followed by the United States in sunset reviews, both as such and as applied in the sunset review in question, in particular, the evidentiary standards applied for the self-initiation of these reviews.
1.4.
The European Communities and the United States held further consultations on 21 March 2001, but failed to reach a mutually satisfactory solution.
1.5.
On 8 August 2001, the European Communities requested the establishment of a panel3 pursuant to Article 6 of the DSU, Article XXIII of GATT 1994, and Article 30 of the SCM Agreement.

B. ESTABLISHMENT AND COMPOSITION OF THE PANEL

1.6.
The Dispute Settlement Body ("DSB") established a panel on 10 September 2001, with standard terms of reference. The terms of reference of the Panel are:

To examine, in the light of the relevant provisions of the covered agreements cited by the European Communities in document WT/DS213/3, the matter referred to the DSB by the European Communities in that document, and to make such findings as will assist the DSB in making the recommendations or in giving the rulings provided for in those agreements.

1.7.
On 18 October 2001, the European Communities requested the Director-General to determine the composition of the Panel, pursuant to paragraph 7 of Article 8 of the DSU. This paragraph provides:

If there is no agreement on the panelists within 20 days after the date of the establishment of a panel, at the request of either party, the Director-General, in consultation with the Chairman of the DSB and the Chairman of the relevant Council or Committee, shall determine the composition of the panel by appointing the panelists whom the Director-General considers most appropriate in accordance with any relevant special or additional rules or procedures of the covered agreement or covered agreements which are at issue in the dispute, after consulting with the parties to the dispute. The Chairman of the DSB shall inform the Members of the composition of the panel thus formed no later than 10 days after the date the Chairman receives such a request.

1.8.
On 26 October 2001, the Director-General accordingly composed the panel as follows4:
1.9.
Chairman: Mr. Hugh McPhail

Members: Mr. Ronald W. Erdmann

Mr. Wieslaw Karsz

1.10.
Japan and Norway reserved their rights to participate in the panel proceedings as third parties.

C. PANEL PROCEEDINGS

1.11.
The Panel met with the parties on 29 January 2002, and on 19 March 2002. The Panel met with the third parties on 29 January 2002.
1.12.
The Panel submitted its interim report to the parties on 14 May 2002. Comments were received from the parties on the interim report on 23 May 2002, and on each other's comments on 30 May 2002 (See Section VII, infra). The Panel submitted its final report to the parties on 14 June 2002.

II. FACTUAL ASPECTS

2.1.
At issue in this dispute is the US law as such in respect of sunset reviews of CVDs, as well as its application in a sunset review carried out by the United States of a CVD order on imports of certain corrosion-resistant carbon steel flat products from Germany. The United States Department of Commerce ("DOC") determined that revocation of the order "would be likely to lead to continuation or recurrence of a countervailable subsidy".5 The DOC transmitted this determination to the United States International Trade Commission ("ITC"), along with a determination regarding the magnitude of the net countervailable subsidy likely to prevail in case of revocation of the order – 0.54 per cent in the review at issue. The ITC determined that revocation of the order "would likely lead to continuation or recurrence of material injury to an industry in the United States within a reasonably foreseeable time".6 Accordingly, the United States decided not to revoke the CVD order under review7 on imports of the product in question.
2.2.
The European Communities considers that the relevant US laws, regulations, administrative procedures, and statement of policy practices in respect of sunset reviews of CVDs, as well as their application in this instance, violates the SCM Agreement and the Marrakesh Agreement Establishing the World Trade Organization ("WTO Agreement"). The European Communities' claims therefore relate to the US sunset review system as such, as well as the specific sunset review determination by the DOC in respect of certain corrosion-resistant carbon steel flat products from Germany.
2.3.
The measures that the European Communities challenges as violating Articles 10, 11.9, 21, and 32.5 of the SCM Agreement, and XVI:4 of the Agreement establishing the World Trade Organization are:

1. the US CVD law in respect of sunset reviews: Section 751(c), as complemented by Section 752, of the Tariff Act of 1930 ("Tariff Act")8, as amended;

2. the accompanying Implementing Regulations: Procedures for Conducting Five-year ("Sunset") Reviews of Anti-Dumping and Countervailing Duty Orders, or "Sunset Regulations"9;

3. the accompanying statement of policy practices: Policies Regarding the Conduct of Five-year ("Sunset") Reviews of Anti-Dumping and Countervailing Duty Orders; Policy Bulletin, or "Sunset Policy Bulletin"10; and

4. their application in this instance, in the sunset review determination in respect of certain corrosion-resistant carbon steel flat products from Germany.11

III. PARTIES' REQUESTS FOR FINDINGS AND RECOMMENDATIONS

A. EUROPEAN COMMUNITIES

3.1.
The European Communities requests that the Panel find that the measures identified by the European Communities (See paragraph 2.3, supra):

1. infringe Article 21, paragraphs 3 and 1, as well as Article 10, of the SCM Agreement, by requiring that sunset reviews are automatically initiated for all existing CVD measures under the conditions specified therein;

2. violate Article 21, paragraphs 3 and 1, in conjunction with Articles 10 and 11, of the SCM Agreement, by applying expedited reviews, through automatic initiation and presumption of likelihood of continuation or recurrence;12

3. violate Article 21.3, in conjunction with Article 11, of the SCM Agreement, by requiring the automatic self-initiation of sunset reviews;

4. establish a standard of investigation for sunset reviews that violates the requirements of the SCM Agreement; and

5. violate Article 21.3, in conjunction with Article 21.1 and Article 11.9, of the SCM Agreement, by not requiring the application of the 1 per cent de minimis rule in sunset reviews and by enabling the continuation of CVDs for five more years in circumstances where there is no need to counter subsidisation which is likely to cause injury, and, because, in the present instance, the US authority continued a measure despite having found that the rate of subsidisation likely to prevail was less than 1 per cent.

3.2.
Accordingly, the European Communities requests the Panel to find the US CVD law, regulations, and statement of policy practices to be inconsistent with Article 32.5 of the SCM Agreement and, consequently, also inconsistent with Article XVI:4 of the WTO Agreement.

B. UNITED STATES

3.3.
The United States disputes the claims of the European Communities, and requests that the Panel find that:

6. the US procedure for the automatic self-initiation of sunset reviews by the DOC is not inconsistent with the SCM Agreement;

7. in not applying the 1 per cent de minimis standard of Article 11.9 of the SCM Agreement to sunset reviews, the United States has not acted inconsistently with its obligations under the SCM Agreement; and

8. the DOC sunset review determination in respect of certain corrosion-resistant carbon steel flat products from Germany is not inconsistent with US obligations under the SCM Agreement.

IV. REQUEST OF THE UNITED STATES FOR A PRELIMINARY RULING

A. REQUEST OF THE UNITED STATES

4.1.
The United States requests that the Panel make a preliminary ruling that the European Communities' claims with respect to the expedited sunset review procedure are not before the Panel because this procedure is not a measure within the Panel's terms of reference.
4.2.
The United States submits that the European Communities did not identify any measure or type of proceeding in consultations other than (i) the sunset review determination in carbon steel; (ii) the initiation of sunset reviews by the DOC; and (iii) the de minimis standard employed by the DOC in sunset reviews. Nor, argues the United States, did the European Communities identify the expedited sunset review procedure in its request for consultations or in its request for the establishment of a panel.
4.3.
The arguments of the United States in this regard are reflected below (See Section V, infra).

B. RESPONSE OF THE EUROPEAN COMMUNITIES

4.4.
The European Communities disagrees with the United States, and submits that the European Communities' claims regarding the United States' expedited sunset review procedure are within the Panel's terms of reference. The European Communities asserts that, given the reference in its request for establishment to Section 751(c) of the Tariff Act, which contains procedural rules on sunset reviews, including expedited sunset reviews, the issue of expedited reviews has also been raised by the European Communities and is therefore within the Panel's terms of reference.
4.5.
The arguments of the European Communities in this regard are reflected below (See Section V, infra).

V. ARGUMENTS OF THE PARTIES

5.1.
The arguments presented by the parties in their written submissions, oral statements, responses to questions, and comments on each other's responses to questions are reflected below.13

A. FIRST WRITTEN SUBMISSION OF THE EUROPEAN COMMUNITIES

1. Introduction

5.2.
This dispute concerns the compatibility of the United States (the "US") basic countervailing duty law (Tariff Act of 1930)14, its accompanying regulations (Sunset Regulations)15 and policy practices (Sunset Policy Bulletin)16, and their concrete application in the sunset review of countervailing duties on imports of certain corrosion-resistant carbon steel flat products ("corrosion resistant steel") from Germany. This decision violates the US WTO obligations in several ways.
5.3.
First, the Agreement on Subsidies and Countervailing Measures (the "SCM Agreement") provides for a presumption of termination of countervailing duty measures after 5 years. In the EC’s view, the automatic initiation of sunset reviews under US practice (without a requirement for any evidence to be produced) disregards this presumption for termination. As such, it contravenes Article 21.3 of the SCM Agreement.
5.4.
Second, the SCM Agreement requires that an investigating authority determines that the expiry of the duty would be likely to lead to recurrence of subsidisation. In not considering changes or terminations in subsidy programmes in the course of the sunset review, the US Department of Commerce ("DOC") effectively refused to conduct a proper investigation of the likelihood or recurrence of subsidisation. Therefore, the US breaches Article 21.3 of the SCM Agreement also under this instance.
5.5.
Third, the SCM Agreement provides that a level of subsidisation which is less than 1 per cent is de minimis. In this case,the investigation should be terminated without the imposition of measures. Under US law, this de minimis threshold does not apply in sunset reviews. Instead, the US practice is to apply a de minimis threshold of 0.5 per cent. In not applying the de minimis threshold provided for in the SCM Agreement, DOC effectively imposes and collects countervailing duties which are below the de minimis level. As a result, these provisions are in violation of Article 21.3, in conjunction with Articles 21.1 and 11.9, of the SCM Agreement.
5.6.
Finally, for the same reasons identified above, the US has failed to ensure that its laws, regulations and administrative procedures are in conformity with and Article 32.5 of the SCM Agreement and consequently with its WTO obligations under Article XVI:4 of the Marrakesh Agreement Establishing the World Trade Organization (the "WTO Agreement").

2. Factual Background

5.7.
Following the initiation of an investigation on 24 July 1992, DOC issued, on 9 July 1993, its final affirmative determination that exports of corrosion resistant steel from Germany were subsidised. DOC found that 5 subsidy programmes conferred a benefit to the German exporters. The subsidy level was determined at 0.60 per cent17. DOC found that two of the above programmes, the Capital Investment Grants ("CIG") and the Aid for Closure of Steel Operations (which account for 0.45 per cent of the total subsidy amount), were non-recurring grants because the recipient could not be expected to receive benefits on an ongoing basis. On the basis of the US methodology for non-recurring grants, the benefit was allocated over a 15-year period.
5.8.
Following the introduction of sunset provisions in the Uruguay Round, DOC was obliged to review all of its outstanding CVD orders by the end of 1999. On 1 September 1999, DOC automatically initiated a sunset review of the definitive countervailing duties on corrosion-resistant steel from Germany18. Following the publication of the notice, the domestic interested parties filed a notice of intent to participate. Subsequently, the foreign interested parties, i.e., the German producers, the Government of Germany and the European Commission filed substantive responses within 30 days after the publication of the Notice of Initiation.
5.9.
Following the investigation, on 20 March 2000, DOC issued the preliminary results of the full sunset review19. DOC preliminary concluded that revocation of the CVD order was likely to lead to recurrence of continuation or recurrence of a countervailable subsidy at a net subsidy rate of 0.54 per cent. The German producers argued that the benefit stream would become de minimis after the sunset review. As regards the CIG programme, DOC stated that some payments were given up to 1990 and, based on DOC's 15-year allocation period, would continue to confer a benefit to the German producers beyond the end of the sunset review.20 Only regarding the Structural Improvement Aids, DOC concluded that recurring benefits were granted up to 1986 and that no evidence was brought forward that benefits would occur after the sunset review.
5.10.
In calculating the amount of benefit, DOC simply took the subsidy rate from the original investigation in 1993 and deducted the amounts related to the Structural Improvement Aid programme and the Zonal Area programme. Therefore, the original countervailing duty rate was decreased by the amount of 0.06 per cent resulting in a net subsidy rate of 0.54 per cent being reported to ITC as the rate likely to prevail.
5.11.
On 2 August 2000, the DOC published the final results of its sunset review, in which the preliminary findings were confirmed.

3. Standard of Review

5.12.
By virtue of Article 30 of the SCM Agreement, the provisions of the DSU are applicable to the settlement of disputes under the SCM Agreement, "except as otherwise specifically provided [t]herein." In the absence of any specific standard of review provided for in the SCM Agreement on the issues raised in the present dispute, the EC considers that the standard of review set forth in Article 11 of the DSU is applicable in this case. The same has already been held by the panel and the Appellate Body reports in the United States – Leaded Bars case.21
5.13.
As regards the standard of review contained in Article 11 of the DSU, in the European Communities – Hormones case, the Appellate Body stated that "the applicable standard is neither de novo review as such, nor 'total deference', but rather the 'objective assessment of the facts' ".22 This does not mean that panels must simply accept the conclusions of the competent authority. To the contrary, a panel can assess whether the US competent authority's explanation for its determination under Article 21.3 is reasoned and adequate only if the panel critically examines that explanation, in depth, and in the light of all the facts before the panel. Panels must, therefore, review whether the competent authority's explanation fully addresses the nature, and, especially, the complexities, of the data, and responds to other plausible interpretations of that data.23
5.14.
The Panel is required to "make an objective assessment of the matter before it", including an objective assessment of the applicability of and conformity with the SCM Agreement of the US laws and regulations that are the subject matter of the present proceedings.24 For the purposes of interpreting the SCM Agreement, therefore, Article 11 of the DSU imposes upon panels a comprehensive obligation to make an "objective assessment of the matter" as a whole, an obligation which embraces all aspects of a panel's examination of the "matter", both factual and legal. Thus, panels should make an "objective assessment of the facts", of the "applicability" of the covered agreements, and of the "conformity" of the measure at stake with those covered agreements. Therefore, the text of Article 11 DSU and of the SCM Agreement clearly necessitate an active review or examination of all the pertinent facts.25
5.15.
The requirement to conduct an "objective assessment" of a claim has, in principle, two elements. First, a panel must review whether competent authority has evaluated all the relevant factors26 and, second, a panel must review whether the authority has provided a reasoned and adequate explanation of how the facts support their determination.27 Thus, the panel's objective assessment involves a formal aspect and a substantive aspect.28
5.16.
The relevant provisions of the DSU and SCM Agreement and the Appellate Body findings in the above cases require that the Panel in the present case is obliged to examine all the relevant facts and evidence presented to it by the parties to this dispute or obtained through the Panel’s own initiative. It follows that the Panel is obliged to assess whether the national authority provided a reasoned and adequate explanation of how the facts in the record or otherwise available to the US competent authority supported the determinations that were made by them in the contested measure. Obviously, this goes to the heart of the panel’s role in a dispute settlement case, as it relates to the panel’s discretionary authority to examine and weigh all factual evidence and to assess the applicability of and conformity with the relevant covered agreements of the measure under consideration.

4. Legal Arguments

(a) The standards of initiation of sunset review lead to a violation of Article 21.3 of the SCM Agreement

(i) Initiation

5.17.
Under US law, sunset reviews are automatically and initiated by DOC on its own initiative five years after the publication of the CVD order29. Under Article 21.3 SCM Agreement, WTO Members have an unequivocal obligation to terminate countervailing measures on a date no later than 5 years from their imposition, unless it is determined, following a review, that the expiry of the duties would be likely to lead to continuation or recurrence of subsidisation and injury. The rationale of this provision should be read in conjunction with Article 21.1 SCM Agreement which provides that a countervailing duty shall remain in force only as long as necessary and to the extent to counteract subsidisation which is causing injury. Therefore, Article 21.3 reinforces the requirement of Article 21.1 by creating a presumption that duties lapse after 5 years. These provisions provide that the continuation of measures beyond the five years is an exceptional situation (demonstrated by the term "unless"). This corresponds to the object and the purpose of Part V of the SCM Agreement to guarantee that trade defence measures are to be imposed and maintained only if they are necessary to offset injurious subsidisation.
5.18.
By requiring that sunset reviews are automatically initiated, the US effectively transforms this exception into a general rule, thus infringing Article 21, paragraphs 1 and 3, of the SCM Agreement.

(ii) Expedited reviews will "normally" result in the continuation of countervailing duties

5.19.
The automatic initiation under US law is also biased towards the continuation of countervailing duties and not towards the presumption of their termination. This is clear in the case of "expedited" sunset reviews. Under this procedure, if domestic interested parties file a "notice of intent to participate", the sunset procedure continues and interested foreign parties, i.e., exporters and the foreign government, are required to file substantive responses. If the foreign government, or exporters accounting for at least 50 per cent of exports, do not file substantive responses, DOC will shorten the timeframe for the review and normally presume that subsidisation continues.30 Therefore, the US procedure of expedited reviews, through automatic initiation and presumption of likelihood of continuation or recurrence, is clearly biased towards continuation of countervailing duties, in violation of the duty to terminate them set in Article 21 of the SCM Agreement.

(iii) The self-initiation of sunset reviews requires sufficient evidence of continuation or recurrence of subsidisation

5.20.
Article 21.3 SCM Agreement states that a sunset review can be initiated either on the initiative of the domestic producers or upon a duly substantiated request made by or on behalf of the industry. In the first case, the EC considers that Article 11 of the SCM Agreement on the initiation and conduct of the original subsidy investigation is of application in the case of sunset review as context to Article 21.3 of the same Agreement. It results from Article 11.2 that the initiation of an investigation to determine the existence of subsidisation should normally be based on the existence of a substantiated request made by or on behalf of the domestic industry. On the contrary, self-initiation by the domestic authority is the exception, which is justified only if domestic authority have the same level of sufficient evidence of subsidisation, injury and causal link (Article 11.6). In the same way, it follows that in order to initiate a sunset review on its own initiative, the domestic authority should be in possession of the same level of evidence that would be required in a "duly substantiated request" from the domestic industry. The automatic initiation requirement under US law converts the exception into a general rule, thus leading to the violation of Article 21.3 SCM Agreement.

(b) The domestic authority is under an obligation to "determine" the likelihood of continuation or recurrence of subsidisation

(i) The requirements of the SCM Agreement

5.21.
Article 21.3 SCM Agreement set out the requirements under which the investigating authority can deviate from the presumption of termination of CVD orders after five years. It requires that authorities determine that the expiry of the duty would be likely to lead to continuation or recurrence of subsidisation and injury. The Appellate Body in the case United States – Leaded Bars, considered that in a review under Article 21.2 the domestic authorities are required to make a finding of subsidisation during the period of review31. If this applies for Article 21.2 reviews, a fortiori a positive finding that all the conditions are fulfilled is necessary in the context of an Article 21.3 investigation. The parallel drawn between sunset reviews and original investigations is confirmed by the findings of the Panel in Brazil – Desiccated Coconut. There, called to judge of the applicability of the SCM Agreement to measures taken prior to the entry into force of the WTO Agreement, the Panel recognised that:

even measures maintained and imposed under the pre-WTO regime, and not subject to a review under Article 21.2 of the SCM Agreement, will ultimately be brought under WTO disciplines under this sunset provision32.

(ii) US law and practice

5.22.
US law, however, requires the DOC to consider (a) the net countervailable subsidy determined in the investigation or a review and (b) whether any changes in the subsidies occurred that may affect the countervailing duty. This is clearly stated in Section 752(b)(1) of the Tariff Act, URAA Statement of Administrative Action and Paragraph III.B.1 of the Sunset Policy Bulletin. In practice, DOC normally selects the CVD rate from the original investigation or a review, on the basis that it is the only rate that reflects the behaviour of governments and exporters without the discipline of an order in place. In the EC's view, this provision establishes a standard for sunset reviews that violates the SCM Agreement. This is clearly demonstrated in the case of corrosion resistant steel from Germany. Since there have not been any administrative reviews since the original investigation, the only available subsidisation rate is the original 1993 subsidy rate. DOC refused, in the course of the sunset review, to consider changes or terminations of subsidy programmes despite the concrete evidence submitted by the German exporters in the sunset review. DOC justified this decision by stating that no full investigation is conducted in a sunset review and decisions should be based on investigated and verified results.
5.23.
The EC considers that the DOC refusal to conduct any meaningful investigation in a sunset review is in violation of the requirements of Article 21.3 of the SCM Agreement. This practice not only contravenes the duty placed by Article 21.3 on the domestic authorities to "determine" the likelihood of continuation or recurrence of a subsidy, but it also shifts the burden of proving a change in circumstance on the foreign exporters and to a stage of the procedure, the administrative review, which is considered by the SCM Agreement as non-obligatory and which is in any event overtaken by the sunset review. The fact that a party did not request an administrative review cannot relieve the domestic authorities from their duties to "determine" the likelihood of continuation or recurrence of subsidisation. In any case, in the present case the German exporters did all they could to submit relevant information.

(iii) The US DOC refused to conduct a proper determination of the likelihood of continuation or recurrence of subsidization

5.24.
In the sunset review concerning corrosion resistant steel from Germany, the German producers stated that the major subsidy programme involved in the original investigation, the CIG programme, had been terminated and did not provide any further benefits. In particular they claimed that this programme, which provided non-recurring benefits, applied only to investments made prior to 1 January 1986.33 DOC had already determined in the original investigation that this programme applied in fact only to investments made prior to 1 January 1986.34 This was confirmed in the sunset review.35
5.25.
Despite this evidence, DOC rejected the claims of the German exporters. It stated that since no administrative reviews of these orders had been conducted, it was unable to determine whether any additional benefits under these programmes were received subsequent to the period of investigation. It should be stressed that DOC had both the non-confidential version of the questionnaire response and the calculation memorandum as part of the record of the original investigation and could have easily verified whether the benefits received by the German manufacturers of the subject merchandise under the CIG after 1 January 1986 were de minimis. The fact that DOC referred back to the original subsidy rate for CIG, violated Article 21.3 SCM Agreement because it did not take into consideration the termination of the CIG programme despite evidence on the record.

(iv) The CIG programme should have been considered as terminated

5.26.
In the present case, the German producers claimed that the CIG programme, had been terminated and did not provide any further benefits. In particular they claimed that this programme, which provided nonrecurring benefits, applied only to investments made prior to 1 January 198636. The calculation memorandum of the original investigation37 shows that Thyssen last received payments under the CIG programme in 1989, and that the amount received after 1985 was so small that it would have been automatically expensed in the year of receipt under the DOC 0.5 per cent rule38. Therefore, in accordance with their obligations under Article 21.3, the DOC should have taken this into account. DOC had already determined in the original investigation that this programme applied in fact only to investments made prior to 1 January 1986.39 DOC reaffirmed this finding in the final results of the sunset review, stating that: "We agree with Dillinger and the German Group's contention that … the CIG is applicable only to investments made prior to January 1, 1986."40
5.27.
Thus, the issue in the sunset review was not whether the programme continued to exist but whether any of the old benefit would continue after the end of the sunset review41. In their response, the US producers claimed that Preussag, the predecessor to Salzgitter AG, had received benefits under this programme as late as 199042. In their rebuttal of 15 October 1999, the German producers clarified that 99.4 per cent of the CIG paid to Preussag were received prior to the end of the 1985/86 fiscal year.43 Given that, using the 15-year allocation period applied by DOC, only 0.6 per cent of the original grants should have been considered by DOC as remaining to be countervailed after the end of the sunset review, the original subsidy rate for this programme of 0.39 per cent should have diminished to only 0.00234 per cent.44 Because DOC only calculates subsidy rates to the second decimal place, a rate of 0.00234 per cent would have been equivalent to zero.
5.28.
In its response, the US authorities did not directly discuss this evidence or respond in substance to the German producers’ evidence and arguments submitted in this regard during the sunset review. In fact, DOC rejected the German producers claim on the basis that the record of these sunset reviews is not sufficient for us to definitely conclude whether the benefits received by the German manufacturers of the subject merchandise under the CIG and/or IPA after January 1, 1985 were less than 0.5 percent of the corresponding beneficiary's annual net sales and, consequently, whether the benefits should be expensed in the year they were received. Furthermore, since no administrative reviews of these orders were conducted, we are unable to determine whether any additional benefits under these programmes were received subsequent to the period of investigation. As a result, as we did in our preliminary results, we determine that benefit streams from the CIG and IPA continue beyond the end of these sunset reviews and that, therefore, a countervailable subsidy from the CIG and IPA to manufacturers of subject merchandise would be likely if the orders were revoked45.
5.29.
On the issue of the insufficiency of the record of the sunset review, it has to be noted that, on 28 April 2000, the US producers filed a submission containing a copy of the public version of Preussag’s questionnaire response in the original investigation. Although the public version of the questionnaire response did not contain any payment amounts, it showed that of the fourteen CIG payments made to Preussag, only three were received after the end of the 1985/86 fiscal year.46 DOC, however, had both the non-confidential version of the questionnaire response and the calculation memorandum as part of the record of the original investigation and could have easily verified whether the benefits received by the German manufacturers of the subject merchandise under the CIG after 1 January 1986 were de minimis. The German producers were, therefore, not asking the DOC to rely upon speculative, unverified information, but simply to acknowledge the evidence that it had collected and verified itself in the original investigation.
5.30.
With regard to the argument that no administrative review had been conducted, DOC seems to argue that it was somehow prevented by law from considering the evidence because the German producers of corrosion-resistant flat products had not requested an administrative review.47 As seen above, Section 752 of the Act specifically requires the DOC to consider "the net countervailable subsidy determined in the investigation and subsequent reviews, and whether any change in the programme which gave rise to the net countervailable subsidy … has occurred that is likely to affect the net countervailable subsidy". There can be no doubt that the termination of a subsidy programme and the cessation of payments thereunder should be regarded as "change in the programme" that must be considered by the DOC. Moreover, while the relevant provisions of US law states that the DOC shall "normally" choose a net countervailable subsidy from the original investigation or an administrative review when determining the net countervailable subsidy that is likely to prevail if the order is revoked, this did not prohibit DOC from using in this case other rates or making adjustments to the rates found in an original investigation or an administrative review48.
5.31.
It follows from the above that the facts on record in this case require that the responsible US authorities should have adjusted the original subsidy rate to account for the termination of the CIG programme despite the fact that there were no administrative reviews of the countervailing duty order on corrosion-resistant flat products. In any case, the fact that the amount of the net countervailable subsidy likely to prevail was close to zero and, hence, de minimis, should have led the US authorities to terminate the duty in accordance with Article 21.3 of the SCM Agreement. The US has, therefore, erred in not accounting for the termination of the CIG programme, when it calculated the net countervailable subsidy that was likely to prevail if the countervailing duty order were revoked, and this is a violation of Article 21.3 of the SCM Agreement.

(v) Application of DOC's declining balance methodology

5.32.
DOC should also have reached the conclusion that the effects of the CIG programme had ended on the basis of its declining balance methodology. Under this methodology, DOC allocates non-recurring subsidies over time and in order to account for the "time value of money", it "front-loads" the amounts allocated to the early years of the allocation period. This inevitably results in a gradual year-by-year decline of the subsidy amount. In the present case, even without any consideration of individual programmes being made, it would have taken the subsidy amount even further below de minimis.

(vi) Treatment of evidence

5.33.
Finally, the EC considers that the US violated the standards of Article 21.3 SCM Agreement regarding the treatment of evidence. The German producers presented to DOC the reasons why the CIG programme should have been considered as terminated and as not providing any further benefits. Based upon this evidence, the German producers believed that DOC would find that subsidisation under the CIG programme would not be likely to continue or recur. However, in its preliminary results of 20 March 2000, DOC continued to assign a 0.39 per cent subsidy rate for this programme.
5.34.
The counsel for the German producers contacted DOC and stated that, if there were any doubts about how much was paid under the CIG programme after 1986, then DOC should make its calculation memoranda from the original investigation part of the record in the sunset review. On 13 April 2001, the German producers made a written request that DOC places the calculation memoranda on the record of the sunset review.49 DOC never took any action on the German producers’ request concerning the calculation memoranda and, in its final results of 27 July 2000, it claimed that the German producers’ request was untimely since it should have been submitted prior to 15 October 1999.50.
5.35.
The EC considers that such a short deadline violates the provisions of Article 21.3, in conjunction with Articles 12.1 and 21.4, of the SCM Agreement because it fails to afford the producers with an "ample opportunity" to present in writing all evidence which they consider relevant in respect of the sunset review.

(c) The de minimis rule

5.36.
The EC considers that the U.S. interpretation and application of the sunset provisions in Article 21.3, as regards the de minimis requirement, is inconsistent with the SCM Agreement.

(i) The US law and practice

5.37.
Under the Sunset Regulations51 and the Sunset Policy Bulletin,52 the US will treat as de minimis any countervailable subsidy rate that is less than 0.5 per cent. In initial investigations, the US applies a de minimis threshold of 1 per cent53. The US’ Statement of Administrative Action made it clear that the US considers that the de minims requirements of Articles 11.9 of the SCM Agreement only apply to initial investigations.

(ii) Interpretation of the relevant provisions of the SCM Agreement

5.38.
Article 21.3 of the SCM Agreement provides that any definitive countervailing duty shall be terminated not later than five years from its imposition, unless the authorities make a positive determination in a review that the expiry of the countervailing duty would be likely to lead to continuation or recurrence of subsidisation and injury. The continuation of the duty after a sunset review for five more years will be possible only if two conditions are met: a) a review is initiated; and b) the determination is made that subsidisation and injury would be likely to continue or recur if the duty were to expire. The Appellate Body in the United States – Leaded Bars case held that in order to establish the continued need for countervailing duties, an investigating authority will have to make a finding on subsidisation, i.e., whether or not the subsidy continues to exist. If there is no longer a subsidy, there would no longer be any need for a countervailing duty54.
5.39.
Therefore, in the context of a review under Article 21.2, the object and purpose of the requirement to demonstrate subsidisation and injury must be the same as when subsidisation and injury were determined in the original investigation. It follows then that if the de minimis rule is applied in original investigations, it must also remain applicable in reviews under Article 21.2. The EC submits that the above analysis in case of Article 21.2 reviews applies all the more so in the context of sunset reviews under Article 21.3. Indeed, there are compelling reasons to conclude that a countervailing duty, which is anyhow supposed to expire upon five years, will not be maintained without a clear demonstration of "subsidisation" and "injury". The Appellate Body further held that Article 21.2 sets out a review mechanism to ensure that Members comply with the rule of Article 21.1. The EC submits that the same reasoning applies a fortiori as regards the object and purpose of the sunset review mechanism laid down in Article 21.3 and renders sunset reviews similar to the original investigations, thus imposing on the authorities the burden of demonstrating the likelihood of a continuation or recurrence of subsidisation and injury (from a subsidy that is not de minimis), if the duty is not to expire.
5.40.
The term "subsidisation", in Articles 21.1, 21.2 and 21.3, can only interpreted in a systematically coherent manner, in context and in the light of the other relevant provisions of the SCM Agreement as a whole. One of these relevant provisions is clearly Article 11.9. This is because Article 11.9, in the context of original investigations, prevents the authorities from making a finding on subsidisation and injury on the basis of an amount of subsidy that is less than 1 per cent ad valorem, since it requires "immediate termination" of investigations in such circumstances. In addition, the US practice of applying a de minimis rule of (albeit at a WTO-inconsistent rate of 0.5 per cent) in, inter alia, sunset reviews provides an implicit confirmation of the need to apply a de minimis rule in the context of Article 21.3. Indeed, it is inherently contradictory for the US to argue that the rule of Article 11.9 of the SCM Agreement does not apply to sunset reviews, but nevertheless continue applying such a de minimis rule in practice. The US inconsistent and contradictory approach may seem to reduce the legal debate from one of principle to one about the actual amount or level of subsidy that is considered to be de minimis (i.e., in practice to 0.5 instead of 1 per cent). However, a closer examination of the US law and practice, as applied to the facts of the present case, would demonstrate that it could produce very perverse and inconsistent results.

(iii) Application of the de minimis rule to the corrosion-resistant steel products case

5.41.
In the case of corrosion resistant steel from Germany, the countervailing duty rate determined in the original investigation, which stemmed essentially from non-recurring subsidies allocated over time, was 0.59 per cent. This was only 0.09 per cent above the de minimis level of 0.50 per cent applied in the US before the entry into force of the WTO Agreements.55 Thus, if the original investigation had been conducted under the currently applicable rules for de minimis subsidies, the investigation should have been immediately terminated without the imposition of any countervailing duties, in accordance with Article 11.9 of the SCM Agreement. In the sunset review, which was conducted under WTO rules, the US, despite finding that the subsidy rate likely to prevail would be 0.53 per cent, nevertheless continued the measure, since its practice is to apply a 0.5 per cent de minimis threshold in sunsets. The EC submits that, for the reasons stated above, this threshold is not appropriate, and that since 0.53 per cent is below the 1 per cent de minimis level which should apply in sunset reviews, the US was in breach of Article 21.3 in continuing the measure

(d) The Obligation to Bring its Law Into Conformity with the WTO Agreements

5.42.
Article XVI:4 of the WTO Agreement and Article 32.5 of the SCM Agreement oblige Members to bring their domestic law into conformity with their obligations under the WTO Agreements. Indeed, as the panel report in Brazil – Desiccated Coconut has found,56 countervailing duty measures applied before 1995 would have to be brought under the disciplines of WTO over time pursuant to reviews under Article 21.2 or under the sunset provision of Article 21.3 of the SCM Agreement. The EC has demonstrated that the basic US law, the accompanying regulations and practices relating to sunset reviews of countervailing duties and their concrete application to the facts of the present case are inconsistent with a number of provisions of the SCM Agreement, i.e., Articles 21, paragraphs 1, 3 and 4, Article 10 and Article 11.9. For those reasons, the EC submits that the US failed to bring its domestic law in conformity with WTO obligations and, thereby, violating Article 32.5 of the SCM Agreement and Article XVI.4 of the WTO Agreement.

5. Conclusions

5.43.
For these reasons, the EC respectfully requests that the Panel finds that the US basic countervailing duty law (Tariff Act of 1930), its accompanying regulations (sunset regulations) and policy practices (sunset policy bulletin) as such, and their concrete application to imports of certain corrosion-resistant carbon steel flat products from Germany in the present case are inconsistent with Article 21 paragraphs 3, 1 and 4, Article 10 and Article 11.9 of the SCM Agreement. For the above reasons, the US countervailing duty law, regulations and practice should also be considered to be inconsistent with Article 32.5 of the SCM Agreement and, consequently, should be found to violate also Article XVI.4 of the WTO Agreement.

B. FIRST WRITTEN SUBMISSION OF THE UNITED STATES

1. Introduction

5.44.
The crux of the EC’s case that is properly before the Panel consists of allegations that the US countervailing duty law, as well as the sunset review determination in certain corrosion-resistant carbon steel flat products from Germany based upon that law, are inconsistent with the SCM Agreement because: (1) Commerce automatically initiates sunset reviews without first gathering evidence regarding the continuation or recurrence of subsidization; and (2) Commerce does not apply the SCM Agreement’s de minimis standard for countervailing duty investigations to sunset reviews. The EC argues that the Panel should read into Article 21.3 – the provision of the SCM Agreement that deals with sunset reviews – the requirements of Articles 11.6 and 11.9.
5.45.
The EC’s claims, however, run afoul of a basic principle of treaty interpretation. As stated by the Appellate Body in India Patent Protection, "the principles of treaty interpretation set out in Article 31 of the Vienna Convention... neither require nor condone the imputation into a treaty of words that are not there...." This is precisely what the EC is asking the Panel to do here; impute into Article 21.3 of the SCM Agreement "words that are not there".
5.46.
The EC tries to overcome this problem by repeatedly asserting that sunset reviews are "exceptions" to some other principle and, thus, must be interpreted in such a manner as to read into Article 21.3 "words that are not there." As discussed below, sunset reviews are not "exceptions" to something else, but instead are merely one part of an overall balance of rights and obligations negotiated during the Uruguay Round. However, even if one were to treat the provision on sunset reviews as an "exception" to something else, the EC’s arguments run afoul of another principle, articulated in EC Hormones, which is that "merely characterizing a treaty provision as an ‘exception’ does not by itself justify a ‘stricter’ or ‘narrower’ interpretation of that provision than would be warranted... by applying the normal rules of treaty interpretation."

2. Factual Background

(a) Sunset Reviews under US Law

5.47.
Commerce and the USITC jointly conduct sunset reviews pursuant to sections 751(c) and 752 of the Act. Pursuant to section 751(d)(2) of the Act, a countervailing duty order must be revoked after five years unless both Commerce and the USITC make respective affirmative determinations that subsidization and injury would be likely to continue or recur. Under the statute, Commerce automatically initiates a sunset review on its own initiative within five years of the date of publication of a countervailing duty order.
5.48.
Commerce has the responsibility of determining whether revocation of a countervailing duty order would be likely to lead to continuation or recurrence of subsidization, whereas the USITC has the responsibility of determining whether revocation of a countervailing duty order would be likely to lead to continuation or recurrence of injury. If Commerce’s determination is negative – i.e., if Commerce finds that there is no such likelihood – Commerce must revoke the order. If Commerce’s determination is affirmative, Commerce transmits its determination to the USITC, along with a determination regarding the magnitude of the net countervailable subsidy that is likely to prevail if the order is revoked. Under US law, the applicable de minimis standard in sunset reviews is the same as the standard in other types of reviews (e.g., duty assessment reviews) – 0.5 per cent.
5.49.
Commerce’s 1998 Sunset Regulations describe specifically the information to be provided by all interested parties in a sunset review and invite parties to submit, with the required information, "any other relevant information or argument that the party would like [Commerce] to consider." These regulations function as the standard questionnaire. The Sunset Regulations also provide that substantive responses to a notice of initiation are due 30 days after the date of publication in the Federal Register of the notice of initiation; rebuttals are due five days later. The regulations provide that Commerce normally will not accept or consider any additional information from a party after the time for filing rebuttals has expired.
5.50.
On 14 May 1998, Commerce published the schedule for initiation of sunset reviews of pre-1995 anti-dumping and countervailing duty orders, indicating that the sunset review of the countervailing duty order on corrosion-resistant steel would be initiated in September 1999.
5.51.
Thus, with the applicable information requirements, deadlines, and initiation schedule published in the Federal Register by May 1998, the EC and German producers had over 15 months to prepare for the sunset review of the countervailing duty order on certain corrosion-resistant carbon steel flat products from Germany.

(b) Certain Corrosion-Resistant Carbon Steel Flat Products from Germany

5.52.
On 9 July 1993, Commerce published its final affirmative countervailing duty determination on certain corrosion-resistant carbon steel flat products from Germany. Commerce calculated a country-wide total ad valorem countervailing duty rate of 0.59 per cent, based on the German producers receipt of countervailable benefits under the following five programmes: (1) Capital Investment Grants (hereinafter "CIG"); (2) Structural Improvement Aids; (3) Special Subsidies for Companies in the Zonal Border Area; (4) Aid for Closure of Steel Operations; and (5) ECSC Redeployment Aid Under Article 56(2)(b). On 9 August 1993, the USITC notified Commerce of its final affirmative injury determination. On 17 August 1993, Commerce amended its final determination to correct a ministerial error, which increased the ad valorem rate to 0.60 per cent, and issued the countervailing duty order.
5.53.
On 26 August 1999, Commerce notified representatives of the EC, the German Government, and German producers, by mail, that the sunset review of the countervailing duty order on certain corrosion-resistant carbon steel flat products from Germany would be initiated on or about 1 September 1999 (consistent with the Sunset Initiation Schedule). On 1 September 1999, Commerce published its notice of initiation. In both its letters to German producers and the published initiation notice, Commerce highlighted the 30-day deadline for filing substantive responses, as well as the applicable information requirements.
5.54.
By 4 October 1999, the EC, the German Government, German producers, and domestic interested parties filed their substantive responses. The EC, the German Government, the German producers, and the domestic interested parties filed rebuttal comments on 15 October 1999. On 20 October 1999, Commerce determined to conduct a full sunset review based on its receipt of complete substantive responses from the EC, the German Government, and German producers accounting for a significant portion of German exports to the United States.
5.55.
On 27 March 2000, Commerce published its preliminary sunset determination finding likelihood of continuation or recurrence of subsidization. Based on its finding that benefit streams from non-recurring grants under the CIG programme would continue beyond the five-year mark and that the Aid for Closure of Steel Operations and ECSC programmes continue to exist, Commerce determined there was likelihood of continuation or recurrence of subsidization.
5.56.
As required under US law, Commerce also determined the net countervailable subsidy likely to prevail if the order were revoked. As a general matter, and starting with the total ad valorem rate determined in the original investigation, Commerce considers whether, since the investigation, it has found subsidy programmes to be terminated and/or new programmes to be countervailable. Based on findings, which normally are made in the context of administrative reviews under section 751(a) of the Act, Commerce may adjust the rate determined in the original investigation to take these subsequent findings into account. Although no administrative reviews of the order on certain corrosion-resistant carbon steel flat products from Germany were ever conducted, Commerce agreed with the EC and the German producers that the Structural Improvement Aids and Special Subsidies for Companies in the Zonal Border Area programmes had been terminated with no continuing benefits and adjusted the net countervailable subsidy rate accordingly. Because no administrative reviews had been conducted, Commerce did not consider the domestic interested parties’ allegations concerning additional countervailable subsidies. For the same reason, Commerce did not recalculate the subsidy rates determined in the original investigation. Based on this analysis, Commerce determined a net countervailable subsidy rate of 0.54 per cent.
5.57.
In its final determination, published 2 August 2000, Commerce did not change the basis for its likelihood determination or its determination concerning the net countervailable subsidy likely to prevail. On 1 December 2000, the USITC published its determination that revocation of the countervailing duty order would be likely to lead to continuation or recurrence of injury. On 15 December 2000, the United States published notice of the continuation of the countervailing duty order on certain corrosion-resistant carbon steel flat products from Germany based on the decisions by Commerce and the USITC finding likelihood of continuation or recurrence of subsidization and injury.

3. Standard of Review

5.58.
With respect to disputes involving a determination made by a domestic authority based upon an administrative record, the Appellate Body, in US Cotton Yarn, recently summarized the standard of review under DSU Article 11. The United States does not disagree with this standard. The EC erroneously argues, however, that the Panel cannot "disregard or refuse to consider facts and evidence submitted to it" by the parties to the dispute. The United States disagrees with the EC’s implication that a panel has unfettered discretion to consider any evidence in deciding the issues before it and, as discussed below, the Panel should decline to consider the document submitted by the EC.

4. Substantive Argument

(a) Automatic Self-Initiation of Sunset Reviews is Consistent with the SCM Agreement

5.59.
Article 21.3 authorizes authorities to initiate a sunset review "on their own initiative or upon a duly substantiated request made by or on behalf of the domestic industry" (emphasis added). This disjunctive language is unambiguous, and, under the customary rules of interpretation, must be read according to its ordinary meaning, which is that a Member may either self-initiate a sunset review or initiate a sunset review in response to a duly substantiated request. The right of an investigating authority to initiate a sunset review on its own initiative is unqualified and the Panel may not "diminish" this right. Despite the plain language of Article 21.3, the EC argues that the Article 11.6 requirements for self-initiation of an investigation are applicable to self-initiation of sunset reviews. The obvious flaw in the EC’s argument is that there is no reference to the Article 11.6 requirements in the text of Article 21.3 or vice versa. Furthermore, the SCM Agreement itself distinguishes between the investigatory phase and the review phase of a countervailing duty proceeding, e.g., Article 11 deals with investigations, while Article 21 deals with reviews.
5.60.
The EC’s arguments, therefore, find no support under customary rules of treaty interpretation. Article 21.3 explicitly provides for initiation of sunset reviews on an authority’s own initiative. Furthermore, nothing in the text of Article 21.3, or Article 11.6, imposes any evidentiary requirements on authorities who initiate sunset reviews on their own initiative. It is impossible to violate an obligation that does not exist. Therefore, the United States’ automatic initiation of sunset reviews is not inconsistent with the SCM Agreement.

(b) There is no de minimis Standard for Sunset Reviews

5.61.
The focus of a sunset review under Article 21.3 is future behaviour, i.e., the likelihood of continuation or recurrence of subsidization – not whether or to what extent subsidization currently exists. The analysis is perforce predictive. Under these circumstances, mathematical certainty or precision as to the exact amount of likely future subsidization is not necessarily practicable and certainly not required.
5.62.
Under Article 11.9, Members must apply a one per cent de minimis standard in countervailing duty investigations. The EC erroneously argues that the Article 11. 9 de minimis standard is applicable in sunset reviews under Article 21.3. There is no textual or contextual support for the EC’s claim.
5.63.
In Korea DRAMs, Korea argued that the de minimis standard in Article 5.8 of the AD Agreement applied to reviews as well as to investigations. Article 5.8 of the AD Agreement is the parallel provision to Article 11.9 of the SCM Agreement. The panel rejected Korea’s arguments, finding that "the term ‘investigation’ [used in the context of Article 5.8] means the investigative phase leading up to the final determination of the investigating authority." Thus, the Korea DRAMs panel found no textual or contextual support for Korea’s claim that the de minimis standard applied beyond the investigative phase.
5.64.
The EC’s argument is not only devoid of support in the text of the SCM Agreement, it also fails to mention, much less reconcile, its position with relevant language in the text. Specifically, note 52 of Article 21.3 provides that "a finding in the most recent assessment proceeding that no duty is to be levied shall not by itself require the authorities to terminate the definitive duty". Thus, the current level of subsidization is not decisive as to whether subsidization is likely to recur. The EC’s claim that a de minimis standard is required in the context of Article 21.3 sunset reviews would render note 52 meaningless.
5.65.
The EC would also have the panel read into the use of the word "subsidization" in Article 21 an implicit reference to Article 11.9 because authorities must terminate an investigation if the amount of the subsidy is de minimis. However, nothing in the word "subsidization", as defined in the SCM Agreement implies anything about a de minimis standard. The term "subsidization" simply means the existence of a subsidy as defined in Article 1 of the SCM; Article 1 contains no de minimis standard.
5.66.
In sum, giving the text of the Agreement its ordinary meaning, the only conclusion one can reach is that there is no obligation to apply the Article 11.9 de minimis standard in an Article 21.3 sunset review. The EC’s arguments concerning the object and purpose of Article 21.3 fail to overcome the obvious lack of any textual support for their claim.
5.67.
The EC argues that a sunset review is equivalent to an investigation because it could result in re-"imposition" of an order and, as such, the same de minimis standard is applicable in a sunset review. This argument completely ignores the fundamental difference between investigations, in which a de minimis standard is required under Article 11.9, and sunset reviews. In the context of Article 11.9, the function of the de minimis test is to determine whether foreign government subsidies warrant the imposition of a countervailing duty order in the first instance. For example, in an investigation, if the investigating authority found that a government programme had provided recurring subsidies at a rate of more than one per cent, imposition of a countervailing duty would be warranted if the subsidized imports were found to cause injury.
5.68.
In contrast, the focus of the sunset review is the future. The mere continued existence of this same programme could warrant maintaining the duty beyond the five-year point, even if the amount of the subsidy was currently zero, as stated in footnote 52, because subsidization may be likely to recur absent the discipline of the duty. This distinction between the object and purpose of an investigation and the object and purpose of a sunset review supports the conclusion that, absent an express reference to the contrary, there is no basis to assume or infer an intent that the de minimis standard for investigations applies in sunset reviews.
5.69.
In an attempt to bolster its non-existent textual argument, the EC cites the fact that the United States applies a de minimis standard in sunset reviews as "confirmation" of the requirement to apply a de minimis rule in the context of Article 21.3 sunset reviews. In addition, the EC argues that, given the provisions of Article 32.4, it had a "reasonable and legitimate expectation" that the United States would terminate the duty. The EC is wrong on both accounts.
5.70.
The United States’ de minimis "practice" is legally irrelevant. As demonstrated above, there is no de minimis standard in sunset reviews. Thus, Members are free to determine what, if any, de minimis standard they will apply. Furthermore, while Article 31.3(b) of the Vienna Convention permits consideration of "any subsequent practice in the application of the treaty which establishes the agreement of the parties regarding its interpretation," policy decisions made by one Member for purposes of its domestic legislation do not constitute "subsequent practice" within the meaning of Article 31.3(b).
5.71.
In addition, the EC’s only legitimate expectations with respect to Articles 32.4 and 21.3 are those reflected in the Agreement itself. As the Appellate Body, in India Patent Protection, stated: "[P]rinciples of interpretation neither require nor condone the imputation into a treaty of words that are not there or the importation into a treaty of concepts that were not intended."
5.72.
In sum, applying customary rules of treaty interpretation, the Panel should find that there is no de minimis standard for sunset reviews in the SCM Agreement and, therefore, the United States’ application of a 0.5 per cent de minimis standard in sunset reviews does not constitute a violation of its obligations under the SCM Agreement.

(c) Commerce Properly Determined that the Expiry of the Countervailing Duty Order would be likely to lead to Continuation or Recurrence of Subsidization Based upon an Appropriately Conducted review of all Relevant and Properly Submitted Facts

5.73.
As demonstrated above, the United States’ automatic self-initiation of sunset reviews and its application of a particular de minimis standard do not breach any provision of the SCM Agreement. The remaining claims raised by the EC concern Commerce’s findings and procedural actions in this case. An "objective assessment" of Commerce’s findings and actions, pursuant to Article 11 of the DSU, would focus on the consistency of the sunset review with the requirements of Articles 21.3 and 12.
5.74.
As a starting point for making its likelihood determination in the sunset review, Commerce considered the countervailable subsidies and programmes used, and the amount of the subsidy determined, in the original investigation. As explained in Commerce’s preliminary sunset determination, the rationale for this approach is that the findings in the original investigation provide the only evidence reflecting the behaviour of the respondents without the discipline of countervailing measures in place. This approach makes sense given that, in a sunset review under the Article 21.3, an authority is considering whether, without the discipline of the duty, subsidization would likely continue or recur, i.e., what would happen without the discipline of the order.
5.75.
In the original investigation, Commerce determined that German producers of corrosion-resistant steel benefitted from five different subsidy programmes. In the sunset review, Commerce found that the benefit streams from non-recurring grants under the CIG programme will continue beyond the five-year mark; the Structural Improvement Aids and Special Subsidies for Companies in the Zonal Border Area programmes had been terminated; and the Aid for Closure of Steel Operations and ECSC Redeployment Aid Under Article 56(2)(b) programmes continue to exist. Significantly, the EC has not disputed or disproved these findings. As an initial matter, therefore, it was reasonable for Commerce to find likelihood given the continued existence and availability of countervailable subsidy programmes previously found to have been used by German producers of corrosion-resistant steel and the continuation of benefit streams from grants under the CIG programme.
5.76.
Although the EC essentially concedes the continued existence of some benefits from the CIG programme, it claims that, based on routine amortization, Commerce should have considered the programme terminated without residual benefits to the German producers. With respect to non-recurring benefits (such as the benefits from the CIG programme), Commerce uses a "declining balance" formula to determine the amount of subsidization to be allocated in each period. An ad valorem subsidy rate, for a particular period, is derived by dividing a numerator – the subsidy benefit properly attributable to the subject merchandise – by a denominator – the value of the sales of the merchandise at issue (in the case of a domestic subsidy). Without knowing the sales volume, the ad valorem subsidy rate for any period cannot be determined despite the use of a "declining balance" methodology generally. The EC’s claim, therefore, fails as a factual matter because there is no basis for its assumption that the sales volumes will remain constant.
5.77.
The EC’s amortization arguments, furthermore, are based in part on a calculation memorandum from the original countervailing duty investigation that is not part of the record considered in the sunset review (Exhibit EC-20). The request to submit this business confidential document was untimely submitted and Commerce properly declined to consider it. The German producers’ request to submit this document also implicated Commerce’s rules concerning treatment of confidential information ("business proprietary information" or "BPI" in US parlance). Pursuant to US law, release of that information is not permitted without the consent of the person that submitted it. Commerce could not ignore previous requests for confidential treatment and automatically place this information from the original 1993 investigation on the record of the sunset review. Further, other parties without prior access to the document would have been prejudiced by its untimely inclusion on the record.
5.78.
Under these circumstances, Commerce did not consider it practicable or appropriate to consider the document. Commerce’s decision to enforce procedural rules governing deadlines for submission of evidence and the release of confidential business information was proper and consistent with Article 12. (The evidentiary and procedural requirements of Article 12 are applicable to sunset reviews by virtue of Article 12.4.) As such, the Panel should find that Commerce appropriately declined to consider the information and that it is not this Panel’s role to consider evidence which could have been timely presented to the decision maker but was not. Furthermore, even if the Panel should consider the document, it does not prove the EC’s arguments. The calculation memorandum only provides the absolute subsidy amounts (i.e., the numerator) – it does not shed any light on the value of the sales of the merchandise at issue (i.e., the denominator). As demonstrated above, without a denominator, there is no way to calculate the ad valorem subsidy rate.
5.79.
Consistent with the Appellate Body’s ruling in UK Lead Bar and Article 21.3, Commerce properly considered that the existing benefit streams from the CIG programmes constituted evidence of the "continuation" of subsidization. Furthermore, the continued existence of other programmes previously found to be countervailable is not in dispute. As a result, Commerce’s finding of likelihood of continuation or recurrence of subsidization is consistent with its obligations under Articles 21.3 and 12 of the SCM Agreement. In addition, as demonstrated below, Commerce’s evidentiary and procedural actions also were consistent with its obligations under Article 12.
5.80.
Article 12.1 requires domestic authorities to give interested Members and parties an ample opportunity to present in writing all evidence which they consider relevant to the proceeding. The facts do not support the EC’s claims that Commerce failed to do so.
5.81.
The Sunset Regulations describe specifically the information required to be provided by all interested parties in a sunset review, i.e., they constitute the standard questionnaire. In addition, the Sunset Regulations specifically invite parties to submit, with the required information, "any other relevant information or arguments that the party would like [Commerce] to consider". Consistent with Article 12.1.1, Commerce’s regulations also provide 30 days for parties to submit the required information and provide for extensions of time to meet this deadline. The EC and the German producers were on notice of the information requirements and options, as well as the applicable deadlines and extension options, over 15 months ahead of the scheduled date for initiation of the sunset review.
5.82.
Yet over six months after the deadline for responding to the sunset questionnaire and submitting optional information, the German producers attempted to place new factual information on the record. The EC asserts that Commerce’s rejection of these untimely submissions was contrary to their "right" under the SCM Agreement to have an "ample opportunity to present in writing all evidence which they consider relevant in respect of the sunset review". As a factual matter, however, the German producers and the German Government had ample time to submit factual information in the sunset review. Furthermore, as a legal matter, Commerce’s filing deadlines and its decision not to accept late-filed information fully comport with its obligations under Article 12.
5.83.
Specifically, the German producers had 30 days to respond to the questionnaire. In addition, Commerce’s rejection of the German producers late-filed information was reasonable under the circumstances of this case, i.e., the German producers attempted to file new factual information over six months after Commerce’s deadline. Although an authority "should" grant extensions "whenever practicable", nothing required Commerce to find that it was practicable to accept and consider documents filed six months late, particularly given the fact that the EC and the German producers had over 15 months to gather any data they considered appropriate and to prepare their submission of required and optional information. Finally, the EC’s argument, that Commerce arbitrarily applied its regulation to submissions from the various parties, ignores relevant factual distinctions between the submissions from the German producers rejected by Commerce and those submissions from the US producers and the German Government that were accepted by Commerce. In particular, the accepted submissions involved public documents containing no new factual information.
5.84.
In sum, the Panel should dismiss the EC’s claims with respect to treatment of evidence. Commerce followed reasonable, appropriate procedures that fully comply with the evidentiary and procedural requirements of Articles 21 and 12.

(d) The Panel should make a Preliminary Ruling that the EC’s Claims Regarding the Expedited Sunset Review Procedure are not within the Panel’s Terms of Reference

5.85.
The United States requests that the Panel make a preliminary ruling that the EC’s claims regarding the US expedited sunset review procedure are not properly before the Panel, because this procedure is not a measure within the Panel’s terms of reference. Not until its first written submission to the Panel did the EC ever give any indication that it was complaining about this procedure.
5.86.
In its initial request for consultations, the EC identified Commerce’s determination in the full sunset review of the countervailing duty order on certain corrosion-resistant carbon steel flat products from Germany as the challenged measure, alleging that Commerce’s determination is inconsistent with Articles 10, 11.9 and 21 of the SCM Agreement. In its second request for consultations, the EC identified Commerce’s procedures for initiation of sunset reviews, both as applied by Commerce in the sunset determination in question and in general, as an additional challenged measure, alleging that such initiation procedures are inconsistent with Articles 21.1, 21.3 and 32.5 of the SCM Agreement and Article XIV:4 of the Marrakesh Agreement. The EC did not allege that any other sunset determination or procedure violated US WTO obligations. Furthermore, the parties did not discuss the expedited sunset review procedure at either the 8 December 2000, or 21 March 2001 consultations. Finally, there is no mention of the expedited sunset review procedure in the EC’s request for the establishment of a panel.
5.87.
Articles 4.7 and 6.2 of the DSU, therefore, preclude the EC’s claims with respect to the expedited sunset review procedure, because the EC never identified this procedure as a measure in its consultation requests, in the consultations themselves, or in its panel request. It is well-established that a complaining party cannot add new measures after a panel’s terms of reference have been established.

5. Conclusion

5.88.
For the reasons set out in this submission, the United States respectfully requests that the Panel make the following findings:

(1) The US procedure for the automatic self-initiation of sunset reviews by Commerce is not inconsistent with the SCM Agreement;

(2) In not applying the 1 per cent de minimis standard of Article 11.9 of the SCM Agreement to sunset reviews, the United States has not acted inconsistently with its obligations under the SCM Agreement;

(3) The Commerce sunset review determination in certain corrosion-resistant carbon steel flat products from Germany is not inconsistent with United States obligations under the SCM Agreement.

5.89.
In addition, the United States respectfully requests that the Panel make a preliminary ruling that the EC’s claims with respect to the expedited sunset review procedure are not within the Panel’s terms of reference.

C. ORAL STATEMENT OF THE EUROPEAN COMMUNITIES AT THE FIRST MEETING OF THE PANEL

1. Introduction

5.90.
The complaint concerns the decision of the US authorities to continue the countervailing duties on imports of corrosion-resistant carbon steel flat products from Germany (the product). The decision of the US authorities in this case is a result that is mandated by the basic US statute, the implementing legislation and administrative practice on sunset reviews of countervailing duty orders. For these reasons, the present proceeding covers certain aspects of the US basic sunset review legislation, their procedures and their implementation which led to the unlawful continuation of the countervailing duties in this case.57
5.91.
The oral statement is organised as follows: (a) a presentation of the facts, (b) a review of the substantive legal claims, and (c) the issue of standard of review and the US request to make a preliminary ruling on expedited reviews.
5.92.
The basic thrust of the EC case is that countervailing duty measures are exceptional, non-MFN measures that are permitted only and so long as it is necessary to offset injurious subsidies. This is true as regards both the original imposition of countervailing duties and their review under the sunset provisions of the SCM Agreement.

2. Factual Background

5.93.
On 9 July 1993, DOC decided that exports of the product were subsidised. DOC countervailed five subsidy programmes (CVD rate 0.60 per cent). DOC found that two programmes, the Capital Investment Grants (CIG) (0.39 per cent) and the Aid for Closure of Steel Operations (0.06 per cent) were non-recurring grants and the benefit was allocated over a 15-year period. The other three programmes were found to be recurring subsidies (0.15 per cent of the total subsidy amount).
5.94.
On 1 September 1999, DOC automatically initiated a sunset review of the definitive duties. On 20 March 2000, concluded that revocation of the CVD order was likely to lead to recurrence of continuation or recurrence of a countervailable subsidy at a net subsidy rate of 0.54 per cent. DOC found that the CIG programme would continue to confer a benefit beyond the end of the sunset review. The US simply took the subsidy rate from the original investigation in 1993 and deducted the amounts related to two programmes. On 2 August 2000, the DOC confirmed these findings in the final results.

3. Legal Arguments

(a) The US standard of initiation applied to sunset reviews leads inevitably to a violation of Article 21.3 of the SCM Agreement

5.95.
Pursuant to Article 21.3 SCM Agreement, WTO Members have an obligation to terminate CVD measures no later than 5 years from their imposition, unless the authorities determine that the expiry of the duties would be likely to lead to continuation or recurrence of subsidisation and injury. The EC considers that Article 21.3 reinforces the requirement of Article 21.1 by creating an obligation to terminate duties not later than 5 years from their imposition. Therefore, the effect of these provisions is that the continuation of CVD measures beyond the five years period is the exception. Indeed, a proper interpretation of the provisions of Article 21.1 and of the terms "not later than" in Article 21.3 suggest clearly that a CVD should be terminated even before 5 years, if its application is not necessary to counteract injurious subsidisation. Therefore, the combined effect of these provisions is that the continuation of CVD measures beyond the five years period is an exceptional situation. This corresponds to the object and the purpose of the SCM Agreement, which – as explained above – is to guarantee that trade defence measures are to be imposed and maintained only if they are necessary to offset injurious subsidisation.
5.96.
The EC considers that self-initiation without the necessary supporting evidence is in violation of the above principle. In this respect, self-initiation is justified only if the domestic authorities have an equivalent level of evidence that would be required in such a request from the domestic industry. Indeed, without requiring a sufficient amount of evidence of subsidisation and injury in sunset reviews, these become biased in favour of keeping in place unjustified CVD measures. It is clear that the investigating authority has the burden of demonstrating that subsidisation and injury is likely to continue or recur and cannot devise and apply a system that shifts this burden on the exporters and third countries.
5.97.
The US attempts (at para. 63-66) to justify this inconsistency by reading the terms of Article 21.3 in "clinical isolation" and out of context, and without taking into account the proper object and purpose of these provisions. The US also draws (at para. 67) the wrong conclusions from the distinction between the investigation phase and the review phase of a CVD proceeding. Contrary to what the US argues in para. 67, it follows from Article 22.7, in conjunction with Article 21.1 and Article 11.6 of the SCM Agreement, that without sufficient evidence the authorities should not self-initiate automatically sunset reviews, but they should leave CVD orders to expire (unless a valid request is made by the domestic industry). Sunset reviews are not a neutral exercise and do not involve, as the US claims (at para. 58), "taking stock of the situation". Rather, in such cases there is a presumption of expiry of duties and the investigating authority is required to demonstrate that it has sufficient evidence when initiating the sunset review, and that the conditions for initially imposing the CVD measure would still be present in the absence of such measure. It is clear from the language of Article 21.3, read in context and in the light of the object and purpose of these provisions, that the investigating authority has the burden of demonstrating that subsidisation and injury is likely to continue or recur.
5.98.
The EC submits that there is no difference as regards the object and purpose of the initial investigations and of sunset reviews. Their common objective is to determine if there is subsidisation and injury which will enable the application of CVD measures in principle for the next five years. It is generally admitted that every text, no matter how clear on its face it is claimed to be, as the US argues here of Article 21.3, requires to be scrutinised in its context and in the light of the object and purpose which is designed to serve. The US arguments throughout their written submission also run counter to the basic principle of good faith interpretation, because they lead unavoidably to results which are manifestly absurd and unreasonable58 and, as explained above, would reduce parts of Article 21.3 to inutility. So, the question is not whether the Community proposes to read into Article 21.3 words which are not there, as the US now argues, but whether the interpretation it proposes makes sense in the light of the object and purpose of these provisions in their context.
5.99.
The EC claim will be better understood if the automatic initiation under US law is examined in two situations: first, in the case of "expedited" sunset reviews, and, second, as applied to the facts in the present case.
5.100.
Under the expedited procedure, if domestic interested parties file a "notice of intent to participate", the sunset procedure continues and interested foreign parties, i.e. exporters and the foreign government, are required to file substantive responses. This of course imposes serious economic and other costs on foreign exporters and other WTO Members. If the foreign government, or exporters accounting for at least 50 per cent of exports, do not file substantive responses for whatever reason, DOC will shorten the timeframe for the review and will normally presume without further investigation that subsidisation continues.59 Therefore, the US procedure of expedited reviews, through automatic initiation and presumption of likelihood of continuation or recurrence, is applied so as to absolve the US authorities from their basic duty of investigation and demonstration of likely subsidisation and injury and, hence, is clearly biased towards continuation and perpetuation of countervailing duties, in violation of the duty to terminate them as set in Articles 21.3 and 21.1 of the SCM Agreement. It also enables the DOC, without any support in the SCM Agreement, to potentially ignore information submitted by a significant number of exporters.

(b) The domestic authorities are under an obligation to "determine" the likelihood of continuation or recurrence of subsidisation

5.101.
Article 21.3 SCM Agreement requires that the authorities determine that the expiry of the duty would be likely to lead to continuation or recurrence of subsidisation and injury. However, US law does not require a new investigation but requires in sunset reviews the DOC to consider only: (a) the net countervailable subsidy determined in the original investigation, and (b) whether there were any changes in the subsidies since the original investigation. In the EC's view, DOC’s refusal to conduct any meaningful fresh investigation in a sunset review is in violation of the requirements of Article 21.3 of the SCM Agreement.
5.102.
The EC considers that the DOC’s refusal to conduct any meaningful fresh investigation in a sunset review is in violation of the requirements of Article 21.3 of the SCM Agreement. The US practice not only contravenes the duty placed by Article 21.3 on the domestic authorities to make a fresh "determination" about the likelihood of continuation or recurrence of a subsidy, but, by arguing that administrative reviews should have been requested, it also shifts the burden of proving a change in circumstances on the foreign exporters and at a previous stage of the procedure. Such reviews, however, are considered by the SCM Agreement as non-obligatory and are in any event overtaken by the sunset review. The fact that a party did not request an administrative review cannot, therefore, relieve in any way the domestic authorities from their basic and primary duty to "determine" the likelihood of continuation or recurrence of subsidisation [In the present case, the exporters could not request an administrative reviews, as no shipments to the US had been made since the original CVD order]. In fact, DOC sunset practice has been to consistently ignore reduced subsidy rates found in administrative reviews. At any rate, in the present case the German exporters did all they could to submit relevant information to DOC.
5.103.
In this case, the German producers stated that the CIG programme had been terminated and could not provide any further benefits. DOC had already determined in the original investigation that this programme applied only to investments made prior to 1 January 1986. Despite this evidence, DOC stated that it was unable to determine whether any additional benefits under these programmes were received subsequent to the period of investigation.The US admits that DOC did not recalculate the rate but used the original subsidy rate for CIG. This violates Article 21.3 SCM Agreement because DOC did not take into consideration the termination of the CIG programme despite evidence on the record.
5.104.
Despite this evidence, DOC rejected the claims of the German exporters. It stated that since no administrative reviews of these orders had been conducted, it was unable to determine whether any additional benefits under these programmes were received subsequent to the period of investigation. It should be stressed that, DOC had also the complete record from the original investigation, including the questionnaire responses and the calculation memoranda, and could have easily verified whether the benefits received by the German manufacturers of the subject merchandise under the CIG after 1 January 1986 were de minimis. The US now admits (at para. 40) that DOC did not recalculate the rate but referred back to the original subsidy rate for CIG. This violates Article 21.3 SCM Agreement because DOC did not take into consideration the termination of the CIG programme despite evidence on the record. Indeed, the grounds of the US explanation violate the "likely to lead" requirement in Article 21.3 and the "necessity" requirement laid down in Article 21.1 of the Agreement. It also runs counter to Article 21.4, in conjunction with Article 12.5, because the US authorities did not make any effort to satisfy themselves of the accuracy of the claims made by the German exporters. Therefore, the US failed to provide an adequate and reasonable explanation of how the facts support their decision in this case.
5.105.
Moreover, DOC should have also reached the conclusion that the effects of the CIG programme had ended on the basis of its declining balance methodology.
5.106.
As the Appellate Body held several times, the requirement to conduct an "objective assessment" of a claim imposes upon the Panel the obligation to review, first, whether the competent authorities have evaluated all the relevant factors60 and, second, whether the authorities have provided a reasoned and adequate explanation of how the facts support their determination.61 It follows that although panels are not entitled to conduct a de novo review of the evidence, nor to substitute their own conclusions for those of the competent authorities, this does not mean that panels must simply accept the conclusions of the competent authorities. Panels should instead examine the explanation offered in depth and in the light of all the facts before it.
5.107.
It should further be stressed that in the present case, the calculation memorandum of the original investigation62 shows that Thyssen last received payments under the CIG programme in 1989, and that the amount received after 1985 was so small that it would have been automatically expensed in the year of receipt under the DOC 0.5 per cent rule.63 Therefore, in accordance with their obligations under Article 21.3 and 4, the DOC should have taken this into account.
5.108.
The EC also considers that in the present case the US violated the standards of Article 21.3 SCM Agreement regarding the treatment of evidence. The German producers presented to DOC during the sunset review the reasons why the CIG programme should have been considered as terminated and as not providing any further benefits. Based upon this evidence, the German producers believed that DOC would find that subsidisation under the CIG programme would not be likely to continue or recur. However, in its preliminary results of 20 March 2000, DOC continued to assign a 0.39 per cent subsidy rate for this programme.
5.109.
It follows from the above that DOC’s refusal to conduct any fresh investigation in a sunset review and take into account all relevant and available evidence is in violation of the requirements of Article 21.3 of the SCM Agreement

(c) The de minimis rule under Article 21.3 of the SCM Agreement

5.110.
The EC considers that the U.S. interpretation and application of the sunset provisions in Article 21.3 as regards the de minimis requirement is inconsistent with the SCM Agreement. The de minims requirements of Articles 11.9 of the SCM Agreement are applied by the US only to original investigations (but a 0.5 per cent de minimis rule is applied in sunset and administrative reviews under its domestic laws).
5.111.
The EC believes that the object and purpose of the sunset reviews and of the original investigations is the same, i.e. the application of a CVD measure for the next five years and, therefore, the de minimis rule of 1 per cent in Article 11.9 should apply to sunset reviews.
5.112.
Indeed, a systematic and good faith interpretation of Article 21.3 with Articles 21.1, 22.1, 22.7 and 11.9 of the SCM Agreement, would suggest clearly that the de minimis rule of 1 per cent should be applied also in sunset reviews. The US has failed to demonstrate so far how it is that essentially non-recurring and expensed subsidies of 0.60 per cent in the original investigation are likely to continue and cause injury in a sunset review.64 The absurdity of the US approach is that a subsidy that, if examined in a new investigation after 1995, would have been found to be de minimis under Article 11.9 of the SCM Agreement, would now become not de minimis under a sunset review and would continue to be applied for 5 more years, even in the absence of a change of circumstances. It is therefore particularly revealing to examine the application of the US approach to the facts of the present case.
5.113.
There is common agreement that of the five subsidy programmes examined in the original investigation, 2 have been terminated, i.e. the Structural Improvement Aids (0.05 per cent) and the Zonal Border Area (0.01 per cent). Two of the three programmes that, according to DOC, continued to exist, i.e. the Aid for Closure for Steel Operations (0.06 per cent) and the ECSC Redevelopment Aid (0.08 per cent) would have expired definitively in 2002. The last and most important programme, the Capital Investment Grants, was found by DOC to be non-recurring, and its original net subsidy rate of 0.39 per cent would have been reduced to 0.00234 per cent in accordance with the DOC’s 15-year allocation period and calculation methodology. Because DOC only calculates subsidy rates to the second decimal point, a net rate of 0.00234 per cent would have been equivalent to zero. Therefore, on the basis of DOC’s own calculation methods, the only 2 likely to continue subsidies would have provided a net amount of 0.14 per cent (i.e. 0.06% + 0.08%).
5.114.
The US provides now no credible explanation. In particular, we fail to see the reason for which the US cites (at para. 94) that there were 2 more programmes under the ECSC treaty which provided zero benefit to these products in the period of investigation, and the reason for which it states (at para. 34) that "domestic parties made allegations concerning new subsidy programmes providing benefits to the German steel industry", but later rejected these claims (at para. 40) because no administrative reviews had been conducted.65 We cannot help but come to the conclusion that this is yet another ingenious way of DOC and the US to convey the impression (e.g., at paras. 95-96) that there were in place a number of important subsidy programmes ready to be re-deployed soon after the CVD order were to expire. One need only look carefully into the facts of this case to realise that this is absolutely not correct.
5.115.
The US states that original investigations and sunset reviews are different and the SCM Agreement does not require recalculation of CVD in sunsets. However, there is nothing in Article 21.3 to prevent the US DOC from recalculating the amount of subsidy in a sunset by establishing a relevant investigation period and doing a recalculation. Moreover, it is required to quantify the amount of subsidy likely to prevail in the absence of measures. DOC should have based its quantification on the facts, and examine whether it can meet its burden of proof by positively demonstrating likelihood of continuation of subsidisation and injury. It has failed to do so in this case.
5.116.
The US advanced further the argument that "the mere continued existence of the CIG programme could warrant maintaining the duty beyond the five-year point, even if the amount of the subsidy was currently zero, as stated in footnote 52, because subsidisation may be likely to recur absent the discipline of the duty" (at para. 81). Footnote 52 refers only to assessment proceedings, which are carried out annually on demand and involve retrospective calculations of the amount of subsidy for a particular period. Sunset reviews are completely different and involve an element of prospective analysis, i.e. whether subsidisation and injury is likely to recur. There is, consequently, no basis for the argument which the US is attempting to draw from the text of Footnote 52 of the SCM Agreement.
5.117.
The US also alleges that the current level of subsidisation is not decisive as to whether subsidisation is likely to recur (at para. 74). Although the current level of subsidy might not be decisive as to whether a subsidy is likely to recur, it is nevertheless a very important factor. Especially in the case of a non-recurring subsidy which is already at de minimis level, because the future level of the subsidy is known and, under the US declining balance methodology, the subsidy amount will always go down in the future. In the present case, the CVD was based on non-recurring subsidies allocated over time (0.60 per cent). Because no new subsidies were granted and on the basis of all available facts in this case, the US should have terminated this measures as the level of subsidy likely to continue in this case was by far too low, even much lower than the 0.50 per cent applied unilaterally by the US domestically. The EC submitted that the 0.50 per cent de minimis threshold applied by the US domestically is not appropriate, and that since the rate of 0.54 per cent in the review is below the 1 per cent de minimis level which should apply in sunset reviews, the US was in breach of Article 21.3 in continuing the CVD measure.

(d) The Obligation to Bring its Law Into Conformity with the WTO Agreements

5.118.
Article XVI:4 of the WTO Agreement and Article 32.5 of the SCM Agreement oblige Members to bring their domestic law into conformity with their obligations under the WTO Agreements. The EC has demonstrated that the basic US law, the accompanying regulations and practices relating to sunset reviews of countervailing duties and their concrete application to the facts of the present case are inconsistent with a number of provisions of the SCM Agreement, i.e. Articles 21, paragraphs 1, 3 and 4, Article 10 and Article 11.9. For those reasons, the EC submits that the US failed to bring its domestic law in conformity with WTO obligations and, thereby, violating Article 32.5 of the SCM Agreement and, consequently, Article XVI.4 of the WTO Agreement.

4. Standard of review

5.119.
The applicable standard is Article 11 DSU. However, the US claims that the panel cannot consider facts and evidence which were not considered by the US authorities in the sunset review. The EC disagrees based on Appellate Body decisions regarding Article 11 DSU. In European Communities – Hormones, it stated that "the applicable standard is neither de novo review as such, nor 'total deference', but rather the 'objective assessment of the facts' ". In Argentina – Footwear Safeguards, the Appellate Body clarified the concept of "objective assessment". The requirement to conduct an "objective assessment" of a claim has thus two elements: a review of whether competent authorities have evaluated all the relevant factors and, second, a review of whether the authorities have provided a reasoned and adequate explanation of how the facts support their determination. The EC believes the US has failed to provide an adequate and reasonable explanation of how the facts support their decision in this case.
5.120.
As with regard to requirement of a reasonable and adequate explanation, in the United States - Lamb Safeguards case, the Appellate Body held that panels must review whether the competent authorities' explanation fully addresses the nature, and, especially, the complexities, of the data, and responds to other plausible interpretations of that data.66
5.121.
It follows also that although panels are not entitled to conduct a de novo review of the evidence, nor to substitute their own conclusions for those of the competent authorities, this does not mean that panels must simply accept the conclusions of the competent authorities. Panels should instead examine the explanation offered in depth and in the light of all the facts before it.

5. US Request for preliminary ruling

5.122.
The EC consider without foundation the US request that the Panel issue a preliminary ruling that the EC's claims regarding the expedited sunset review procedure are not within its terms of reference.
5.123.
This Panel was established by the DSB on 10 September 2001 with standard terms of reference. The matter referred to it is, therefore, the one described by the EC in its request for the establishment of the Panel (document WT/DS213/3). The EC referred explicitly to Section 751. c) of the Tariff Act of 1930, which sets out the procedures to conduct sunset reviews and this includes also expedited reviews. The expedited reviews have also been discussed explicitly or implicitly in the consultations, as the procedural and substantive requirements on evidence which foreign producers have to provide in all types of sunset reviews were indeed the subject of the consultations. The US was therefore not deprived of its rights of defence.

6. Conclusions

5.124.
For these reasons, the EC respectfully requests that the Panel finds that the US basic countervailing duty law (Tariff Act of 1930), its accompanying regulations (sunset regulations) and policy practices (sunset policy bulletin), and their concrete application to imports of certain corrosion-resistant carbon steel flat products from Germany in the present case are inconsistent with Article 21 paragraphs 3, 1 and 4, Article 10 and Article 11.9 of the SCM Agreement. For the above reasons, the US countervailing duty law, regulations and practice should also be considered to be inconsistent with Article 32.5 of the SCM Agreement and, consequently, should be found to violate also Article XVI.4 of the WTO Agreement.

D. ORAL STATEMENT OF THE UNITED STATES AT THE FIRST MEETING OF THE PANEL

5.125.
On behalf of the United States delegation, I would like to thank the Panel for this opportunity to comment on certain issues raised by the EC in its First Written Submission. We do not intend to offer a lengthy statement today; you have our written submission, and we will not repeat all of the comments that we made there. We will be pleased to receive any questions you may have at the conclusion of our statement.
5.126.
Mr. Chairman, this proceeding presents three basic questions. The first is a purely legal question: Does the United States act inconsistently with Article 21.3 by self-initiating sunset reviews without regard to the evidentiary provisions of Article 11.6? The second is also a purely legal question: Does the United States act inconsistently with Article 21.3 by not applying the de minimis provisions of Article 11.9 in sunset reviews? The answer to both of these questions is no, for a simple, yet fundamental reason – it is impossible to act inconsistently with an obligation that does not exist.
5.127.
I will return to these two issues in a moment, but first let me address the third question, which goes to the specific CVD determination at issue in this proceeding – Commerce’s determination that expiry of the countervailing duty on certain corrosion-resistant carbon steel flat products from Germany would be likely to lead to continuation or recurrence of subsidization. The question here is whether Commerce’s determination was based upon an appropriately conducted review of all relevant and properly submitted facts. An "objective assessment" of Commerce’s findings and actions supports an answer in the affirmative.

1. Commerce Properly Determined That Expiry Of The Duty Would Be Likely To Lead To Continuation Or Recurrence Of Subsidization

5.128.
Article 21.3 defines the point in time at which the authorities must take stock of or terminate a duty – that is every five years. Article 21.3 also defines the circumstances under which maintaining a duty may be considered "necessary" – that is when continuation or recurrence of subsidization and injury is likely. An authority’s decision to maintain a duty must be supported by evidence of these requisite circumstances.
5.129.
Setting aside the issue of injury, which is not being challenged in this case, what does it mean to determine likelihood of continuation or recurrence of subsidization?
5.130.
First, we consider, individually, the words establishing the circumstances under which maintaining a duty may be considered necessary. The word "likely" carries with it the ordinary meaning of "probable". Where continuation or recurrence of subsidization is probable, this probability would then constitute a proper basis for entitlement to the continued imposition of a countervailing duty.
5.131.
The word "continuation" expresses a temporal relationship between past and future. Something that happened in the past may continue in the future. An example might be where benefits from an untied, non-recurring financial contribution continue to flow.
5.132.
The word "recurrence" also expresses a temporal relationship between past and future. Something that happened in the past may happen again in the future. An example might be where there was not a recent financial contribution under a particular subsidy programme, but the programme still exists and may be used in the future. In this situation, the existing programme is a source for new financial contributions – in other words, its continued existence enables recurrence of subsidization.
5.133.
Considered together then, these words indicate that determining likelihood of continuation or recurrence requires a consideration of future, rather than present, circumstances. What are the prospects of subsidization in the future? Without the discipline of the duty, is subsidization likely to continue or recur? The analysis required in a sunset review, therefore, is necessarily prospective in nature.
5.134.
Support for this proposition is found in the text of Article 21.3 itself. As discussed in our First Written Submission, note 52 provides that a finding in the most recent assessment proceeding that no duty is to be levied shall not by itself require the authorities to terminate the definitive duty. This indicates that the current level of subsidization is not decisive as to whether subsidization is likely to continue or recur.
5.135.
In the sunset review involving German corrosion-resistant steel, Commerce found likelihood based on two unrefuted facts. The first fact is the continued existence and availability of countervailable subsidy programmes previously found to have been used by German producers. The second fact is the continued existence of benefit streams from a countervailable subsidy programme previously found to benefit German producers.
5.136.
The EC argues that there are a number of factual and procedural flaws with Commerce’s sunset determination, such that Commerce’s determination is in violation of provisions of the SCM Agreement. In out First Written Submission, we address and rebut the EC’s claims in greater detail. Today, I will only briefly touch upon the EC’s two main claims. The first concerns the Capital Investment Grants, or CIG, Programme; the second concerns whether the German producers and the EC were afforded "ample opportunity" to participate in the underlying sunset review.
5.137.
At the outset, I would note that rather than demonstrating that Commerce’s findings or procedural actions were inconsistent with the SCM Agreement, the EC’s claims present essentially another view of the facts. However, Article 11 of the DSU directs panels to make an "objective assessment" of the facts of the case and of the applicability and conformity with relevant agreements. This "objective assessment" must necessarily focus on the consistency of the sunset review with the requirements of Article 21.3 and Article 12.
5.138.
With respect to the CIG Programme, the EC theorizes that, using Commerce’s declining balance methodology, the benefits remaining from the programme would be de minimis. You will recall that in the original investigation, Commerce found that German producers of corrosion-resistant steel benefitted from non-recurring grants under this programme. In the sunset review, Commerce found that benefit streams from the CIG grants would continue beyond the five-year period. While the EC’s declining balance theories are not unreasonable, they only address half of the equation. In particular, it is not possible to calculate a particular rate of subsidization using this methodology without information concerning sales of the subject merchandise. The EC’s claim presumes that sales volumes have remained constant, a presumption for which there is no evidence in the record.
5.139.
In addition, the EC’s argument in this regard is based partly on a calculation memorandum which Commerce properly declined to consider in the sunset review below. As we demonstrate in our First Written Submission, it is not appropriate for the Panel to now consider this business confidential document. Nevertheless, even if this document were appropriately part of the record before the Panel, it does not prove the EC’s argument. This is because the memorandum only provides absolute subsidy amounts; it does not contain the information on sales volumes necessary to calculate the ad valorem rate of subsidization.
5.140.
Nothing in the SCM Agreement requires consideration of the magnitude of subsidization in determining the likelihood of continuation or recurrence of subsidization. Furthermore, as the Appellate Body recognized in UK Lead Bar, benefits from a non-recurring subsidy continue to flow. As a result, Commerce’s "continuation of subsidization" finding with respect to the CIG programme is correct.
5.141.
With respect to Commerce’s procedural actions in the sunset review, the EC argues that Commerce did not provide an "ample opportunity" to present in writing all evidence which the parties considered relevant to the proceeding. As a factual matter, this assertion is simply incorrect.
5.142.
Commerce’s published Sunset Regulations contain the standard sunset questionnaire and provide an opportunity for parties to submit any argument and information they consider relevant to Commerce’s sunset determination. Commerce’s regulations also set a 30-day deadline for the submission of such information and provide for extensions of that 30-day deadline. The German Producers and the EC were on notice of these information requirements and options, as well as the applicable deadlines, at least 15 months prior to the initiation date for the sunset review.
5.143.
The EC has not demonstrated how Commerce’s actions in this regard violate any of the evidentiary and procedural requirements of Article 12. Fifteen months certainly seems to provide "ample opportunity" to gather and present any evidence the German Producers and the EC considered pertinent to Commerce’s sunset determination. I would also note that 15 months is longer than the normal deadline in Article 21.4 for the conduct and completion of sunset reviews. That the German Producers and the EC failed to avail themselves of the ample opportunity to present evidence is not a dereliction that can be ascribed to Commerce’s procedural actions in this case.

2. It is impossible to act inconsistently with an obligation that does not exist

5.144.
I will turn now to the two purely legal claims raised by the EC. The first claim is that an authority must satisfy the evidentiary requirements of Article 11.6 before self-initiating a sunset review under Article 21.3. The second claim is that application of the Article 11.9 de minimis standard in a sunset review is required under the SCM Agreement. As set out more fully in our First Written Submission, there is no support in the SCM Agreement for these claims.
5.145.
Article 31 of the Vienna Convention contain some very basic principles of treaty interpretation. As articulated by the Appellate Body, these principles "neither require nor condone the imputation into a treaty of words that are not there...." The EC’s arguments in this case run afoul of this fundamental proposition.
5.146.
With respect to the self-initiation issue, the EC argues that the Panel should read into Article 21.3, the requirements of Article 11.6 – the provision of the SCM Agreement that deals with evidentiary requirements for self-initiation of an investigation. With respect to the de minimis issue, the EC argues that the Panel should read into Article 21.3, the requirements of Article 11.9 – the provision of the SCM Agreement that deals with the de minimis standard for an investigation. There is no support in the Agreement for the EC’s theories.
5.147.
With no textual foundation, the EC simply asserts that, under the SCM Agreement, sunset reviews are essentially nothing more than new investigations. It then takes that unsupported assertion and makes a further leap of logic to the conclusion that various provisions of Article 11 (addressing "Initiation and Subsequent Investigation") must therefore apply in sunset reviews carried out under Article 21 (addressing "Duration and Review of Countervailing Duties").
5.148.
Japan, in its third party submission, makes an even more impressive leap of logic, arguing that Article 11 requirements are made applicable to Article 21.3 sunset reviews by virtue of the fact that Article 22 – which addresses "Public Notice and Explanation of Determinations" – applies to reviews under Article 21 (pursuant to paragraph 7 of Article 22). Apparently, according to Japan, the mere mention of Article 11 in Article 22.1 creates an obligation to apply Article 11 in Article 21.3 sunset reviews. I would also note that the provisions of Article 22 apply "mutatis mutandis" to reviews. This means that the public notice and explanation provisions are applicable to reviews, but with "necessary changes" or "changes as appropriate."
5.149.
Mr. Chairman, it goes without saying that treaty interpretation does not work that way. One cannot create a new set of international obligations through unsupported theories and unfounded suppositions that ignore the very words of the treaty being interpreted. It is well-accepted that, under the Vienna Convention, the ordinary meaning of a treaty is the basis for interpreting that treaty. The fact that the EC and Japan have to try so hard to find a connection between Article 11 and Article 21 is itself evidence of the lack of the very connection that they seek. Put differently, the United States agrees with the EC and Japan that Article 11 provides context for purposes of interpreting Article 21.3; but it is context that disproves their assertions.
5.150.
As discussed in our First Written Submission, the drafters clearly knew how to cross-reference a particular provision to make it applicable in the context of Article 21 reviews. For example, Article 21.4 expressly makes the evidentiary requirements of Article 12 applicable in Article 21 reviews. Even Japan’s argument proves the point – Article 22.7 makes the public notice and explanation requirements of Article 22 applicable in Article 21 reviews on a mutatis mutandis basis.
5.151.
The uncontested fact is that the drafters did not make the initiation and subsequent investigation requirements of Article 11 applicable in Article 21 reviews. And no amount of lawyering can override the plain text of the SCM Agreement or create an obligation that does not exist. The Article 11.6 evidentiary prerequisite simply does not apply to Article 21.3 sunset reviews; and neither does the Article 11.9 de minimis standard.
5.152.
The EC makes claims based on its "reasonable and legitimate expectations" on these issues; yet it is the legitimate expectations of the Members as a whole, as expressed in the agreed text of the treaty, that are at risk of being infringed in this case. As the Appellate Body has stated in India Patent Protection at paragraph 45:

The legitimate expectations of the parties to a treaty are reflected in the language of the treaty itself. The duty of a treaty interpreter is to examine the words of the treaty to determine the intentions of the parties. This should be done in accordance with the principles of treaty interpretation set out in Article 31 of the Vienna Convention. But these principles of interpretation neither require nor condone the imputation into a treaty of words that are not there or the importation into a treaty of concepts that were not intended.

(Emphasis added.) If the Members had actually agreed that various provisions of Article 11 should apply in sunset reviews carried out under Article 21, the text would reflect that agreement, just as it does with respect to the application of Article 12 to Article 21 reviews. The EC is improperly trying to have the Panel do what the negotiators did not.

5.153.
The EC is asking the Panel to read into Article 21.3 words, and hence obligations, that are not there. For this reason, the EC’s claims concerning self-initiation and de minimis must fail.

E. RESPONSES OF THE EUROPEAN COMMUNITIES TO QUESTIONS FROM THE PANEL FOLLOWING THE FIRST MEETING

Both parties

Q1. In your view:

(a) Is the de minimis standard contained in Article 11.9 of the Agreement on Subsidies and Countervailing Measures ("SCM Agreement" or "the Agreement") applicable to reviews under Article 21.2?

Reply

5.154.
Article 21.2 deals with reviews during the lifetime of the measure with the aim of examining whether the continued imposition of the measure remains necessary (based on the principle set out in Article 21.1). In the EC's view, with regard to the US system, we should distinguish between two types of review under Article 21.2, i.e. duty assessment reviews (administrative reviews under US law) and reviews where the effective need for continued imposition of the measure is examined (changed circumstances review under US law). As regards a changed circumstances review, the de minimis standard under 11.9 is applicable because Article 21.2 requires that if the authorities determine that the countervailing duty is no longer warranted, it shall be terminated immediately. Changed circumstances reviews can thus affect the scope and, indeed, the continued imposition of the measure. Under Article 11.9 of the Agreement, the subsidy amount must be above 1 per cent to permit the imposition of a countervailing duty in the first place. It follows that if an interested party can show, in a changed circumstances review, that the subsidy amount will not rise above 1 per cent upon removal of the measure, the measure should be terminated immediately. It should be noted that virtually the same terminology "immediate termination" and "immediately terminated" is used in Articles 11.9 and 21.2 respectively.
5.155.
On the other hand, a de minimis amount of subsidy found during a simple administrative review would not by itself oblige the authorities to terminate the proceeding, since the purpose of such a review is only to determine what amount of duty has to be collected. (Footnote 52 of the SCM Agreement). Thus, as the Korea – DRAMS panel determined, simple administrative reviews have no effect on the scope and continued imposition of a measure and cannot lead to its termination. As the EC explained in its first submission, administrative reviews in the SCM Agreement appear to fall under Article 21.2 by default, whereas in the Anti-Dumping Agreement (Footnote 21) they are not considered to be reviews at all.

(b) Are the negligible import volume and injury standards contained in Article 11.9 and the negligible injury standard contained in Article 15.3 applicable to reviews under Article 21.2?

Reply

5.156.
Negligible import volume and injury standards are provided for in the SCM Agreement because WTO members consider that such a small volume of imports can not be the cause of any injury sustained by the domestic industry. The use of the term "shall" indicates that this is not a refutable presumption. Therefore, if a certain volume of imports cannot cause injury at the initiation of an investigation, this irrefutable presumption applies to reviews under Article 21.2. in cases where it is demonstrated that the amount of the subsidy is less than 1 per cent ad valorem and is not going to increase above this level or that imports would not rise above de minimis upon removal of the measure. The terms, object and purpose of Article 21.2 in context compel such a conclusion.

(c) Are the negligible import volume and injury standards contained in Article 11.9 and the negligible injury standard contained in Article 15.3 applicable to reviews under Article 21.3?

Reply

5.157.
Article 21.3 lays down clearly a presumption of termination ("shall be terminated"). The authorities are required, therefore, to show that the conditions for imposing the original measure would, in its absence, continue to exist or recur and it is for the domestic industry and the investigating authority to show that, if the measure were removed, imports would be likely to again rise above de minimis. For essentially the same reasons stated in the reply to the previous question above and those explained more detail in the first written submission of the EC, the EC considers that the standards contained in Articles 11.9 and 15.3 of the SCM Agreement are applicable to reviews under Article 21.3 thereof.

(d) Are the negligible import volume standards for developing countries set out in Article 27.10 applicable to reviews under Article 21 in general, and to reviews under Article 21.3 in particular? Is the 2 per cent de minimis standard for developing countries set out in Article 27.10(b) applicable to reviews under Article 21 in general, and to reviews under Article 21.3 in particular? And is the 3 per cent de minimis level for certain developing countries set out in Article 27.11 applicable to reviews under Article 21 in general, and to reviews under Article 21.3 in particular?

Please explain in detail.

Reply

5.158.
The higher de minimis thresholds for developing countries are equally applicable to initial investigations and reviews examining the need for continued imposition of the CVD (under Article 21.2) and sunset reviews (under 21.3). The purpose of special de minimis thresholds for developing countries lies in Article 27.1 of the SCM Agreement. WTO members recognize that subsidies play an important role in the further economic development of developing countries and, therefore, agreed in Articles 27.10 and 27.11 to increase the minimum import volumes and subsidy amounts below which injury is presumed not to be caused. Likewise, this is an irrefutable presumption.
5.159.
Article 27.10 states that "Any countervailing duty investigation of a product originating in a developing country Member shall be terminated as soon as the authorities determine…". It would be contrary to the explicit provisions of the SCM Agreement to grant certain exemptions because they are considered essential for development needs and only apply them to original countervailing investigations. It would also be contrary to the terms of Articles 27.10 and 27.11, read in conjunction with Articles 21.1, 21.2 and 21.3, not to apply them in reviews. Clearly the reference to "subsidization" in these Articles must be taken to refer to a level of subsidization which has been determined to be injurious with reference to the Member concerned.
5.160.
An example may illustrate this point. Article 27.11 provides for a developing country Member, which eliminates export subsidies in less than eight years to benefit from an increased de minimis subsidy threshold of 3 per cent instead of 2 per cent. This provision aims to reward developing countries which have accelerated compliance with their obligations. Suppose a certain Member were able to benefit from this provision and the following year was involved in a US CVD sunset review involving one of its products. If the rate of subsidy likely to prevail was found to be 2.5 per cent, i.e. below the de minimis rate now applicable to the Member concerned, the US would continue to impose a CVD order and the Member involved would have obtained absolutely no benefit from its efforts to remove early the export subsidies.

European Communities

Q2. The Panel notes the European Communities' argument regarding the presumption of termination of all countervailing duties ("CVDs") contained in Article 21 of the SCM Agreement and the claim that the US law and statement of policy practice in reviews under Article 21.3 disregard this presumption through the automatic initiation of reviews under Article 21.3, and thereby violate Article 21.3. In what precise manner would the US law and statement of policy practice have had to be different for them to be WTO-consistent? Please respond in light of the phrase "in a review initiated... on their own initiative" contained in Article 21.3 of the SCM Agreement.

Reply

5.161.
The current US law requires the automatic initiation of sunset reviews without any evidence. This ignores the presumption of termination under Article 21.3 and reverses the burden of proof, which should be on the petitioners or the investigating authority to justify the initiation of a review and not on the exporters to justify the termination and the non-initiation of such a review.
5.162.
The problem with US law is that it transforms the text and basic rationale of Article 21.3 and provides for self-initiation without taking account of the standard of evidence required under Articles 21.3 and 4 to meet the burden of proof of either the petitioner or the investigating authority in conjunction with Articles 11.6, 22.1 and 22.7 of the SCM Agreement. It follows that the investigating authority must be in possession of the same or an equivalent amount of sufficient evidence as the domestic industry at the time of initiating the review. The basic US law and its implementing regulations would have to provide for this in order to be WTO consistent. In particular, the US law will have to provide that:

· An applied CVD order shall terminate 5 years after its imposition and will not be reviewed unless the authorities, at the time of initiating the review, possess an amount of relevant evidence that is the same or equivalent to that required in original investigations;

· The burden of providing such evidence is placed on domestic industry or the authorities themselves;

· No reverse or negative inference is drawn from the fact that foreign exporters or producers or third countries do not participate at the initial call for the sunset review, except where Article 12.7 is applicable.

Q3. How, if at all, do the European Communities' own statutory and regulatory provisions in respect of the initiation of reviews under Article 21.3 give meaning to the presumption of termination of all CVDs that the European Communities considers is contained in Article 21 of the SCM Agreement?

Reply

5.163.
Article 18 of Council Regulation (EC) No 2026/97 of 6 October 1997, the Community’s basic CVD regulation, contains the relevant provisions on sunset reviews (known as "expiry reviews"). In accordance with Article 18.4, the Commission publishes a notice of impending expiry of a countervailing measure during the last year of its duration; our practice is to do this nine months before the date of expiry (the five-year anniversary date).
5.164.
The notice of impending expiry states that unless a review is initiated, the measures will expire. In order for a review to be initiated, Community producers must, no later than three months before the expiry date, lodge a request containing sufficient evidence that the removal of the measures would be likely to result in continuation or recurrence of subsidy and injury. If the EU Commission considers that the request contains sufficient evidence, it initiates an expiry review. If there is no request, or if the request contains insufficient evidence, the Commission publishes a notice of expiry and the measures lapse on the five-year anniversary date. This notice of expiry is of purely declaratory nature.
5.165.
There is consequently a presumption of termination, since in the absence of the provision of sufficient positive evidence by the Community producers justifying a review, the measures lapse automatically.

Q4. Does the European Communities consider self-initiation of reviews under Article 21.3 by investigating authorities to be in and of itself WTO-inconsistent?

Reply

5.166.
No. Article 21.3 explicitly provides for the possibility of self-initiation, as does Article 18.1 of Regulation 2026/97. In fact, Article 21.3 provides for either self-initiation or initiation on the basis of a "duly motivated request" from the domestic industry. The Community’s claim is that the investigating authority must be in possession of sufficient evidence, that is the same level or an equivalent amount of relevant evidence that would be required from the domestic industry if it initiates on its own initiative, as in the case in the initiation of new investigations (Articles 11.1 and 11.6). Otherwise, self-initiation would become the easy option and would lead to inconsistent results.

Q5. The Panel notes the European Communities' arguments contained in paragraph 15 of its oral statement at the first meeting of the Panel.

(a) Does the European Communities consider that there is a presumption in the SCM Agreement that all provisions of the Agreement are applicable to reviews under Article 21.3?

Reply

5.167.
The EC believes that no provision of the SCM Agreement can be read in isolation, and that all provisions are potentially applicable mutatis mutandis to Article 21.3, to the extent that they are relevant to the issues covered by Article 21.3 and that their application to Article 21.3 does not create a situation of conflict or is not specifically excluded.

(b) In what circumstances might some provisions of the Agreement not apply to reviews under Article 21.3?

Reply

5.168.
Subject to the above, there are provisions which clearly have no relevance to Article 21.3. As explained below, the provisions concerning multilateral subsidy disciplines, to the extent that they do not apply to countervailing duties at all, are not applicable to sunsets. Furthermore, a small number of provisions which concern countervailing duties cannot, for practical reasons, apply to sunsets. However, one has to be very careful before determining that a certain provision has no such relevance, since certain WTO Members may choose to take account of them in their own practice. For instance, the US, in its sunset practice, takes account of whether the subsidies fall under Articles 3 or 6.1 of the SCM Agreement, although neither of these provisions has any obvious relationship to countervailing duties.

(c) If there are particular provisions that, in the view of the European Communities, do not apply to reviews under Article 21.3, what are they, and why do they not apply?

Reply

5.169.
The SCM Agreement contains two major divisions: multilateral obligations for WTO members and provisions regarding the conduct of countervailing investigations. It is obvious that certain provisions imposing multilateral obligations on WTO members by nature do not apply to countervailing investigations.
5.170.
The EC will provide an overview of the different parts of the SCM Agreement and explain whether provisions apply to countervailing proceedings and reviews under Article 21.3 in particular.

Part I:

5.171.
Article 1 applies to sunset reviews since only subsidies as defined in Article 1 could fall within the concept of "subsidization".
5.172.
Article 2 regarding specificity applies equally to sunset reviews.

Part II:

5.173.
Article 3: the prohibition of certain subsidies as such does not apply to CVD. However, read in conjunction with Article 2.3 SCM Agreement, the definition of prohibited subsidies plays a role for CVD and reviews under Article 21.3 because these subsidies are deemed to be specific.
5.174.
Article 4: is not applicable for CVD and sunsets since it covers specific dispute settlement rules regarding those prohibited subsidies.

Part III:

5.175.
This part deals with actionable subsidies and possible multilateral remedies against those subsidies. It does not apply to CVD and sunsets. In addition, Article 6.1 has now expired.

Part IV:

5.176.
Expired pursuant to Article 31 SCM Agreement on 31.12.99. When applicable, it created a category of non-actionable subsidies which made them non-countervailable in all CVD actions, including sunsets. For example, a regional aid countervailed pre-1995 could have been green-lighted and found non-countervailable in a post-1995 sunset review.

Part V:

5.177.
All provisions of Part V regarding countervailing investigations is in principle applicable mutatis mutandis to sunset reviews under 21.3. For Article 12, there is an explicit reference. For other provisions, e.g. Article 15 on injury, it is clear that the word "injury" in Article 21.3 can only be interpreted with reference to Article 15. The same is true for "domestic industry" with relation to Article 16.
5.178.
In those cases where there is no explicit reference to a specific article or term in Article 21.3, the EC submits that all provisions are in principle applicable to Article 21.3. The only exceptions are provisions which cannot, for practical reasons, serve any purpose in a sunset review. For instance:

Article17 regarding provisional measures can only apply to new investigation, where no measure is yet in force. Provisional measures can be imposed when a preliminary determination is made that a subsidy exists and there is injury. It is mainly intended to prevent injury during the remainder of the investigation and also serves to provide interested parties with an opportunity to comment on the preliminary finding regarding subsidies and injury. It is obvious that this provision is not relevant to a sunset review, where a duty is already in force and can remain effective pending the outcome of the review.

Article 20 covers issues of retroactive collection of duties. This provides for the retroactive collection of provisional duties, and in critical circumstances, retroactive assessment for up to 90 days prior to the provisional stage. Since in a sunset review the duty is already in force this provision does not apply.

5.179.
Part VI regards the establishment and the operation of the Committee on Subsidies and Countervailing Measures. As such, it does not have an impact on the conduct of countervailing proceedings including sunsets. However, the Committee may deal with issues regarding sunset investigations if a WTO Member puts a particular investigation on the agenda for discussion in the Committee.
5.180.
Part VII deals with notifications and does not impact on the conduct of sunset reviews.
5.181.
Part VIII deals with special and differential treatment for developing country members. Articles 27.10 to 27.12 refer to special CVD de minimis thresholds for developing countries, which are equally applicable to new investigations and sunset reviews (Question 1(d)). Article 27.15 deals with Committee review of any CVD measure, including the result of a sunset review. Article 27.1 is a declaratory provision which contributes to justifying the S&D provisions, including those relating to CVD.
5.182.
Part IX does not apply to CVD since it covers some transitional multilateral obligations.
5.183.
Part X concerning dispute settlement covers all aspects of the SCM Agreement. It also covers sunset reviews.
5.184.
Part XI. Its provisions on "other final provisions" apply to sunsets.
5.185.
Article 31. - The expiry of Article 8 removed the non-countervailable category of specific subsidies ; these are now countervailable in sunsets, even if, to reverse the example given under Part IV, the original subsidy had been green lighted in the first investigation.
5.186.
Article 32 applies to sunsets (Article 32.4 explicitly).

Q6. With respect to reviews under Article 21.3, does the European Communities consider that:

(a) All the requirements of Articles 11.4, 11.5, and 11.11 apply to the initiation or self-initiation of reviews under Article 21.3? Why or why not?

Reply

5.187.
The standing requirements in Article 11.4 apply to reviews initiated under Article 21.3. These two articles require that applications or requests must be made "by or on behalf of the domestic industry", which is defined in Article 11.4.
5.188.
Article 11.5 also applies to both. There is no reason why Members should be required to avoid publicising applications under Article 11, but should be allowed to publicise review requests under Article 21.3. Applications under 11.5 and requests under 21.3 are aimed at securing the same objective, that is to avoid unjustified disruptions in international trade on the basis of allegations and claims that are manifestly incorrect. Article 11.11 does not apply to Article 21.3 reviews. Article 21.4 explicitly governs the timing and duration of all reviews under Article 21, and its provisions differ from those of Article 11.11.

(b) Article 13.1 applies to reviews under Article 21 in general, and to reviews under Article 21.3 in particular? Why or why not?

Reply

5.189.
Article 13.1 refers explicitly to applications under Article 11 and "in any event before the initiation of any investigation". Footnote 44 refers to the fact that "no affirmative determination" shall be made without reasonable opportunity for consultations having been given. Such a determination may be made in either a new investigation or review, which implies that consultations should have been offered in the framework of a sunset. Thus, it would seem that Articles 13.1 and 13.2 would apply mutatis mutandis to Article 21.3 reviews.

(c) Article 19.4 applies to reviews under Article 21 in general, and to reviews under Article 21.3 in particular? Why or why not?

Reply

5.190.
Yes, it applies to both types of reviews ("no countervailing duty…"). Moreover, this interpretation is in conformity with the overriding principles laid down in Articles 10 and 21.1 of the SCM Agreement.

Q7. The Panel notes the European Communities' arguments contained in paragraph 37 of its oral statement at the first meeting of the Panel.

(a) Could the European Communities explain its view that the object and purpose of reviews under Article 21.3 and original investigations is the same, in light of the fact that an original investigation determines the existence of subsidization and injury ab initio, while a review under Article 21.3 occurs in the context of the existence of a prior finding of subsidization and injury?

Reply

5.191.
New investigations and Article 21.3 reviews with a positive outcome both have the same result – the application of a countervailing duty for a period of up to 5 years. In order for such duties to be imposed or maintained, the investigating authority must in both cases determine that the conditions for the imposition of a countervailing duty, as laid down in Article 19.1, are fulfilled. The only difference is that Article 21.3 reviews take place with a measure already in force, so that the investigating authority must determine whether, in the absence of that measure, the Article 19.1 conditions would be met. The difference in the wording of the provisions on new investigations and sunset reviews merely reflects the fact that, in the latter case, there is a need to take account of an existing measure in establishing whether the conditions still exist for applying countervailing measures; the object and purpose of both provisions, however, remains the same.

(b) Could the European Communities address specifically the difference, if any, between a finding of "the existence of and amount of the subsidy and that, through the effects of the subsidy, the subsidized imports are causing injury", so as to justify imposition of a CVD under Article 19.1, and a finding that "expiry of the duty would be likely to lead to continuation or recurrence of subsidization and injury" sufficient to justify continuation of the CVD under Article 21.3?

Reply

5.192.
There are not differences except that in the case of Article 21.3 there are CVD measures already in force. The purpose of an Article 21.3 review is to determine whether, absent the measures in force, all the conditions for the imposition of a duty under Article 19.1 would still be present. Since an Article 21.3 review always takes place with a measure in force, this inevitably involves a certain element of prediction based on the facts presented of what would happen if the measures were left to expire. The investigating authorities are required to meet the "likely" standard, while applying the provisions of Article 21.1. However, the concepts of subsidization, causality and injury have to be interpreted in the same way in the context of both provisions. Moreover, it has to be based on "evidence", not on pure conjecture and speculation. This follows from Article 21.4, which makes Article 12 relating to evidence applicable to reviews. Moreover, the requirement to re-establish the existence of the Article 19.1 conditions in a sunset review flows from the general presumption of termination in Article 21.3, in conjunction with the basic requirements of Article 21.1.
5.193.
In this regard, it should be noted that, regardless of the presence or absence of measures, some factual situations are inherently more predictable than others. For example, while it may be relatively difficult to determine to what extent a company will resume dumping practices if a measure is dropped, sovereign States normally behave differently and in a more predictable way as regards the granting of subsidies. Moreover, a non-recurring subsidy in a CVD case (as is the case here) will always be predictable, since the benefit stream calculated is known for the whole allocation period of the subsidy. In such a case, it would be for the investigating authority to justify, on the basis of positive evidence, a finding that any other related factors, e.g. the denominator, would change.

Q8. The European Communities refers, in paragraph 3 of its oral statement at the first meeting of the Panel, to the fact that CVD measures are exceptional, non-MFN measures. The European Communities further relies on a characterization of continuation of a measure after a review as an "exception" in arguing that the provisions of Article 21.3 must be strictly interpreted.

(a) First, does the European Communities consider that this requirement of strict interpretation applies to the entirety of Part V of the SCM Agreement?

Reply

5.194.
The EC considers that the provisions of Article 21.3 SCM Agreement relating to continuation of a CVD measure constitute an exception to the general rule which provides that CVD orders should in principle expire after 5 years. The exceptional character of continuation of the duty beyond the five-year period is reflected in the use of "unless" which introduces the exception. The EC considers that a number of provisions of the SCM Agreement (e.g. Articles 10, 11.3, 11.5, 11.9, 19.4, 21.1, etc.) provide implicit support for the proposition that CVD orders should be imposed after careful and strict interpretation of all the provisions of the SCM Agreement.

(b) Second, could the European Communities explain how, in its view, reading the text of Article 21.3 to include obligations not explicitly set out therein, but set out in Article 11.6 (or 11.9), is a "strict interpretation" of Article 21.3?

Reply

5.195.
The phrase "strict interpretation" in this context simply stands for the proposition that the terms of Article 21.3 have to be interpreted also in light of their object and purpose and in context, which is the entire SCM Agreement and, in particular, Articles 11.6, 11.9 and 15.3 thereof. It is further meant to clarify the proposition that a subsidy which is found to be less than 1 per cent in the original investigation - before the entry into force of the SCM Agreement - would not benefit from a lenient or relaxed interpretation of Article 21.3 so as to permit its unjustified continuation even if it is likely to be below the 1 per cent de minimis rule in the sunset review. This relaxed interpretation is the basic US argument which proposes to read the terms of Article 21.3 in complete isolation of the rest of the Agreement.

Q9. What precise evidentiary standards does the European Communities consider as being required for the initiation of reviews under Article 21.3, particularly in light of the linkages the European Communities sees between Articles 21.3 and 11.6? In the view of the European Communities, is there any difference between the evidentiary standards required for the initiation of original investigations and those required for the initiation of reviews under Article 21.3?

Reply

5.196.
When an investigating authority decides to initiate a review under Article 21.3, it should have sufficient evidence of subsidization, injury and a causal link between those requirements. The level of evidence required in a sunset review is the same or equivalent to the evidence required in an original investigation. Article 21.3 requires the investigating authority to make a determination of likelihood of continuation or recurrence of subsidization and injury. Therefore, before opening the sunset review there should be sufficient evidence of the two aspects which are likely to continue or recur. The final review determination should positively demonstrate both likelihood of subsidization causality and injury.
5.197.
Article 11.6 requires that when an investigating authority self-initiates an original investigation, the standard of evidence is sufficient evidence. Since an original investigation and a sunset review have the same object and purpose, this standard should apply to self-initiated sunset reviews, as it is the practice under US law. Articles 22.1 and 22.7 confirm this proposition.

Q10. Please comment on the relevance, if any, of the retrospective and prospective nature of the assessment to be made, respectively, in original investigations and in reviews under Article 21.3. In particular, how is an assessment of the existence of injurious subsidization similar to and/or different from an assessment of the effects of the revocation of a duty in terms of injurious subsidization? How precisely would the European Communities suggest that the relevant subsidization rate for a review under Article 21.3 should be calculated?

Reply

5.198.
Both types of investigation involve some retrospective and prospective assessment. This is because original investigations and sunset reviews are quite similar by nature and operation. The original investigation must necessarily involve retrospective calculations of the amount of subsidy for a particular period (an investigation period prior to the initiation of the case or the review). The purpose is to determine the amount of benefit granted to a particular company during a specific period of time. Likewise, sunset reviews also involve a retrospective analysis since the subsidization of the past years is often the basis for the sunset finding, especially when the subsidy is non-recurring. The retrospective analysis is consequently extrapolated for the five coming years (prospective nature).
5.199.
In view of the prospective character of the measure which is the outcome of the original investigation and the sunset review, it is essential that a determination is made of the amount of subsidy. According to Article 19.1, an original investigation aims to measure the need for offsetting injury while in a sunset review the purpose is to examine in addition at what level subsidization will recur or continue if the measure were to be removed.
5.200.
However, in a sunset review, as in an original investigation, the subsidy rate should be calculated exactly during the period of review since the investigating authority (particularly for non-recurring subsidies for which the benefit is known) should use this as the basis for establishing the level at which subsidization will continue or recur if the measure were to be removed. The "likelihood" element of the determination relates to the continuation or recurrence of the subsidy and injury, but if the amount found likely to continue or recur, on the basis of the amount found in the review period, is de minimis, the case should be terminated immediately.

Q11. Article 11.6 allows investigating authorities to proceed to initiation without having received a written application by or on behalf of a domestic industry "only if they have sufficient evidence of the existence of a subsidy, injury and causal link, as described in paragraph 2". Article 11.2 of the SCM Agreement sets out, in subparagraph (iii), that this evidence should include "evidence with regard to the existence, amount and nature of the subsidy in question". Could the European Communities comment on the proposition that the CVD order itself, or the results of the most recent review under Article 21.2, is evidence of the existence, amount, and nature of the subsidy in question, consistent with the requirements of Article 11.2(iii)?

Reply

5.201.
The EC would agree that a sufficient amount of evidence on the existence, nature and amount of the subsidy is necessary for initiating a sunset review. However, the initial CVD order, dating from five or more years ago, is not by itself sufficient to meet the burden of proof for initiating a sunset review. There must be the same or equivalent evidence that the original subsidy is likely to continue to exist and to confer a benefit to the current exporters of the product, or there must be evidence of new subsidies.
5.202.
The evidence of a more recent Article 21.2 review may in principle be more relevant, but it will not by itself be sufficient either. Furthermore, such evidence, as applied in the US system, must work both ways. For instance, if a recent administrative review shows that a subsidy amount is de minimis and the subsidy is non-recurring, this should confirm the presumption against initiating, unless compelling evidence of new subsidies is brought forward. Under the present US rules, the US takes no account of such factors when initiating reviews or when continuing CVD measures.

Q12. Please respond to the following:

In the view of the European Communities, is there any similarity between the evidentiary standards required for the self-initiation of reviews under Article 21.2 ("where warranted") and those required for the self-initiation of reviews under Article 21.3?

Reply

5.203.
Subject to the kind of review that is to be carried out under Article 21.2, there is similarity to the extent that, in both cases, the investigating authority would need to be in possession of the same or equivalent amount of sufficient information required from an interested party ("positive information" in the case of a 21.2 review) or the domestic industry ("a duly motivated request" in the case of a 21.3 review). This similarity stems from the fact that in both cases the authorities are required to possess a sufficient amount of evidence in order to initiate review. This is because both types of review are not neutral, in the sense that both may impose costs on the parties involved and disrupt international trade.

(b) Could the European Communities explain its view – expressed in response to an oral question from the Panel at its first meeting – that the requirements for evidence necessary to justify self-initiation under Article 21.2 are lower than the requirements necessary to justify self-initiation under Article 21.3?

Reply

5.204.
This will essentially depend on the type of review carried out under Article 21.2. In 21.2 reviews, the investigating authority, reflecting the burden of proof imposed on interested parties, would need to be convinced that the continued imposition of the duty was no longer necessary ("where warranted" in Article 21.2). Such a conviction may sometimes require a high standard of evidence but may sometimes be based on very simple evidence, for instance, the allocation period for a subsidy may have expired or domestic production in the country imposing the measure may have ceased. In addition, a 21.2 review may be limited to an examination of either subsidy or injury. Therefore, a 21.2 review may, in some circumstances, require a smaller amount of evidence.
5.205.
In contrast, the purpose of sunset reviews under 21.3 is to rebut the presumption of termination of the measure. Such reviews always require a minimum (and rather high) standard of evidence to demonstrate that both subsidy and injury is likely to continue or recur if the measure were to be removed.

(c) Where in the SCM Agreement does the European Communities find the support for lower evidentiary standards applicable to self-initiation under Article 21.2?

Reply

5.206.
See answer above. It will essentially depend on the kind of Article 21.2. reviews under consideration (i.e. administrative review, charge of circumstances reviews, etc.). The term "were warranted" can provide the basis for such a proposition.

(d) How is an assessment of the need for the continued imposition of the duty similar to and/or different from an assessment of the effects of the revocation of a duty in terms of injurious subsidization?

Reply

5.207.
Leaving aside the situations where the authorities initiate the review, the assessment concerning the need for continued imposition (Article 21.2) and the effects of revocation of a duty (Article 21.3) are quite similar. Both examinations focus on the prospective nature of the exercise, i.e. would the removal of the duty lead to continued subsidization and injury. The main difference is the burden of proof. For example, in an Article 21.2 review, the burden of proof is essentially on the interested party (usually a foreign exporter) to justify that there is no need for continued imposition. In an Article 21.3 review, in contrast, the domestic industry must meet the burden of proof to rebut the presumption of expiry of a duty. In both cases, however, the authorities are required to make a positive determination about subsidization, causality and injury.

Q13. The European Communities argues that, "in order to initiate a sunset review on its own initiative, the domestic authorities should be in possession of the same level of evidence that would be required in a 'duly substantiated request' from the domestic industry".67

(a) What does the European Communities understand by the term "duly substantiated request"?

Reply

5.208.
In the EC's view, therefore, a "duly substantiated request" is a request by or on behalf of the domestic industry that contains sufficient evidence that the expiry of the measures would be likely to result in a continuation or recurrence of subsidization and injury, in order to justify a review under Article 21.3. Such a likelihood may, for example, be indicated by evidence of continued subsidization and injury, or evidence that the removal of existing injury is due to the existence of measures. A "duly substantiated" request should normally, inter alia, contain information on:

- the volume and value of domestic production

- the volume and value of the production of petitioners

- volume and value of total imports

- volume and value of subsidized imports

- evidence with regard to the existence, amount and the nature of the subsidy, plus -evidence relating to its continuation or recurrence

- information regarding the effect of imports on prices, production, lost sales

- causality between subsidized imports and injury, and evidence of why such injury will continue or recur.

Although the exact content of a duly substantiated request under Article 21.3 will differ in some ways from the content of an application under Article 11.2, since an Article 21.3 request will relate to a situation where measures are in force and injurious subsidization has already been established, the elements contained in an Article 11.2 should normally be present.

(b) In the view of the European Communities, is there any difference between the evidentiary standards required for the initiation of reviews under Article 21.3 upon request from the domestic industry and those required for the self-initiation of reviews under Article 21.3 by investigating authorities?

Reply

5.209.
The evidence required in a request from the industry and self-initiation should for all practical purposes be the same. The purpose of the evidentiary standard is to guarantee that genuine cases are brought which are backed-up by concrete evidence. If a different standard would apply to self-initiation, this would seriously alter the rights of affected parties and would allow circumvention of these mandatory provisions. This would open the door for situations where the domestic industry does not succeed in creating a prima facie case, but the investigating authority could circumvent it by self-initiating since the evidentiary standard would be different (i.e. lower).
5.210.
The EC’s point is that a correct interpretation of the purpose of Article 21.3 requires investigating authorities to self-initiate reviews only on the basis of sufficient evidentiary support. Where investigating authorities are free to initiate, as it is the case in the US, without any evidentiary support at all, a major distortion of the system as established by the SCM agreement was a parallel between the EC "automatic" notice of impending expiry and the US "automatic" notice of initiation of a CVD measure. The EC duly emphasized the fundamental difference between the "legal effect" attached to these notices: one (the EC) notifies that an expiry is impending, and placing the onus on the domestic industry to produce evidence, without putting any burden on the exporters; the other (the US) actually initiates an investigation without evidence and requires an important and burdensome action from exporters to produce evidence that the measure should not expire. This is a clear reversal of the burden of proof.
5.211.
However, it should be noted that de facto the actual function and scope of the EC notice of impending expiry and the US notice of initiation is the same: to notify the domestic industry the imminent deadline. Actually, the two documents look quite alike and contain the same type of information. The substantial difference is that, under EC law, following a notice of impending expiry the domestic industry must produce a "duly substantiated request" in order to have the sunset review initiated, whereas under US law it is enough for one company to simply manifest its interest. Without such so-called "adequate response" from a domestic producer (which can be a two-line letter), DOC must terminate the sunset review by law.68 This clearly shows that the automatic self-initiation by DOC is simply a device for lowering the level of evidence required from the domestic industry in order to initiate a sunset review. Such is possible only if one admits that an investigating authority can initiate a sunset review automatically, i.e. by definition without any shred of evidence.

Q14. In paragraph 52 of its oral statement at the first meeting of the Panel, the European Communities comments that "a panel must not only revise the evidence submitted to it by the parties, but also all other evidence, including evidence it might consider necessary to get it [sic] itself under Article 13 DSU". Could the European Communities explain this view in light of the recent decision of the Appellate Body in United States – Cotton Yarn69 concerning the evidence a panel may consider in reviewing a determination by a Member?

Reply

5.212.
The EC agrees with the Appellate Body’s findings in paragraphs 62-81 of the report in United States – Cotton Yarn. The Appellate Body has taken particular care to clarify that its findings in that case were limited only "to facts which predate the determination, but which was not in existence at the time the determination was made" by the competent authorities of a Member (at paragraphs 67 and 78). It also clarified that the evidence it was dealing with in that case was "evidence in the form of data that had not been compiled at the time of the determination and, hence, could not have been known." (at footnotes 39 and 51). It follows from the above description that the factual basis of the EC’s claim in the present case is very different from the one examined by the Appellate Body in the Cotton Yarn case. The evidence the EC claims the US authorities should have taken into account is evidence that definitively existed at the time the determination made by the DOC and, moreover, it was well known to the US authorities since they were the actual authors of that evidence (which is also in the form of data on the declining balance methodology and the level of subsidization of the German exporters under the CIG programme). The accuracy of the above is not denied even by the DOC in its final determination in this case nor by the US’s submissions to the Panel so far.
5.213.
The EC also agrees with the summary of the "key elements" the Appellate Body laid down on the standard of review that panels should follow under Article 11 of the DSU (at para. 74). The EC notes that the Appellate Body cited with approval its previous finding that the national authorities "must undertake additional investigative steps, when the circumstances so require, in order to fulfil their obligation to evaluate all relevant factors" (at para. 73). That is why the EC considers that for the purpose of determining whether the US authorities have evaluated all relevant factors and examined all pertinent facts, the Panel can take advantage of the provision of Article 13 of the DSU. It is also noteworthy that in the view of the Appellate Body "in describing the duties of competent authorities, we simultaneously define the duties of panels in reviewing the investigations and determinations carried out by competent authorities" (at para. 73). Furthermore, the EC agrees that the standard of review laid down by the Appellate Body with respect to the Safeguards Agreement are equally applicable to the standard of review to be applied under the SCM Agreement (at para. 76).
5.214.
What is particularly important is the finding by the Appellate Body that "a panel reviewing the due diligence exercised by a Member in making its determination … has to put itself in the place of that Member at the time it makes its determination" (at para. 78). Application of this principle to the facts of the present case would clearly require the present Panel to consider the evidence that existed and was made available to the US DOC but which it refused to take into account on the erroneous allegation that it was submitted outside the time limit. This was not only relevant evidence but evidence indispensable for the proper conduct of the sunset review. The legal consequence of the US DOC’s refusal to consider it should lead to the reversal of the DOC’s findings. The EC does not request the Panel to carry out a de novo review in this case but simply to find that the DOC’s determination is inconsistent with the US obligations under Article 21.3 of the SCM Agreement to determine subsidization, causality and injury, and because it refused to consider highly relevant evidence that existed at the time of the determination and of the existence of which they were fully aware at that time.

Q15. The Panel notes the following US statement: "As a starting point for making its likelihood determination in the sunset review, Commerce considered the countervailable subsidies and programmes used, and the amount of the subsidy determined, in the original investigation. As explained in Commerce's preliminary sunset determination, the rationale for this approach is that the findings in the original investigation provide the only evidence reflecting the behaviour of the respondents without the discipline of countervailing measures in place".70 The Panel further notes the European Communities' arguments contained in paragraph 23 of its oral statement at the first meeting of the Panel. On what legal basis does the European Communities challenge the US Department of Commerce's ("DOC") use of the CVD rate from the original investigation (or a review under Article 21.2), and the rationale for the use of this rate?

Reply

5.215.
The legal basis is Article 21.3, which imposes a clear obligation on the US authorities to make a fresh determination of likelihood of subsidization, causality and injury. The CVD rate from the original investigation may in some cases be an element to take into account in the fresh determination in a sunset review. However, the US law and practice is to use invariably this rate from the original investigation in sunset reviews, even in cases, such as the present one, where exporters and foreign governments have presented concrete and verifiable evidence to show that the subsidy has changed or ceased to exist.
5.216.
The US argument that "…the findings in the original investigation provide the only evidence reflecting the behaviour of the respondents without the discipline of countervailing measures in place" is without merit, particularly in the case of a subsidy recognized by the US DOC to be a non-recurring one. In such a case, in the absence of new subsidies, it is definitively known that the amount of the subsidy will fall and is bound to disappear within the determined period of time in the original investigation.
5.217.
The US DOC systematically refuses to change the original duty rate likely to prevail, even where administrative reviews have shown massive fall in non-recurring subsidies. In such cases, it always recommends the original rate, even if, as in some cases, this rate dates back to 15 years. For example, in the case of Certain carbon steel products from Sweden, for which the expedited review was concluded on 29 March 2000, the US DOC alleged that the Swedish exporters will continue to benefit from two subsidy programmes beyond the end of the review. In light of this position, DOC has decided to confirm the original subsidy rate of 8.77 per cent, which dates from October 1985, even though in the most recent administrative review (covering the year 1997) the subsidy rate was determined to be only 0.72 per cent. Since the subsidies in question are long-term loans allocated over time, the future amounts of benefit are clearly predictable and are not going to increase. It is, therefore, hard to see how the subsidy amount will ever go beyond 0.72 per cent.
5.218.
Similarly, in the case of Corrosion-resistant carbon flat steel products from France, also concluded on 29 March 2000, the DOC has opted for the rate from the original investigation of 15.13 per cent, first imposed in 1993. Although no administrative reviews have been carried out in this case, a recent new countervailing duty investigation71, presumably reflecting the level of subsidy without the discipline of an order in place, found a subsidy rate of 5.38 per cent for the company Usinor, the same exporter as in the above case. Given that the subsidies appear to be untied in both cases, there is no reason for the US DOC not to refer to sheet and strip rate in the context of the sunset review and to report a lower rate to the ITC. Furthermore, the US DOC appears to have much more flexibility when it comes to increasing the rate of a countervailing duty likely to prevail. In the Cut-To-Length Carbon Steel Plate from Belgium case (C-423-806), DOC increased the CVD rate which is likely to prevail from 0.96 per cent in the original investigation (of 1993) to 1.05 per cent in the sunset review. Although one administrative review was conducted by DOC in which the CVD rate was established at 0.69 per cent, DOC did not take these findings into account because it considers that all the subsidy programmes countervailed in the original investigation remained in place. Likewise, in the sunset review in the case of Sugar from the EC (C-408-046), a measure which dates from 1978, the CVD rate was increased from 10.8 cts/pound to 23.69 cts/pound.
5.219.
If a final determination of likelihood of recurrence or continuation of subsidization and injury in a sunset review had to be based exclusively on evidence reflecting the behaviour of the exporters in the original investigation without taking into account the order in force, it is self-evident that a respondent will never succeed in getting rid of a CVD duty on such basis. As a matter of fact this is what happens in the US: DOC has never found until now that there was no likelihood of recurrence or continuation of subsidization in a CVD sunset review against EC exports.
5.220.
There is, therefore, a need to understand what exactly the US means by using the terms "the findings in the original investigation provide the only evidence reflecting the behaviour of the respondents without the discipline of the countervailing measures in place". This reference to the so-called "discipline in place" in fact assimilates CVD orders to antidumping duties. The rationale of granting subsidies and imposing CVD orders are, however, drastically different from those pertaining to dumping and the imposition of AD duties. In the latter case there is one actor, the individual exporter, who can swiftly decide the whole strategy of the company regarding when, where and at what price to export, if the AD duty were to expire. Conversely, in the case of subsidies the situation is not the same, because governments normally do not behave commercially in the same way as individual companies. For instance, in case of non-recurring subsidies, it is more reasonable to assume that governments will not rush again to grant a new subsidy if a CVD order were allowed to expire. The principle is that granting a new subsidy or increasing the level of an existing one is normally subject to rather different considerations (e.g. availability of funds, existence of social, regional, restructuring, environmental or other reasons in favour of continuing the subsidy, etc.), which distinguish them clearly from the natural desire of individual companies to swiftly divert exports in order to increase profits when an AD order is left to expire. Moreover, subsidies take normally time to be discussed and decided in national parliaments and this further distinguishes them from dumping practices. To speak of a "discipline in place", therefore, is to disregard completely the peculiarities and specificities of the reasons for which governments usually grant subsidies, compared to the risk of recurrence of dumping. It follows that as a matter of fact and law a CVD order is not a "discipline" in particular in case of non-recurring, declining subsidies, since such a subsidy has been given in the past, has been consumed for the period of time lapsed and is known that it will not recur in the future. It can, therefore, be said with confidence that in such cases the CVD order has no "disciplining" effect on exporters’ behaviour. It also follows that applying again the rate of subsidy determined in the original investigation is a clear violation of Article 21.3 and the obligation to make a fresh determination on likelihood of subsidization, causality and injury.

Q16. Please respond to the following:

(a) Could the European Communities comment on the difficulty inherent in a review under Article 21.3 of making a forward-looking assessment, given that the data for the previous five years reflect government and exporter behaviour with "the discipline of countervailing measures in place"72?

Reply

5.221.
First, as explained above, the EC contests the US assumption that there is "a discipline in place" which is likely to modify in any significant way the behaviour of governments and of the exporters in case a CVD order were allowed to expire. Second, a determination about the likelihood of continuation or recurrence of subsidization, causality and injury must necessarily be based on projections extrapolating from existing data. There is nothing unusual or especially difficult in this regard because governments are quite used to make this type of determinations in a number of situations. For example, in a different but, mutatis mutandis, comparable situation, the Appellate Body said in United States – Lamb Safeguard report the following:

"As facts, by their very nature, pertain to the present and the past, the occurrence of future events can never be definitively proven by facts. There is, therefore, a tension between a future-oriented "threat" analysis, which, ultimately, calls for a degree of "conjecture" about the likelihood of a future event, and the need for a fact-based determination. Unavoidably, this tension must be resolved through the use of facts from the present and the past to justify the conclusion about the future, namely that serious injury is "clearly imminent".73

Had the US used facts from the present, and if necessary from the past, to justify its conclusions about the future in the present case, there should have been only one possible outcome, i.e. the termination of the CVD order in question because the level of the subsidy would have been clearly below the 1 per cent ad valorem de minimis rule applicable in reviews.74 As explained time and time again, the CIG programme was determined by the US to be a non-recurring, declining subsidy. Applying the US’s own calculation methodology, the amount that would have continued after 2000 is equivalent to zero. No new subsidies have been shown by DOC to exist or were likely to be granted in the period in question. There were, therefore, clearly no present or past facts that could justify a conclusion about the future different from the need to terminate immediately the CVD order applicable to the products in question.

5.222.
The EC does not accept, therefore, the view that a review under Article 21.3 involves difficulties. Governments are used to this type of calculations and there can be no excuse to merely side step any need to undertake a fresh froward-looking determination in a sunset review on the basis of such alleged difficulties. Therefore, the EC considers that, as a general rule, the same elements, which the investigating authorities took into account in the original investigation in a retrospective manner, should be analysed in a prospective way during the sunset review.

(b) What range of factors needs to be taken into account, in the view of the European Communities, in making this forward-looking assessment?

Reply

5.223.
As explained above, the EC considers that, as a general rule, the same elements and factors that the investigating authorities took into account in the original investigation in a retrospective manner should be analysed in a prospective way during the sunset review. These are laid down essentially in Articles 11, 12 and 15 of the SCM Agreement. However, the range of factors to be taken into account in making the forward-looking assessment necessary to establish the likelihood of recurrence or continuation of subsidization has to be evaluated on case by case basis, as it mainly depends on the kind of subsidy under examination. For example, if it is a recurring unemployment subsidy, the investigating authorities should consider the likelihood of lay-offs within the reasonable future. In the case at issue, however, the main subsidy was a non-recurring subsidy that by definition cannot recur. There could have been no continuation of benefits in such a case, if the US authorities were simply to take into consideration the calculation memorandum that was part of the file of the original investigation.
5.224.
As noted above, an aspect to be considered in order to determine the likelihood of recurrence or continuation of subsidization may also be the CVD rate from the original investigation. But there are several other factors and some of them may be more relevant and important ones. For example, consideration should be given to old subsidy programmes now terminated, new subsidy programmes that did not exist at the time the original investigation was made, changes in existing programmes, economic situation of the exporting company, its production, etc. These are some very important factors to be taken into account in a sunset review, together with the expected behaviour of the foreign government. In a sunset review, the forward-looking aspect of the assessment, implied by the terms "would be likely to lead to", refers in particular to the continuation or recurrence of subsidization, causality and injury. Because this involves a certain amount of projections extrapolating from existing data, it may be argued that Article 21.3 somewhat "softens" the requirements in particular of Articles 11.2 and 11.3 of the SCM Agreement relating to the "existence" of subsidization and injury. However, a sunset review determination that is based only on the CVD rate established in the original investigation or one that does not apply the 1 per cent de minimis rule violates clearly the whole object and purpose of the SCM Agreement. The US, instead of conducting a fresh "determination" in the proper sense of this term, advance as excuse reasons pertaining to the difficulties in calculating the numerator or denominator in the formula used to calculate the subsidy. These are not exceptional nor insurmountable difficulties, however, and the US takes no concrete steps and does absolutely nothing positive to deal with them, other than simply acknowledge their existence.

(c) In this regard, the Panel notes the view of the European Communities – expressed in response to an oral question posed by the Panel at its first meeting – that the CVD rate from the original investigation can only be "a starting-point". Please explain in detail.

Reply

5.225.
As explained above, the CVD rate from the original investigation may be just one of the factors to take into account in a sunset review. In this sense only it is a "starting point". For example, it is obvious that in a case of a non-recurring, declining subsidy the original CVD rate cannot be used as such, since the benefits of the subsidy are bound to expire.
5.226.
On a more general plane, it should be noted that the text of Article 21.3 does not state expressly how exactly to conduct a sunset determination. However, there is nothing unusual in that, as other provisions of the same Agreement or of other WTO Agreements do not always laid down in detail provisions about their concrete application and implementation. This is because the applicable rules can be easily inferred. The identification of those rules is normally done on the basis of the customary rules of interpretation of public international law (Art. 3.2 DSU). One such recent example provides the Appellate Body report in United States – Cotton Yarn case, where the Appellate Body recognized the lack of explicit provisions on how to conduct a comparative analysis of the effects of imports under Article 6.4, second sentence, of the ATC Agreement (at paragraph 117), and resolved the issue by having recourse to the generally accepted rules of interpretation in the ensuing paragraphs (paras. 117-121) before turning to the specific provision of Article 6.4 and the facts of that case (at paras. 122-127).
5.227.
This is what the EC proposes the Panel to do also in the present case. As explained above, as well as in the written and oral submissions, a sunset determination has, as a starting point, to take account of the same elements and factors that the investigating authorities took into account in the original investigation. This is, however, obviously not enough in the context of a sunset review for a number of reasons, e.g. because old subsidy programmes may now be terminated, new subsidy programmes may enter into force that did not exist at the time the original investigation was made, changes in existing programmes, economic situation of the exporting company, its production, etc. The non-recurring and declining nature of the subsidy is also a very important factor to take into account. To interpret the text of Article 21.3 otherwise would be to reduce the terms "shall determine" and the whole object and purpose of sunset reviews to inutility, as the Appellate Body said in the United States – Cotton Yarn case (at para. 121), because this will lead to the unjustified continuation of CVD measures for five more years. The US law and practice on sunset reviews provides a concrete example of such a flagrant violation of these basic principles of interpretation, because the US does not apply the 1 per cent de minimis rule in sunset reviews. The result is that a subsidy that was found to be below the 1 per cent de minimis rule in the original investigation will continue for five more years because the US uses in sunset reviews the CVD rate established in the original investigation.

Q17. Please explain what the European Communities understands by the term "total steel sales", contained in paragraph 5 of the European Communities' oral statement at the first meeting of the Panel, and how it applies to the calculation of likely ad valorem subsidy rates.

Reply

5.228.
The term "total steel sales" contained in paragraph 5 of the EC’s oral statement refers to the total sales of the respondent companies in 1991. Pursuant to its regulations, the US DOC calculates an ad valorem subsidy rate for domestic subsidies by dividing the amount of the benefit allocated to the period of investigation by the sales value of all products sold by the respondent during the period of investigation (19 C.F.R. § 351,525(a) & (b)(3)).

Q18. In the view of the European Communities, were the EC respondent parties given "ample opportunity [by the DOC] to present in writing all evidence which they consider[ed] relevant in respect of the investigation in question", within the meaning of Article 12.1:

(a) Prior to the DOC's determination of 20 October 1999 to conduct a full review under Article 21.3;

(b) Following such determination and prior to the DOC's preliminary determination of 27 March 2000; and

(c) Following such determination and prior to the DOC's final determination of 2 August 2000?

Why or why not?

Reply

5.229.
As regards (a) above, the EC respondent parties were not given any opportunity, let alone an ample one, to present in writing "all evidence which they considered relevant in respect of the investigation in question". This is because the US authorities initiated automatically and of their initiative the sunset review. The US argument that the affected respondents had 15 months to prepare (e.g. para. 103 of 1st US written submission) is, therefore, totally misleading, false and irrelevant because the type of comments the respondents could and were allowed to submit before the initiation concerned only "the proposed schedule and grouping of reviews", not the substance of the conditions applicable for a sunset review (see Exhibit US-1, Federal Register/Vol. 63, No. 103/Friday, 29 May 1998/Notices, page 29373).
5.230.
As regards (b) above, the US informed the respondents by letter of 26 August 1999, i.e. four days before the actual initiation of the reviews (1 September 1999) (See Exhibit US-2). This letter provided 15 days to file a "notice of intent to participate" and 30 more days therefrom to file "substantive responses". The rest of the information contained in the letter provided information on the formalities on such filings. The vague nature of the US DOC’s information request, however, coupled with its refusal to accept any factual information after the initial 35 days of a sunset review proceeding violate the provisions of Article 12.1. Unlike in new countervailing duty investigations and administrative reviews, the US DOC does not issue questionnaires in a sunset review. Instead, the US DOC requires the interested parties to respond to a short list of questions contained in its regulations (see 19 C.F.R. § 351,218(d)(3)). Most of these questions are perfunctory in nature, seeking information concerning the name and address of the party’s legal representative and the party’s historic countervailing duty rates (see 19 C.F.R. § 351,218(d)(3)(ii)(A) & (iii)(A)). For example, with respect to the key issue of the likely continuation or recurrence of subsidization, the US DOC regulations simply request the interested party to provide a "statement regarding the likely effects of revocation of the order...under review, which must include any factual information, argument, and reason to support such statement." (see 19 C.F.R. § 351,218(d)(3)(ii)(F)).
5.231.
In addition to this vague request for information, the US DOC regulations provide that "the Secretary normally will not accept or consider any additional information from a party after the time for filing rebuttals has expired." (see 19 C.F.R. § 351,218(d)(4)). Under the sunset regulations, rebuttals are required to be filed within 35 days of the publication of the notice of initiation in the Federal Register (see 19 C.F.R. § 351,218(d)(3) & (4)). Thus, under the US DOC’s regulations, an interested party, guided only by the vague request to provide any factual information regarding the likely effects of revocation of the countervailing duty order, has merely 35 days to provide all of its factual information in the sunset review. Moreover, until the preliminary determination is made it knows completely nothing of the information possibly submitted by the other parties, the domestic industry or taken ex officio into account by DOC. In light of the fact that the sunset review statute provides DOC with 240 days in which to make its sunset review determination, there can be no justification for such a restrictively short period for providing factual information and to respond to the information submitted by the other interested parties (see 19 U.S.C. § 1675(c)(5)). Moreover, the fact that the US DOC’s decision as to whether or not to conduct a full sunset review is not issued until 50 days after publication of the notice of initiation means that, even after the US DOC makes a determination to conduct a full sunset review, the parties will be given no additional opportunity whatsoever to submit factual information in this regard (see 19 C.F.R. § 351,218(e)(1)(ii)(C)(1)). The EC submits that in order to comply with the provisions of Article 12.1, the US DOC must inform the parties of the specific information it requires to complete its determination in the particular case, allow the parties access to the information submitted by the other interested parties and give all the parties ample opportunity to provide the information at a meaningful point in the proceedings. That is not done by the US in sunset reviews under the existing laws, regulations and practices and it was definitively not done in the present case.
5.232.
As regards (c) above, the first time that the US DOC commented on the information submitted by the parties in this case was in its preliminary results of 20 March 2000. In its issues and decision memo, the US DOC ignore the evidence submitted by the Government of Germany and the German producers showing that the CIG programme applied only to investments made prior to 1 January 1986 and that the amount of Capital Investment Grants paid after 1986 was so small that no countervailable benefit would remain after the end of the sunset review (see Issues and Decision Memo at p. 24-25, Exhibit EC-7). The DOC simply stated that, because there was evidence that Capital Investment Grants were received as late as 1990, it would "determine that benefit streams from this programme continue beyond the end of this sunset review." (Ibidem, at 25). It is important to underline that the US DOC did not give the German respondents any opportunity to submit additional evidence to respond to its comments in the preliminary determination. As explained in our written and oral submissions, the US DOC even refused a request to review its own calculation memoranda from the original investigation to determine the exact amount of the original grant payments made after 1986. The EC believes that such a procedure that leaves respondents in complete darkness as to precisely what information an investigating authority requires until it is too late to provide the information cannot satisfy the requirements of Article 12.1. The US law and practice does not respect several provisions of Article 12 of the SCM Agreement, which by virtue of Article 21.4 is fully applicable to sunset reviews. In particular, Articles 12.1 and 12.1.1 lay down a general obligation to provide "ample opportunity" to exporters to present in writing all evidence they consider relevant. Article 12.2 allows also for the right to present information orally. Article 12.3 establishes in effect a procedure of mutual dialogue by providing that the affected exporters should be given timely opportunity "to see all information that is relevant to the presentation of their cases … and to prepare presentations on the basis of this information". Moreover, Article 12.8 provides that before the final determination is made, disclosure should take place "in such sufficient time for the parties to defend their interests". The principle underlying Article 12 of the SCM Agreement, therefore, is that during all stages of the reviews the interested parties should be give ample opportunity to present evidence, to have access to the evidence presented by the other parties (except where confidentiality applies), to present counter-evidence and to defend their interests at all stages leading up to the final determination. As explained above, all these procedural rights and due process requirements are not respected by the US DOC in sunset reviews and in the present case.

Q19. Does the European Communities consider that a review under Article 21.3 must involve a full CVD investigation? Please explain.

Reply

5.233.
In principle yes. As stated above in response to question 7, investigating authorities in a sunset review must determine that all the conditions for the imposition of a countervailing duty, as laid down in Article 19.1, will continue or recur if the duty is terminated. However, since a CVD measure is already in force, this cannot be entirely identical to a new countervailing duty investigation. However, the investigating authority must collect, consider and, where appropriate, verify all available and relevant evidence in making its sunset determination. As explained above, it must also allow parties access to and provide them with ample opportunity to comment and defend their interests at all stages of the review, as happens in an original investigation.
5.234.
The US DOC failed to do this in the present sunset review. For example, with respect to the CIG programme, US DOC ignored evidence that the CIG programme applied only to investments made prior to 1 January 1986 and that the amount of Capital Investment Grants paid after 1986 was so small that no countervailable benefit would remain after the end of the sunset review. The US DOC made no attempt to actually measure the benefit that would remain under the CIG programme after the end of the sunset review or to adjust the subsidy rate for the grants that were made prior to the end of 1986, the benefit of which would have expired under the 15-year allocation period. The US DOC simply continued to apply the 0.39 per cent subsidy rate calculated for this programme for the 1991 period covered by the original investigation, as if the programme were likely to confer the same benefit in 2001 as it did ten years earlier in 1991. The US DOC even refused to consult its calculation memoranda from the original investigation to determine the amount of grants paid after 1986.
5.235.
This irrefutable presumption that the benefit of the CIG programme remained unchanged since 1991 is directly contrary to the evidence presented in the case. Because the CIG programme applied only to investments made prior to 1 January 1986, there is no conceivable way that the benefit in 2001 would continue at the same level as in 1991. This is especially true in light of the evidence that the last grant payment under this programme was made in 1990. Even the US producers conceded that the CIG programme had expired (see Issues and Decision Memo at 25 n.36 (Exhibit EC-7)).

Q20. In the view of the European Communities, does the SCM Agreement require the record of the original investigation, in its entirety, to be made part of the record of the review under Article 21.3? Please explain in detail.

Reply

5.236.
Yes. It is beyond doubt that Article 21.4 requires ("shall apply") the authorities to give parties in a sunset review access to the entire record (that is all information and evidence relevant to an investigation and not covered by the confidentiality provisions) in the same way as in the original investigation. For example, to the extent that the US DOC relies upon the rates calculated in its original investigation, it must, at minimum, make its original determination and all documents showing how the rates therein were calculated (i.e., the calculation memoranda) part of the record of the sunset review. Moreover, Article 12.2 requires that any decision must be based only on information what is in the record and which has been made available to all parties. Article 12.3 requires that all parties should be able to see all information that is relevant, and Article 12.8 requires that all essential facts should be made available in sufficient time for the parties to defend their interests.
5.237.
It follows that the US DOC cannot evade this responsibility by claiming that it is somehow prohibited by US law from making the memoranda part of the record in the sunset review (see U.S. 1st written submission, at para. 104). The US International Trade Commission ("ITC") is subject to the same procedures regarding the treatment of business proprietary information as the US DOC. However, unlike DOC, the US ITC in its sunset reviews automatically makes the confidential version of its original determination, as well as the complete staff report from the original investigation, part of the record of the sunset review. This is done automatically right at the beginning of the review so that all parties will be able to use this information in preparing their arguments. The ITC simply provides the documents the same confidential status in the review that they enjoyed in the original investigation.

F. RESPONSES OF THE UNITED STATES TO QUESTIONS FROM THE PANEL FOLLOWING THE FIRST MEETING OF THE PANEL

5.238.
With respect to the Panel’s question concerning the Article 11.9 de minimis standard and reviews under Article 21.2, for essentially the same reasons set forth in the United States’ First Written Submission and Oral Statement, the de minimis standard contained in Article 11.9 is not applicable to reviews under Article 21.2. In particular, as is the case with respect to Article 21.3, there is no reference to Article 11.9 or any other de minimis standard in Article 21.2. It seems evident that had the drafters of the SCM Agreement wished such a reference, one could have been easily made. Furthermore, the panel in Korea DRAMs addressed this issue (albeit in the context of antidumping administrative reviews) and came to the same conclusion.
5.239.
With respect to the Panel’s question concerning negligible import volume and injury standards under Article 11.9 and the negligible injury standard under Article 15.3, and reviews under Article 21.2, with respect to Article 11.9, as stated above, Article 21.2 neither refers to nor incorporates any provisions of Article 11.9. With respect to Article 15.3, the United States directs the Panel’s attention to the fact that this article does not refer to a "negligible injury standard." Rather, in connection with cumulation, Article 15.3 addresses the question of whether the volume of imports in an original investigation is negligible. Article 21.2 does not incorporate the provisions of Article 15.3.
5.240.
With respect to the Panel’s question concerning negligible import volume and injury standards under Article 11.9 and the negligible injury standard under Article 15.3, and reviews under Article 21.3, the negligibility provisions do not apply to Article 21.3 for the same reasons they do not apply to Article 21.2.
5.241.
With respect to the Panel’s questions concerning the applicability of the Article 27.10 negligible import volume standards and the Article 27.10(b) and Article 27.11 de minimis standards to reviews under Article 21, these negligibility provisions, like those contained in Articles 11.9 and 15.3, apply only to original investigations. Similarly, the de minimis standards for various developing countries, like the standard contained in Article 11.9, only apply to original investigations.
5.242.
With respect to the Panel’s questions concerning the impact of the large volume of "transition" orders on the genesis of the United States’ sunset rules, on January 1, 1995, the date on which the WTO Agreement entered into force with respect to the United States, there were over 300 antidumping and countervailing duty orders in existence. Pursuant to its obligations under Article 32.4 of the SCM Agreement (and Article 18.3.2 of the AD Agreement), the United States deemed all of these "transition" orders to be imposed on January 1, 1995. Consequently, the United States was obligated to initiate sunset reviews of all of these transition orders no later than 1 January 2000, the de jure five-year anniversary date of the orders. In its First Written Submission, the United States described in detail the process it used to determine and publicly announce the schedule for the conduct of the sunset reviews of the transition orders.
5.243.
When developing its procedures for sunset reviews, the United States certainly took into account the monumental task of initiating over 300 sunset reviews of transition orders. However, it is difficult to measure the extent to which the volume of the transition orders initially shaped both the procedural and substantive rules now in place. Procedurally, the rules with respect to sunset reviews of transition orders and those of non-transition orders differ very little. One notable difference that takes into account the volume of transition orders concerns Commerce’s ability to extend its deadlines for its preliminary and final determinations. Substantively, Commerce’s analysis is the same for transition orders and non-transition orders.
5.244.
With respect to the Panel’s question concerning when a sunset review is considered to be initiated, section 351,218(c)(1) of Commerce’s Sunset Regulations provides that no later than 30 days before the fifth anniversary date of an order or suspension of an investigation, the Secretary will publish a notice of initiation of a sunset review. In order to provide to the public advance notice of the initiation of sunset reviews, Commerce provides on its website (http://ia.ita.doc.gov/sunset/schedule.htm) the schedule for sunset review initiations through calendar year 2005. With respect to the sunset reviews of transition orders, Commerce published its initiation schedule in the Federal Register on 28 May 1998.
5.245.
With respect to the Panel’s question concerning Commerce’s adequacy determination and information collected after that determination, following normal procedures, Commerce promulgated its determination to conduct a full sunset review in the form of a decision memorandum issued and made public on October 20, 1999. The specific factual information needed to determine whether to conduct a full sunset review (i.e., aggregate export figures) is only a small part of the information and argument contained in the original substantive responses submitted by the interested parties. Commerce normally does not collect additional information after it makes its adequacy determination. Commerce, therefore, did not collect any additional information between its 20 October 1999, adequacy determination and its 27 March 2000, preliminary determination. Between the time of its adequacy determination and its preliminary determination, Commerce analyzes and considers all of the remaining information and argument provided by the parties; i.e., the bulk of the responses.
5.246.
With respect to the Panel’s questions concerning whether there is a "presumption" in the SCM Agreement that no provision of the Agreement is applicable to reviews under Article 21.3 unless specifically indicated, the United States considers that, in light of the text and context of Article 21.3, no provisions are applicable to reviews under Article 21.3, unless specifically indicated. In the view of the United States, it is not a matter of there being a "presumption". Instead, it is a matter of what the text of Article 21.3 provides, as interpreted in accordance with the rules of Article 31 of the Vienna Convention on the Law of Treaties. There are several ways in which other provisions of the Agreement may be applicable to the provisions of Article 21.3. There could be a cross-reference between the two provisions, a reference in one provision to the other, or a general statement that a provision applies throughout the Agreement or throughout Part V of the Agreement. There are no such references with respect to the Article 11.6 evidentiary requirements for self-initiation or the Article 11.9 de minimis standard.
5.247.
The United States considers that other provisions of the SCM Agreement would apply to reviews under Article 21.3 where the Agreement says they apply. Examples of other provisions that apply to Article 21.3 are: the definition of "subsidy" in Article 1 ("For the purpose of this Agreement"); the definition of "interested parties" in Article 12.9 ("for the purposes of this Agreement"); calculation of the amount of a subsidy under Article 14 ("For the purpose of Part V"); definition of "injury" under Article 15 and footnote 45 ("Under this Agreement"); definition of "like product" under footnote 46 ("Throughout this Agreement"); definition of domestic industry in Article 16 ("For the purposes of this Agreement"); definition of "levy" under footnote 51 ("As used in this Agreement").
5.248.
With respect to the Panel’s questions concerning the methodology for the calculation of the level of subsidization and the ad valorem rate in reviews under Article 21.3 versus in original investigations, the United States notes, as an initial matter, that the SCM Agreement does not specify a methodology for calculating the ad valorem rate. If Commerce were to calculate the level of subsidization in the context of reviews conducted under Article 21.3, it certainly would apply the same calculation methodology as it applies in original investigations conducted under Article 11. However, Commerce does not calculate the level of subsidization in sunset reviews. Article 21.3 does not require such a calculation. What Article 21.3 does require is that a countervailing duty be terminated not later that five years from the date upon which it is imposed, unless the authorities determine that expiry of the duty would be likely to lead to continuation or recurrence of subsidization and injury.
5.249.
With respect to the Panel’s question as to whether the injury test applicable to investigations is also applicable to reviews under Article 21.3, the Article 21.3 injury standard is not the same as the standard for injury in original investigations, although it contains some of the same elements. The injury determinations in original investigations are governed by the provisions of Article 15 of the SCM Agreement and Article VI of GATT 1994. Paragraph 6 of Article VI conditions the levying of countervailing duties on a determination that the effects of the subsidized imports are "such as to cause or threaten material injury to an established domestic industry, or [ ] such as to retard materially the establishment of a domestic industry." Article 15 of the SCM Agreement further specifies the factors that investigating authorities must consider in reaching "[a] determination of injury for purposes of Article VI of GATT 1994".
5.250.
The aim of the Article 21.3 review is to determine whether revocation of the countervailing duty would be likely to lead to continuation or recurrence of injury. Footnote 45 to Article 15 indicates that the term injury as used throughout the Agreement "shall be interpreted in accordance with the provisions of this Article." In turn, Article 15 specifies three general criteria – volume, price effects and impact on the domestic industry – that are pertinent to any injury determination under the Agreement.
5.251.
The focus of a review under Article 21.3, however, differs from that of an original investigation under Article 15. The nature and practicalities of the two types of inquiries demonstrate that the tests for the two cannot be identical. In an original investigation, the investigating authorities examine the condition of an industry that has been exposed to the effects of the subsidized imports. In that investigation, an authority examines the relationship between import-related factors (such as relative and absolute increases in import volumes and underselling and other price effects) to industry-related factors (such as trade, financial and employment data that have a bearing on the state of the industry and that may be indicative of present injury or imminent threat of injury). See Articles 15.5 and 15.7. Five years later, as a result of the countervailing duty order, subsidized imports may have either decreased or exited the market altogether, or if they maintain their presence in the market, may be priced higher than they were during the original investigation, when they were entering the market unencumbered by any additional duties. Thus, the inquiry contemplated in a review conducted pursuant to Article 21.3 is counterfactual in nature, and entails application of a somewhat different standard with respect to the volume, price and relevant industry factors. An authority must decide the likely impact of a prospective change in the status quo, i.e., the revocation of the countervailing duty order and the elimination of its restraining effects on volumes and prices of imports.
5.252.
With respect to the Panel’s question as to whether, under US law, the Article 11.4 industry support provisions are applicable in a review under Article 21.3, the conditions of Article 11.4, with respect to industry support, are not required to be fulfilled in order for Commerce to conduct a sunset review under US law. Article 21.3 itself contains no requirement in this regard and contains no reference to Article 11.4 or the industry support requirements of that provision.
5.253.
With respect to the Panel’s question as to whether the Article 16 definition of domestic industry must be taken into account in the USITC's assessment of the likelihood of continuation or recurrence of injury in a review under Article 21.3, the answer is yes. Article 21.3 addresses the inquiry into whether revocation of the countervailing duty would be likely to lead to continuation or recurrence of injury. Footnote 45 to Article 15 of the SCM Agreement specifies that, "[u]nder 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 industry...". (Emphasis added). Article 21.3 does not contain an exception to the general definition, and therefore, the injury referred to in that Article is relative to the condition of the domestic industry. Article 16 addresses the definition of the domestic industry "[f]or the purposes of this Agreement," and therefore applies in the context of addressing the continuation or recurrence of injury under Article 21.3.
5.254.
With respect to the Panel’s question as to what textual support exists in the SCM Agreement for the proposition that no de minimis standard is applicable to reviews under Article 21.3 as it is in original investigations, under Article 11.9, Members must apply a 1 per cent de minimis standard in countervailing duty investigations. Nothing in the text of Articles 11.9 or 21.3 requires application of the Article 11.9, 1 per cent de minimis standard in Article 21.3 sunset reviews, or any other type of review. As discussed above, there is no reference in Article 21.3 to a de minimis standard and the text of Article 11.9 makes no reference to Article 21.3.
5.255.
With respect to the Panel’s request that the United States respond to arguments in the EC’s Oral Statement concerning the use and relevance of footnote 52, in its Oral Statement, the EC opined that the United States has confused the purposes of an administrative (i.e., assessment) review and a sunset review and the application of footnote 52. It is the EC that is confused. Pursuant to Article 21.3 and footnote 52, the mere existence of a subsidy programme, even with a net countervailable subsidy rate of zero, could form the basis for a determination of likelihood of future subsidization in accordance with Article 21.3 and footnote 52. The United States agrees with the EC that footnote 52 refers to a situation where the authority determines that the subsidy rate for a particular time period is zero and that, in the United States, that determination takes place in the context of an administrative review. (Although not necessarily germane to the instant dispute, the United States does not agree with the EC’s statement that footnote 52 refers to a situation where a subsidy is "de minimis" in an administrative review. Footnote 52 only discusses a finding in the most recent assessment proceeding that "no duty" is to be levied.) The EC seems to think, however, that footnote 52 serves no other purpose than to make a point about administrative reviews. The EC posits that "[s]unset reviews under Article 21.3 are completely different from administrative reviews". If that is so, why then did the Members include footnote 52 in Article 21.3, the provision governing sunset reviews? There must be a reason. The United States considers that footnote 52 means that the current level of subsidization is not decisive as to whether subsidization is likely to recur. The EC has not offered any alternative interpretation. The reason for this gap in the EC’s argumentation is that the EC’s claim that a de minimis standard is required in the context of Article 21.3 sunset reviews would render note 52 meaningless.
5.256.
With respect to the Panel’s question concerning the rationale for the United States’ application of a 0.5 per cent de minimis standard in sunset reviews, as a matter of domestic policy, Commerce has long applied a 0.5 per cent de minimis standard in administrative (i.e., assessment) reviews. The application of this standard pre-dates the Uruguay Round negotiations. The entry into force of the WTO Agreement did not require a change in this standard, because the Article 11.9 de minimis standard is only applicable to investigations. For this same reason, when the United States amended its law in 1994 to provide for sunset reviews, it chose to apply its long-standing 0.5 per cent de minimis standard to sunset reviews. The United States could have chosen to apply no de minimis standard to sunset reviews at all.
5.257.
In a sunset review, the de minimis standard has particular application in several respects. For example, if Commerce determined in a sunset proceeding, based on the original investigation and any administrative reviews, that the existing countervailable subsidy programmes had been terminated and that the likely net countervailable rate of subsidization was de minimis, Commerce normally would determine that there was no likelihood of continuation or recurrence of subsidization. In addition, if the combined benefits of all programmes considered in the sunset review have never been above de minimis at any time the order was in effect, and there is no likelihood that the combined benefits of such programmes would be above de minimis in the event of removal of the duty, Commerce normally would determine that there is no likelihood of continuation or recurrence of subsidization.
5.258.
In 1987, following a notice and comment rulemaking proceeding, Commerce published a final regulation codifying its long-standing practice of applying a 0.5 per cent de minimis standard in investigations and administrative reviews. In response to Commerce stated that "[t]he doctrine of de minimis non curat lex, that the law does not concern itself with trifles, is a basic tenet of Anglo-American jurisprudence, inherent in all US laws. With respect to the antidumping and countervailing duty laws, the Department has concluded that the potential benefits to domestic petitioners from orders on dumping margins or net subsidies below 0.5% are outweighed by the gains in productivity and efficiency provided by a de minimis rule.... [I]t would be unreasonable for the Department and the US Customs Service to squander their scarce resources administering orders for which the dumping margins or the net subsidies are below 0.5%". 52 FR at 30661.
5.259.
With respect to the Panel’s question concerning the relationship between duty assessment proceedings ("administrative reviews") and reviews under Article 21.3, the United States has a "retrospective" assessment system under which the amount of final liability for countervailing duties is determined after the subject merchandise is imported. The normal procedure used for determining the amount of final liability is the administrative review procedure. In contrast, a sunset review is not a procedure for determining the amount of final countervailing duty liability. The purpose of the sunset review is to determine the likelihood of the continuation or recurrence of subsidization in the event that the countervailing duty is revoked. Thus, a sunset review involves a prediction of a government’s future behaviour without the discipline of an order in place. The focus of the analysis is predictive, as opposed to a focus on the present or the past.
5.260.
With respect to the Panel’s question concerning the ability of the German producers, the German government, and the EC Commission to request an administrative review of the countervailing duty order on certain steel flat products from Germany, Commerce’s regulations provide no absolute requirement for shipments as a pre-requisite to an administrative review and provide Commerce with the discretion to conduct a "no shipment" administrative review of an order. Concerning the textual basis for this in the SCM Agreement, as discussed above, Commerce’s regulations do not make the existence of shipments an absolute prerequisite for an administrative review. Instead, the regulations provide Commerce with the discretion to conduct a "no shipment" administrative review of an order. In this regard, Articles 19 and 21.2 of the SCM Agreement do not preclude no shipment administrative reviews.
5.261.
In response to the Panel’s question concerning the option of a changed circumstances review, yes, a "changed circumstances" review is discretionary and may be initiated when Commerce (or the USITC) determines that conditions warranting such a review exist. Such reviews normally are initiated based on a request from an interested party.
5.262.
In response to the Panel’s questions concerning the administrative record from the original investigation, no, the administrative record from Commerce’s original countervailing duty investigation does not automatically become part of the administrative record of the sunset review. Under US law and Commerce regulations, each individual review (whether administrative, sunset, or changed circumstances) by Commerce is considered a separate segment of the proceeding, with a separate and distinct administrative record, and separately reviewable by domestic courts. Pursuant to section 751(c) of the Act and consistent with Article 21.3 of the SCM Agreement, Commerce automatically self-initiates sunset reviews. Therefore, Commerce did not use any information from the original investigation to initiate the sunset review of the order on corrosion-resistant carbon steel flat products from Germany.
5.263.
In response to the Panel’s question concerning a particular calculation memorandum from the original investigation, yes, 13 April 2000 was the first time that the German exporters made a request to have the calculation memorandum placed on the administrative record of the sunset review. In response to the Panel’s questions concerning other requests to place information from the administrative record in the original investigation on the administrative record of the sunset review, with respect to business confidential information, no other interested parties placed, or requested that Commerce place, such information on the administrative record during the sunset review. As discussed in the United States’ First Written Submission, in the instant case, Commerce accepted and considered submissions or parts of submissions from the US producers and the German Government which included public information from the original investigation. Specifically, Commerce accepted a submission from the US producers dated 28 April and portions of a German government submission of 18 April.
5.264.
In accepting the US producers’ submission, Commerce considered that the submission contained the public version of Preussag’s questionnaire response from the original investigation and that the US producers had submitted the document because the German producers had cited to the questionnaire response in one of their submissions prior to the deadline for factual information without submitting the document itself. Commerce also accepted portions of the German Government’s 18 April submission. Commerce, however, only accepted those portions of the German Government’s submission that were part of the original investigation, contained no new factual information, and were publicly available. None of the information accepted by Commerce in this instance was confidential information that would have been unavailable to other parties such as the US producers.
5.265.
With respect to the Panel’s question concerning the use of the word "determine" as used in Article 21.3, Article 21.3 establishes that in the context of the sunset review, Commerce is obligated to determine whether expiry of the countervailing duty would be likely to lead to continuation or recurrence of subsidization. The definition of "determine" in the context of Article 21.3 requires a decision about something. In The New Shorter Oxford English Dictionary, "determine" is defined as to "settle or decide (a dispute, controversy, etc., or a sentence, conclusion, issue, etc.) as a judge or arbiter." Further, this entry contains the notation "foll{owed} by simple obj{ect}, subord{inate} cl{ause} w{ith} that, what, whether, etc." The United States considers that this is precisely the manner in which the word "determine" is used in Article 21.3. Article 21.3 requires the authorities to determine or decide something, i.e., whether subsidization is likely to continue or recur. The United States considers that it may determine, in accordance with Article 21.3, whether subsidization is likely to continue or recur without conducting its own investigation, but, rather, by making its decision based on the evidentiary record developed during the sunset review because all parties, both foreign and domestic, have every opportunity under the US system to provide any information they deem relevant.
5.266.
With respect to the Panel’s question concerning the methodology used by the United States to determine the likelihood of continuation or recurrence of subsidization and injury, the substantive provisions governing likelihood of continuation or recurrence of subsidization are contained in Commerce’s Sunset Policy Bulletin. In determining whether subsidization is likely to continue or recur, Commerce will consider the net countervailable subsidy determined in the investigation and subsequent reviews, and whether any change has occurred in programmes which gave rise to the net countervailable subsidy determined in the investigation and subsequent reviews. The USITC’s regulations, at 19 CFR 207.60-207.69, address the procedures for conducting five-year review investigations to determine whether continuation of the countervailing duty would be likely to lead to continuation or recurrence of material injury.
5.267.
With respect to the Panel’s question as to whether, under the US system, original investigations and reviews under Article 21.3 are different segments of one proceeding, the answer is yes. Under the US system, a "proceeding" begins on the date of the filing of a petition and ends on, inter alia, the revocation of an order. A countervailing duty proceeding consists of one or more "segments". "Segment" refers to a portion of the proceeding that is separately judicially reviewable. For example, a countervailing duty investigation, an administrative review, or a sunset review each would constitute a segment of a proceeding. Each segment contains its own discrete administrative record. A final determination, and the discrete record upon which it is based, is subject to judicial review.
5.268.
With respect to the Panel’s questions concerning whether administrative reviews are prerequisites for conducting sunset reviews, under the US system, administrative reviews are not prerequisites for conducting full sunset reviews. However, as a starting point for making its likelihood determination in a sunset review, Commerce considers the countervailable subsidies and programmes found to be used, and the amount of the subsidy determined, in the original investigation. The rationale for this approach is that the findings in the original investigation provide the only evidence reflecting the behaviour of the respondents without the discipline of countervailing measures in place. This makes sense given that, in a sunset review under Article 21.3, an authority is considering whether, without the discipline of the duty, subsidization would likely continue or recur; i.e., what would happen without the discipline of the duty. Commerce also considers its findings in administrative reviews subsequent to the original investigation because information developed during administrative reviews concerning subsidization – e.g., additional subsidies and accompanying benefits granted after issuance of an order or programme terminations – may be an indicator of possible future subsidization or may demonstrate the cessation of subsidization. Finally, even if there has been no administrative review, Commerce will consider if is evidence demonstrating that programmes have been terminated with no residual benefits. In the instant case, even though no administrative review had been conducted, Commerce agreed with the EC and German producers that two programmes had been terminated with no residual benefits and adjusted the net subsidy rate accordingly.
5.269.
With respect to the Panel’s questions concerning a particular calculation memorandum and the use of this business confidential document in Commerce’s sunset review, on 13 April 2000, the German producers in the sunset review sought to have all the calculation memoranda from the original investigation placed on the record of the sunset review. Commerce, however, could not move business confidential information from the record of one segment of the proceeding (i.e., the investigation) to another separate segment of the proceeding (i.e., the sunset review) without the express permission of the person who submitted the confidential information.
5.270.
Article 12.4 of the SCM Agreement provides that confidential information shall not be disclosed without the specific permission of the party submitting it. Consistent with the obligations concerning the treatment of confidential information under Article 12, US law and Commerce regulations provide stringent requirements and safeguards regarding the disclosure and use of business confidential information in the context of countervailing duty proceedings under what is called an administrative protective order or "APO". The APOs granted during the original 1993 investigation would only have allowed for the disclosure of business confidential information in the context of that investigation, per the agreement of the party submitting the confidential information. As a result, Commerce could not accede to the German producers’ request in the sunset review to move all the calculation memoranda from the record of the original investigation to the record of the sunset review without the permission of the parties who originally submitted the information. The request from the German producers in the sunset review contained no indication of such permission.
5.271.
Under US law and regulations, certain information provided by interested parties in an administrative proceeding, whether an investigation or review, may be accorded business confidential treatment. Section 351,304 of Commerce’s regulations sets forth the requirements for parties to claim that factual information should be considered business proprietary information and afforded protection from public disclosure. The claim for proprietary treatment must be made by the owner of the information, the information must be clearly identified, and the claim must be accompanied by an explanation why the information should be afforded proprietary treatment. The calculation memoranda from the original investigation would have contained the business confidential information of the three German producers of certain corrosion-resistant carbon steel flat products that were involved in the original investigation: Hoesch, Preussag, and Thyssen. These producers would have requested business confidential treatment for their data at the time they submitted the data during the original countervailing duty investigation in 1992-93. The German producers in the sunset review were Thyssen Krupp Stahl AG, Stahlwerke Bremen GmbH, EKO Stahl GmbH, and Salzgitter AG. The request from the German producers in the sunset review to move the business confidential information from the record of the original investigation to the record of the sunset review contains no indication that the German producers in the sunset review were authorized to permit the movement of such information. The particular document submitted by the EC in the instant case appears to contain business confidential information for Thyssen. In its First Written Submission, the EC itself notes that this exhibit contains business confidential information.
5.272.
With respect to the Panel’s questions concerning acceptance and consideration of certain evidence from respondent interested parties, section 351,218(d)(3)(v)(B) of Commerce’s Sunset Regulations provides that interested parties may submit any relevant information or argument that the party would like Commerce to consider. Generally, therefore, Commerce will accept any evidence from foreign respondents, including evidence with respect to the issues set out in the Panel’s question. In the context of a sunset review (or any other segment of a countervailing duty proceeding), Commerce considers all relevant evidence that is timely filed. Regarding the extent to which Commerce might base a particular determination on such evidence, it is difficult to say in the abstract. The relevance and probative value of a particular piece of evidence will vary from case to case. Suffice it to say that in the sunset review at issue in this dispute, Commerce considered information and argument from the EC and German producers in finding that two programmes had been terminated with no residual benefits. As a general proposition, Commerce’s Sunset Policy Bulletin provides detailed guidance on analytical issues related to Commerce’s determination of likelihood of continuation or recurrence of subsidization and the net countervailable subsidy rate likely to prevail if the duty were revoked.
5.273.
With respect to the Panel’s request for a schematic representation of the timing and information requirements under US law for sunset reviews, the United States provided copies of Annex VIII of Commerce’s Sunset Regulations, which contains detailed schedules with timing and information requirements for Commerce sunset reviews, and Annex B of the ITC’s Sunset Regulations, which provides a sample schedule for five-year reviews.
5.274.
With respect to the Panel’s request for certain figures concerning the United States’ initiation and conduct of sunset reviews under Article 21.3, the United States indicated as follows: The number of sunset reviews under Article 21.3 initiated since 1 January 1995 is 56. The number of such reviews which resulted in revocation of the CVD order in question due to no filing by the domestic industry of a notice of intent to participate is 17. The number of expedited sunset reviews conducted since 1 January 1995 is 24. The number of such reviews which resulted in revocation of the CVD order in question, due to a finding of no likelihood of continuation or recurrence of subsidisation is 0, and due to a finding of no likelihood of continuation or recurrence of injury is 5. The number of full sunset reviews conducted since 1 January 1995 is 15. The number of such reviews which resulted in revocation of the CVD order in question, due to a finding of no likelihood of continuation or recurrence of subsidisation is 3, and due to a finding of no likelihood of continuation or recurrence of injury is 4.

G. RESPONSES OF THE UNITED STATES TO QUESTIONS FROM THE EUROPEAN COMMUNITIES FOLLOWING THE FIRST MEETING OF THE PANEL

5.275.
As an introductory matter, the United States observes that the EC in its questions makes various assumptions regarding the "purposes" of SCM Agreement provisions. In this regard, it is important to recall the explanation of the Appellate Body in Japan – Alcoholic Beverages: "the treaty’s ‘object and purpose’ is to be referred to in determining the meaning of the ‘terms of the treaty’ and not as an independent basis for interpretation." In this dispute, the EC purports to find or discern various "purposes" without reference to the text of the SCM Agreement, and then refers to obligations not found in the text which presumably derive from these "purposes". This use of "purposes" is precisely the "independent basis for interpretation" which the Appellate Body described as incorrect, and operates to circumvent the requirement in DSU Article 3.2 that Dispute Settlement Body rulings cannot add to or diminish the rights and obligations provided in the covered agreements.
5.276.
With respect to the EC’s question concerning the meaning of "duly substantiated request", Article 21.3 of the SCM Agreement authorizes authorities to initiate a sunset review "on their own initiative or upon a duly substantiated request made by or on behalf of the domestic industry" (emphasis added). Under US law, Commerce automatically initiates a sunset review on its own initiative within five years of the date of publication of a countervailing duty order. US law does not contain a provision regarding initiation of sunset reviews based upon a "duly substantiated request".
5.277.
With respect to the EC’s question concerning "expression of interest" as applicable to initiation of sunset reviews, section 351,218(d)(1)(i) of Commerce’s regulations provides for filing of a notice of intent to participate by domestic interested parties that intend to participate in a sunset review. Section 351,218(d)(1)(ii) sets forth the required contents of the notice of intent to participate, which includes, inter alia, information related to the identity of the domestic interested party and its legal counsel, whether the domestic interested party is related to foreign producers or exporters, the subject merchandise and country subject to the sunset review, and the applicable Federal Register citation. Section 351,303 of Commerce’s regulations governs the filing, format, translation, service, and certification of documents, and applies to all persons submitting documents to Commerce for consideration in any segment of an antidumping or countervailing duty proceeding, including sunset reviews. All filings, including a notice of intent to participate in a sunset review, must be in writing in accordance with the provisions of section 351,303.
5.278.
With respect to the EC’s question concerning burden of proof in sunset reviews, Article 21.3 does not impose a burden of proof on US authorities. In order to withstand scrutiny under Article 11 of the DSU, however, an Article 21.3 sunset determination must be supported by sufficient evidence and be based on proper legal interpretations. The burden of proof is on the complaining party – in this instance, the EC – to demonstrate that Commerce’s determination was not supported by adequate evidence or proper legal interpretations.
5.279.
Article 21.3 does not contain the word "demonstrate". Instead, Article 21.3 provides for termination of a countervailing duty unless the authorities "determine" that the expiry of the duty would likely lead to continuation or recurrence of subsidization and injury. In the case at issue, Commerce determined that expiry of the countervailing duty on certain corrosion-resistant carbon steel flat products from Germany would be likely to lead to continuation or recurrence of subsidization. Commerce’s determination is supported by adequate evidence and is based on a proper legal interpretation of the applicable provisions.
5.280.
With respect to the EC’s questions concerning the purpose and consequence of an investigation versus a sunset review, the purpose of a countervailing duty investigation is to determine the existence and degree of foreign government subsidization and whether the subsidized imports are causing injury. In contrast, the purpose of a sunset review is to determine whether revocation of a countervailing duty order would be likely to lead to continuation or recurrence of subsidization and injury. The consequence of an affirmative finding of subsidization and injury in an investigation is the imposition of a countervailing duty. The consequence of an affirmative finding of likelihood of continuation or recurrence of subsidization and injury in a sunset review is the continued maintenance of the countervailing duty. The completion of a sunset review does not trigger the assessment of duties or change cash deposit requirements.
5.281.
With respect to the EC’s questions concerning preliminary determinations under US system, the purpose of Commerce’s preliminary determination in a full sunset review is to provide an explanation of Commerce’s findings concerning the likelihood of continuation or recurrence of subsidization and the net countervailable subsidy likely to prevail if the countervailing duty were revoked. Commerce’s preliminary determination takes into account the factual information and arguments submitted by the parties in their substantive responses and rebuttals.
5.282.
Section 351,218(d)(3)(i) of Commerce’s Sunset Regulations provide that substantive responses to a notice of initiation are due 30 days after the date of publication in the Federal Register of the notice of initiation. Rebuttal to a substantive response is due five days after the date the substantive response is filed. (Section 351,218(d)(4).) Section 351,218(d)(4) of the Sunset Regulations also provide that Commerce normally will not accept or consider any additional information from a party after the time for filing rebuttals has expired. Section 351,302(c), however, provides that a party may request an extension of a specific time limit. Unless expressly precluded by statute, Commerce may, for good cause, extend any time limit established by its regulations. (Section 351,302(b).) The US countervailing duty statute does not contain deadlines for submission of information in a sunset review.
5.283.
The preliminary determination provides interested parties with an opportunity to review Commerce’s analysis of the information on the record. Commerce issues preliminary determinations in investigations and administrative reviews as well. Preliminary determinations provide interested parties with an opportunity to comment, in case and rebuttal briefs, on Commerce’s preliminary findings. Commerce is not precluded from requesting factual information after the issuance of a preliminary determination, but normally it does not do so.
5.284.
With respect to the EC’s questions concerning data collection periods, section 351,301(c)(2)(iii) of Commerce’s regulations provides that interested parties will have at least 30 days from the date of receipt to respond to a questionnaire. This rule applies to original countervailing duty investigations. Similarly, section 351,218(d)(3)(i) provides for a 30-day deadline with respect to a response to the sunset questionnaire. Furthermore, as discussed above, a party may request an extension of a specific time limit.
5.285.
With respect to the EC’s question concerning Article 12, the United States agrees that Article 21.4 of the SCM Agreement expressly provides that the provisions of Article 12 – as opposed to the provisions of Article 11 – apply to these types of reviews.
5.286.
With respect to the EC’s questions concerning Commerce’s application of a 0.5 per cent de minimis standard in sunset reviews, as well as its questions concerning the subsidy rate Commerce reports to the USITC, because Article 21 of the SCM Agreement does not contain any de minimis standard, these questions is not relevant to the issues raised in this dispute. Nevertheless, the United States notes that, as a matter of domestic policy, Commerce has applied a 0.5 per cent de minimis standard in administrative (i.e., assessment) reviews. The application of this standard pre-dates the Uruguay Round negotiations. The entry into force of the WTO Agreement did not require a change in this standard, because the Article 11.9 de minimis standard is only applicable to investigations. For this same reason, when the United States amended its law in 1994 to provide for sunset reviews, it chose to apply its long-standing 0.5 per cent de minimis standard to sunset reviews. The United States could have chosen to apply no de minimis standard to sunset reviews at all.
5.287.
Commerce’s de minimis standard in reviews is different from its de minimis standard in investigations. Prior to the entry into force of the WTO Agreement, Commerce applied a 0.5 per cent de minimis standard in investigations. However, in order to conform to Article 11.9 of the SCM Agreement, Congress amended the US statute so as to require the use of a 1 per cent de minimis standard in investigations.
5.288.
In a sunset review, the de minimis standard has particular application in several respects. For example, if Commerce determined in a sunset proceeding, based on the original investigation and any administrative reviews, that the existing countervailable subsidy programmes had been terminated and that the likely net countervailable rate of subsidization was de minimis, Commerce normally would determine that there was no likelihood of continuation or recurrence of subsidization.
5.289.
In addition, the Sunset Policy Bulletin (section III.A.6.b) provides that, if the combined benefits of all programmes considered in the sunset review have never been above de minimis at any time the order was in effect, and there is no likelihood that the combined benefits of such programmes would be above de minimis in the event of removal of the duty, Commerce normally would determine that there is no likelihood of continuation or recurrence of subsidization.
5.290.
In 1987, following a notice and comment rulemaking proceeding, Commerce published a final regulation codifying its long-standing practice of applying a 0.5 per cent de minimis standard in investigations and administrative reviews. 52 FR 30660 (17 August 1987). Pursuant to the regulation, net aggregate subsidies (and ad valorem dumping margins) of less than 0.5 per cent would be disregarded for purposes of publishing or revoking orders, setting cash deposit rates, or assessing countervailing duties. In response to comments regarding Commerce’s decision to set 0.5 per cent as the de minimis threshold, Commerce stated as follows:

The doctrine of de minimis non curat lex, that the law does not concern itself with trifles, is a basic tenet of Anglo-American jurisprudence, inherent in all US laws. With respect to the antidumping and countervailing duty laws, the Department has concluded that the potential benefits to domestic petitioners from orders on dumping margins or net subsidies below 0.5% are outweighed by the gains in productivity and efficiency provided by a de minimis rule. Even in price-sensitive markets, the effect of requiring a deposit or assessment of duty based on a rate of 0.5% ad valorem would be negligible. No party submitting comments has provided any information to support a different conclusion. Accordingly, it would be unreasonable for the Department and the US Customs Service to squander their scarce resources administering orders for which the dumping margins or the net subsidies are below 0.5%. The fact that the Department of Treasury and Commerce may not always have applied a uniform de minimis standard in the past is an additional reason supporting the adoption of a fixed standard which can be applied consistently in the future.

52 FR at 30661. In response to comments that the de minimis threshold be set at 1 per cent, Commerce also stated that,

After many years of applying a 0.5 per cent de minimis threshold, the Department has developed no basis to conclude that 1 per cent represents a level of benefit not worth the expense of investigations or annual reviews....

5.291.
As explained in the United States First Written Submission, Commerce starts with the total ad valorem rate determined in the original investigation and considers whether, since the investigation, it has found subsidy programmes to be terminated and/or new programmes to be countervailable. Based on findings, which normally are made in the context of administrative reviews, Commerce may adjust the rate determined in the original investigation to take these subsequent findings into account.
5.292.
With respect to the EC’s question concerning the amount of subsidy in the context of a sunset review, Article 21.3 provides for consideration of whether expiry of a countervailing duty would likely lead to a continuation or recurrence of subsidization and injury. Footnote 52 states that a finding, in the most recent administrative review, that no duty is to be levied shall not by itself require the authorities to terminate the definitive duty. This is consistent with the prospective nature of a sunset review. Thus, the level of subsidization at the time of a sunset review is not necessarily determinative of the outcome of a sunset review. However, if Commerce determined in a particular sunset review that there was no likelihood that the net countervailable subsidy rate would exceed 0.5 per cent ad valorem in the event of revocation of the countervailing duty, Commerce would determine that there was no likelihood of continuation or recurrence of subsidization and revoke the duty.
5.293.
With respect to the EC’s question concerning use of the 1 per cent, Article 11.9 de minimis standard, for the reasons set forth in the United States’ First Written Submission and Oral Statement, the Article 11.9 de minimis standard does not apply to Article 21.3 sunset reviews. Article 21.3 does not contain any de minimis standard. Commerce does apply a 0.5 per cent de minimis standard in sunset reviews, but does not do so because of any obligation imposed by the SCM Agreement. In the final results of the full sunset review of corrosion resistant steel from Germany, Commerce determined that revocation of the order would be likely to lead to continuation or recurrence of a countervailable subsidy and found that the net countervailable subsidy likely to prevail if the order was revoked was 0.54 per cent ad valorem, which is above the de minimis standard under US law.
5.294.
With respect to the EC’s questions concerning calculation of a current rate of subsidization in a sunset review, Commerce normally does not determine the present net countervailable subsidy rate in a sunset review, and it did not calculate the level of subsidization present at the time of the sunset review in the corrosion-resistant steel from Germany case. Commerce did, however, adjust the net countervailable subsidy rate determined in the investigation to account for two programmes that the EC and German producers argued had been terminated with no continuing benefits. Nevertheless, nothing in Article 21.3 or any other provision of the SCM Agreement mandates a particular methodology for determining whether subsidization is likely to continue or recur if the duty were revoked.
5.295.
The focus of a sunset review is necessarily on the possible future behaviour of foreign governments and exporters. The best evidence of that behaviour is the net countervailable subsidy rate determined in the original investigation because it is indicative of behaviour without the discipline of a countervailing duty in place. Starting with the net countervailable subsidy rate from the original investigation, Commerce may make adjustments, including adjustments for subsidies for which benefits were allocated over time, in accordance with the Sunset Policy Bulletin. (See section III.B.3.) Commerce, however, normally makes such determinations in the context of an annual administrative review where interested parties may submit, inter alia, information concerning the level of subsidization present during the period of review for Commerce’s examination.
5.296.
With respect to the EC’s question regarding Commerce’s declining balance methodology, as explained in the United States’ First Written Submission, the ad valorem subsidy rate for any period cannot be determined without knowing the applicable sales volume.
5.297.
With respect to the EC’s questions concerning the rationale for the 1 per cent de minimis standard, the only de minimis standard found in the SCM Agreement is in Article 11.9 which, by its terms, is limited to investigations. Neither Article 11.9 nor any other provision of the SCM Agreement provides a rationale for the de minimis standard. This standard is a product of negotiations. As discussed in more detail above, when Commerce codified its own de minimis standard in its regulations, it considered several rationales, including administrative efficiency, as providing a basis for the standard.
5.298.
With respect to the EC’s questions concerning new countervailable subsidies granted since 1993, as discussed in Commerce’s preliminary sunset determination, there were no administrative reviews of this order. As a result, Commerce had not considered whether German producers benefitted from additional subsidies granted since the original investigation. Commerce did not consider domestic interested parties’ allegations concerning new subsidies in the context of the sunset review because US law intends that such allegations should normally be made in the context of administrative reviews and the lack of any administrative reviews, in and of itself, was not sufficient to constitute good cause to consider the petitioners’ allegations in the context of the sunset review. Furthermore, once Commerce found likelihood based on previously investigated subsidies, a finding of additional new subsidies would not have changed its affirmative likelihood determination.
5.299.
With respect to the EC’s question concerning the application of footnote 52, footnote 52 stands for the proposition that an existing subsidy programme could be the basis for a determination in a sunset review that the expiry of the countervailing duty would likely lead to the continuation of subsidization even if Commerce found a net countervailable subsidy rate of zero attributable to that programme in the most recent administrative review. In other words, footnote 52 means that the current level of subsidization is not decisive as to whether subsidization is likely to recur.
5.300.
With respect to the EC’s questions concerning Commerce’s findings in other sunset determinations, the dispute before this Panel involves Commerce’s sunset determination concerning certain corrosion-resistant carbon steel flat products from Germany. As a general matter, Commerce sunset determinations involving other merchandise and other countries are available as public, published documents and can be found on Commerce’s website, the website of the US Government Printing Office, or through commercial database services such as Lexis. We note, however, that to date, Commerce has found no likelihood of continuation or recurrence of subsidization in the following three full sunset reviews: C-122-404, Live Swine from Canada, 64 Fed. Reg. 60301 (November 4, 1999); C-333-401, Cotton Shop Towels from Peru, 64 Fed. Reg. 66884 (November 30, 1999); and C-201-505, POS Cookware from Mexico, 65 FR 284, (January 4, 2000).
5.301.
With respect to the EC’s question concerning the effect of the net subsidy rate on volume of exports, issues concerning the volume of exports to the United States are properly considered in the context of the USITC’s injury determination, which has not been challenged by the EC in this dispute. Nevertheless, the United States would point out that the ad valorem countervailing duty rate determined in the investigation was 0.60 per cent and, according to the EC, the German producers stopped shipping to the United States once this 0.60 per cent countervailing duty was in place.
5.302.
With respect to the EC’s questions concerning amounts for certain denominators or numerators, Commerce’s finding of likelihood of continuation or recurrence of subsidization is based on and supported by the administrative record of the sunset review. The EC’s questions are not really questions. Rather, the EC is asking the United States to research and perform new calculations in the context of the EC’s challenge to the United States’ laws and regulations concerning sunset reviews and to a particular Commerce sunset determination. The United States does not believe it is appropriate to do so in this context.

H. COMMENTS OF THE EUROPEAN COMMUNITIES ON THE RESPONSES OF THE UNITED STATES TO QUESTIONS FROM THE PANEL FOLLOWING THE FIRST MEETING OF THE PANEL

Introduction
5.303.
The US points (para. 2) that the EC is not challenging the injury determination in this proceeding. However, the EC in this proceeding is not challenging the injury requirement only for the purpose of verifying the accuracy of the ITC injury determination as such, but it is doing so for the purpose of the de minimis challenge, i.e. as a requirement that needs to be demonstrated positively in the context of a sunset review under Article 21.3 of the SCM Agreement (see First Oral Statement of EC, footnote 27).

Question 1(a)

5.304.
Throughout its responses to the questions from the Panel, the US espouses a formalistic interpretation of Article 21.3 SCM Agreement that would essentially strip Articles 21.1 and 21.3 of any practical effect. An example of this is the US’s statement in paragraph 5 that Article 21.3 contains "no requirement to quantify the amount of subsidization likely to continue or recur". Article 21.3 permits the continuation of a countervailing duty order only if the authorities affirmatively determine that "the expiry of the duty would be likely to lead to continuation or recurrence of subsidization and injury". In turn, Article 15.5 defines "injury" as that injury caused by subsidized imports "through the effects of subsidies". Thus, an authority can only make a proper determination concerning the continuation or recurrence of injury if it considers the amount of subsidization that is likely to continue or recur if the duties were to expire. Under the US system, this means that DOC must provide the ITC with an accurate calculation of the amount of subsidization that is likely to continue or recur if the countervailing duty order were revoked. The US DOC has failed to do that in this case but simply continued to apply the countervailing duty rate calculated for 1991 despite the clear and indisputable evidence provided by the Government of Germany and the German producers establishing that this rate was no longer accurate.
5.305.
As regards the references to paragraph 6.87 of the Korea DRAMS panel report, the US draws the wrong conclusions. In that report, the Panel had not dealt with a sunset review in the sense of Article 11.3 of the AD Agreement, but a claim for a duty assessment review under Article 9.3 of the AD Agreement. But there is no exact equivalent of Article 9.3 AD Agreement in the SCM agreement. Such duty assessment reviews fall by default within the scope of Article 21.2 of the SCM Agreement. In addition, and more importantly, the basic reason for which that Panel interpreted Articles 5.8 and 9.3 of the AD Agreement in the way it did is stated in paragraph 6.90 of the report, to which the US omits to make any reference. There the Panel held that: "In the context of Article 9.3 duty assessment procedures, however, the function of any de minimis test applied by Members is to determine whether or not an exporter should pay a duty. A de minimis test in the context of an Article 9.3 duty assessment procedure will not remove an exporter from the scope of the order." (emphasis added) This is clearly not the case of a CVD sunset review determination, because the basic principle laid down by Article 21.3 and 21.1 of the SCM Agreement is the termination of the CVD order unless a fresh, positive finding about likelihood of subsidization, injury and causality is made by the authorities.

Questions 1(b) and (c)

5.306.
Contrary to the US responses, the US ITC does consider negligibility in its sunset review determinations. Under US law, cumulation is specifically prohibited in a sunset review if "imports are likely to have no discernable adverse impact on the domestic industry". 19 U.S.C. § 1675a(a)(7). This provision was added to US law by the Uruguay Round Agreements Act. In its report to that Act, the US Senate specifically stated that this cumulation provision was being added because Congress "believe[d] that it is appropriate to preclude cumulation where imports are likely to be negligible". S. Rep. No. 103-412, at 51 (1994). Similarly, the US ITC also considers differences in the "conditions of competition" in making its cumulation decision in a sunset review. SeeSugar from the European Union, USITC Pub. No. 3238 at 15 (1999) ftp://ftp.usitc.gov/pub/reports/ opinions/PUB3238.pdf>.
5.307.
Moreover, for the reasons explained in the comment to the previous question, the combined effect of paragraphs 6.43-6.6.50 of the Korea DRAMS panel report supports the proposition that the standards of Article 11.9 and 15.3 should apply to Article 21.2 and to Article 21.3 reviews.

Questions 1(d)

5.308.
Acceptance of the US position would render the provisions of Article 27, paragraphs 10 to 12, useless once the initial CVD order is in place. This position is totally unreasonable when proper consideration is given to the context, object, purpose and scope of special rules for developing country members laid down in the SCM and, more generally, the WTO Agreements.

Question 21

5.309.
As the US readily admits, its rules apply with equal force to both "transition orders" and "non-transition" orders. Article 21.3 also makes no distinction between transition and non-transition orders and does not in any way permit an authority to apply a lower level of review or procedure with respect to transition orders.

Question 23

5.310.
The US’s response to this question reveals how its system of automatic self-initiation results in the conduct of a sunset review without the requirement that the domestic industry ever provide evidence substantiating the likelihood of the continuation or recurrence of subsidization. In paragraph 18, the US concedes that the "only specific factual information" that it considers in making the decision as to whether to conduct a full sunset review is the "aggregate export figures provided by respondent interested parties". The other factual information submitted by the parties does not play any role in DOC’s adequacy determination and is not discussed by DOC until its preliminary determination, which occurs after the time for submitting additional factual information has long expired (See EC Second Written Submission at paras. 41-42).

Question 24

5.311.
The US arguments omit to mention another generally accepted principle of interpretation that "omissions in different contexts may have different meanings, and omission, in and of itself, is not necessarily dispositive". This is what the Appellate body has held in several cases. For example, in the Canada – Automotive Industry report, the Appellate Body held that the panel in that case had erred because it failed to examine the ordinary meaning of the terms in their relevant context and in light of the object and purpose of the entire SCM Agreement in order to identify the proper meaning of an omission in the text.75 Another example provides the US reasoning in paragraph 48 of its response to question no 32(b) from the Panel, where it states that "Articles 19 and 21.2 of the SCM Agreement do not preclude no shipment administrative reviews". Without passing judgment on the substance of this US claim, what is important to note here is the reasoning by inference resorted to by the US, which is an accepted method of interpretation under the generally principles of public international law on treaty interpretation.

Question 25

5.312.
In its response, the US concedes that it does not calculate the level of subsidization in a sunset review but simply reports to the US ITC the rate that was already calculated in the original investigation (at paras. 23-24). This in turn taints the US ITC injury determination because it is not based upon the level of subsidization that would be likely to continue or recur if the countervailing duties were terminated as required by Article 21.3. While the US contends that the rate from the original investigation is the "only calculated rate that reflects the behaviour of exporters and foreign governments without the discipline of an order in place", this reliance upon the rates from the original investigation cannot be applied as an irrefutable presumption. To do so would render the provisions of Articles 21.1 and 21.3, requiring the termination of countervailing duties after five years, meaningless because the fact that countervailing duties were originally imposed would be used as the justification for keeping those duties in place indefinitely.
5.313.
The US’s discussion (at para. 25) demonstrates a blatant violation of Article 12 SCM Agreement as made applicable to sunset reviews by virtue of Article 21.4. In paragraph 25 of its replies, the US states that, in making its determination in a sunset review, it "only uses information developed in the original investigation or prior administrative proceedings" (emphasis added). The US tries to justify this practice by stating that "this information has been subject to the rigors of the administrative process in those proceedings, such as interested party briefing and onsite verification". In our view, what the US fails to discuss is that reviews under Article 21.3 are, by virtue of Article 21.4, subject to the same evidentiary and procedural requirements as investigations and administrative reviews, including the ability to conduct onsite verifications.
5.314.
Moreover, DOC’s stated preference for information developed in the original investigation cannot justify its refusal to adjust the subsidy determination for changes in the CIG programme. All of the information submitted by the Government of Germany and the German producers, showing the termination of the CIG programme with no significant residual benefit after 2000, comes directly from the original investigation and was subject to full briefing and verification in that case. The conditions and requirements of Article 12 are fully applicable to sunset reviews and no distinction is made therein whether the information is collected in the original investigation or in an Article 21.2 or an Article 21.3 review. Most importantly, what the US terms "a relationship of …convenience" (at para. 25) under its domestic law cannot derogate from the basic principles of Articles 21.3 and 21.4 of the SCM Agreement to make a fresh, proper determination of likelihood of subsidization, injury and causality on the basis of positive evidence.

Question 26

5.315.
While the United States concedes that the definition of injury contained in Article 15 applies throughout the SCM Agreement, it reads (at paras. 26-29) unduly restrictively Article 15, by applying only those provisions that it finds favourable. By doing so, however, it seems to contradict its previous replies to question 24 (a) and (b) from the Panel (at paras 19-21). Indeed, the reference in Footnote no 45 to "threat of material injury"76 appears already to accommodate, albeit in a different context, the prospective nature of the analysis claimed by the US under the "likelihood" test of an Article 21.3 sunset review. Moreover, as stated in the comment to question 1(a) above, the definition of injury must include the element of causation. Article 15.3 also makes clear that the de minimis and negligibility provisions of Article 11.9 are an integral part to the definition of injury. As the Korea DRAMS panel has held, in the comparable situation of an Article 11.2 of the AD Agreement review, "while mathematical certainty of recurrence of dumping is not required, the conclusions must still be demonstrable on the basis of the evidence adduced" (at paras. 6.43 and 6.50 of the report). The obligation to adduce such "evidence" in a CVD sunset review is squarely place, by virtue of Article 21.4, on the authorities undertaking the review for the purpose of making a "determination" under 21.3.

Question 27

5.316.
The US reply to question no 27(a) is another illustration of the isolationist interpretation of Article 21.3 advanced by it. There is indeed no valid reason to interpret the term "domestic industry" in Article 21.3 differently from that laid down in Article 11.4 in conjunction with Article 16 of the SCM Agreement. It follows that a request made by "at least one domestic interested party" is not a "duly" made request under Article 21.3 because it would not comply with the initiation requirements of Article 11.4.

Question 28

5.317.
Here the US appears to be placing undue importance on the use of the word "investigation" in Article 11.9, as opposed to the use of the word "review" in Article 21.3. The same undue reliance on the word "investigation" is shown by the US in its reply to question 1(d) above concerning Article 27.10. Although it is true that Article 11 lays down provisions on "initiation and subsequent investigation", the term "investigation" denotes the "process or an instance of investigating" or "a formal examination or study".77 It follows that although this term encompasses a substantive exercise relating to this process of fact finding, it is more accurate to say that its scope is narrower than the scope of the term "determine" as this is used in a number of Articles in the Agreement, such as in Articles 19 or 21.3. Indeed, an investigation may lead to the imposition and collection of countervailing duties when the Member makes a final "determination" under the conditions laid down in Article 19. Therefore, what is particularly important in Article 11.9 is not the use of the term "investigation", as the US claims, but rather the use of the terms "the authorities concerned are satisfied that"78, which is the equivalent to the concept of making a "determination" used in Article 19 or Articles 21.3 of the SCM Agreement. In other words, Article 11.9 comports an element of substantive determination, which, although made in the process of the original investigation, is not different from the substantive point of view of the finding that needs to be made in a sunset determination about likelihood of subsidization, injury and causality. What is significant, therefore, is the substantive determination that should be made by the authorities about the level of subsidization under both Articles 11.9 and 21.3, which must in both cases entail a determination on de minimis level of the subsidy, injury and causal link rather than the procedural process or stage at which this occurs. In other words, a determination in a sunset review requires the conduct of an investigation as much as an investigation is required prior to the original determination. Therefore, the fact that the word "investigation" does not appear in Article 21.3 makes no difference. Investigating authorities are not absolved of their obligation to investigate from the day after an original measure is imposed. This would amount to saying that authorities are entitled to apply a lower standard of examination to issues raised in reviews, compared to that which applies in original investigations. It follows that the US insistence only on the word "investigation" in Article 11.9 or 27.10 of the SCM Agreement, to exclude the application of the de minimis rule in sunset reviews, amounts to simply reading these provisions outside their context, object and purpose.

Questions 29 and 30

5.318.
The US claims (at paras. 34-36) that the interpretation proposed by the EC would render Footnote 52 a nullity and asserts that there must be a reason for its inclusion in Article 21.3. This answer is rather puzzling. The EC does not dispute that "a finding in the most recent assessment procedure that no duty is to be levied shall not by itself require the authorities to terminate the definitive duty". Such a finding may be based on a finding on the zero subsidy or on some other reason. But Footnote 52 does not support the conclusion the US attempts to draw from it.
5.319.
The US argument is based on an understanding of the concept of de minimis in Article 11.9 that reflects the Latin proverb "de minimis non curat lex". As the US has explained in several places (e.g. in its reply to question 30 from the Panel and its reply to question 8 from the EC, respectively at paras. 41 and 16 of its replies), it applies voluntarily the de minimis rule of 0.50 per cent under domestic law for reasons of administrative efficiency because the US law is not supposed to "concern itself with trifles".
5.320.
The preparatory history of Article 11.9 and the corresponding provision in the AD Agreement, however, indicate that the rationale of the de minimis rule in these Agreements is different from that on which the de minimis rule is applied in the US. In these two WTO Agreements, the level of 1 per cent ad valorem was agreed to be the de minimis threshold not for reasons of administrative efficiency but because for imports below this level the causal link between subsidized imports and the material injury to domestic industry would not exist.79 This is a significantly different rationale, because it demonstrates that the US continued under its domestic law its old practice going back to 1980, as codified in the 1987 Guideline on the minimis rule (see Exhibit US-6). Based on that understanding of the de minimis rule, the US then extrapolates from Footnote 52 by making several illogical leaps: First, it argues that if a zero rate of subsidy cannot require the authorities to terminate the CVD duty, that means that the level of the subsidy, even if zero, is not relevant for deciding likelihood of continuation or recurrence of subsidization, injury and causality. Second, it draws the operational conclusion that it is not necessary to calculate the actual level of subsidy in a sunset review. Third, it does consider that the de minimis rule of Article 11.9 applies to Article 21.3 reviews.
5.321.
The EC considers that the current level of subsidy is not always "determinative" of the outcome of a sunset review. However, the "prospective nature of a sunset review" is subject to the requirement for the authorities to positively determine that subsidization is likely to continue or recur. Thus, if the current level of subsidy is zero or de-minimis, this does not by itself end the measure, but it does require the US DOC to demonstrate what change of circumstances will lead to the subsidy increasing to a level above de-minimis, if it wishes to continue the measure. In the case of a non-recurring subsidy which is already de-minimis and subject to the US declining balance methodology, the DOC must adduce and rely on positive evidence of new subsidies or of a sharp fall in the denominator in order to find likelihood of subsidization, injury and causality.
5.322.
Moreover, the US states that the "focus" of a sunset review is on the possible future behaviour of foreign governments and exporters. This is misleading. First, we do not see in what respect the behaviour of the exports would be relevant in this context. Second, a sunset review determination requires positive evidence to demonstrate that subsidization, injury and causality is likely to continue or recur. The review should therefore focus on "demonstrable evidence adduced by the authorities" rather than on pure speculation and conjecture derived by applying the level of subsidization from the original investigation. The US states that the "best evidence" of future behaviour of governments and exporters is the rate from the original investigation. However, in the case of Corrosion-resistant steel from Germany, how can this be? The DOC has itself established that there are no new subsidies and that the non-recurring subsidies found in the original investigation, 9 years ago, have now declined to a level that is equivalent zero by its own calculation methodology.
5.323.
There appears therefore to be only one reasonable conclusion: the US has modified its past practice after the WTO Agreements on AD and SCM entered into force only as regards determinations made in the initial investigation, but left its domestic legislation on sunset reviews unchanged by applying its 0.50 per cent de minimis rule based on a completely different rationale (i.e. administrative convenience and efficiency).

Question 30

5.324.
In addition to what has been explained above, the US’s response to question 30 again reveals its refusal to provide parties to a sunset review with the evidentiary and procedural rights established in Article 12. In paragraph 39, the US reiterates that its determination in a sunset review is "based on the original investigation and any administrative reviews". It states:

"In a sunset review, the de minimis standard has particular application in several respects. For example, if Commerce determined in a sunset proceeding, based on the original investigation and any administrative reviews, that the existing countervailable subsidy programs had been terminated and that the likely net countervailable rate of subsidization was de minimis, Commerce normally would determine that there was no likelihood of continuation or recurrence of subsidization." (emphasis added)

5.325.
A closer look into this statement reveals that the US’s reasoning is, from a purely logical point of view, inherently contradictory. When the US DOC finds in a sunset review that the net subsidy rate is likely to be below its domestic 0.50 per cent rule, it makes a positive empirical finding that the subsidy is likely to continue at that level. But since this level of subsidy is not considered under US law worth pursuing, it draws the conclusion (which is literally a legal fiction) that there "will be no likelihood of continuation or recurrence of subsidization". In reality, however, this legal conclusion of no likelihood of continuation of subsidization is not empirically correct, because the subsidy will continue albeit at a level below the 0.50 per cent. Therefore, in reality this legal conclusion is equivalent to a finding that there is no causal link of material injury for the purposes of a sunset review under Article 21.3 of the SCM Agreement. For this reason as well, it simply makes no sense the US’s refusal to recognize the need to apply in sunsets the 1 per cent de minimis rule of Article 11.9, but continue to apply a 0.50 per cent de minimis rule under its domestic law.

Question 31

5.326.
The US’s response to question 31 begs the issue. In paragraph 44, the US states that it "does not calculate the present rate [of subsidization in a sunset review] because the purpose of a sunset review is to determine the likelihood of the continuation or recurrence of subsidization" and that this "necessarily involves a prediction of a government’s future behaviour without the discipline of a countervailing duty order in place". The issue is not whether DOC must determine the current level of subsidization rather than the likely future rate of subsidization in a sunset review. Instead, the issue is whether DOC may irrefutably presume that the likely level of future subsidization is the same as the rate of subsidization determined in the original investigation.
5.327.
Moreover, the US does not answer the basic question why it can conduct on site verifications and apply the full rigors of an investigation in the context of an administrative review but cannot do or does not wish to do so in the context of a sunset review. Attempting to predict the future behaviour of a government in no way prevents the US authorities from conducting a proper, fresh, new investigation to find out about the numerator or denominator of the exporters’ sales, the likelihood of the subsidy programme been modified or withdrawn or whether there is a non-recurring subsidy the amount of which has already been expensed and the remaining benefit flowing from it is nearly zero.
5.328.
Furthermore, the US Sunset Policy Bulletin (see Exhibit EC-15) provides in Section III.A.5 some criteria about the way future government behaviour could be predicted. They include in particular the legal method by which the government granted or eliminated a programme, i.e. whether this is done by administrative action or through legislative action. The US Sunset Policy Bulletin proposes to the authorities to draw different conclusions depending on the legal method applied. This supports clearly the arguments the EC made in paragraph 34 of its Second Written Submission. This is all the more important here, as the programmes under review in the present case and, in particular the CIG programme, were granted by legislative action and could not as such be reinstated by administrative action only.
5.329.
In addition, the improper nature of the irrefutable presumption used by the US DOC is clear from the facts of this case. In this proceeding, DOC determined